hacker news with inline top comments    .. more ..    14 Apr 2011 News
home   ask   best   8 years ago   
Introducing Prompt. Nice SSH for iOS. panic.com
128 points by taylorbuley 2 hours ago   59 comments top 14
points by jrockway 1 hour ago replies      
So, as a UNIX hacker, here's what I don't get about the iOS community. There is way too much excitement over the simplest things ever. Let's break down some of the copy for this app:

Prompt is a clean, crisp, and cheerful SSH client

What does that mean? What does a clean ssh client do; not commit any protocol violations? What is crisp? When you bite into it, it's like a ripe apple? What does cheerful mean? Is the ssh client really happy that it can make a TCP connection to the intarwebs, especially for me?

What does this sentence mean to someone that wants to ssh from their phone?

it helps you when you need it, and stays out of your way when you don't

So when I want to be sshing, I can run it, and when I don't want to be sshing, I don't have to run it? Splendid. It works like every other computer program ever made.

Perfect for system administrators, web developers, movie-style hackers (“Let me just TCP/IP into the UNIX port!”), or any person who needs to connect remotely and type some magic.

So I googled for "ssh client", found your page, and you're telling me who uses ssh clients. How would you get to this page without knowing what ssh is, and how would you know what ssh is without needing to use it?

Anyway, a lot of happy-sounding words for ... a program that decrypts text from the Internet and writes it to the screen.

points by jimwise 1 hour ago replies      
I'd been using touchterm for a while, and had experimented with other iOS ssh clients. They... worked, but the experience always felt a bit clunky.

I've been using prompt for a day now, and so far it feels better than the others. The app design feels really clean, the method of expanding the keyboard to handle modifier keys and frequently-used non-alpha keys works well without grabbing too much screen real estate (this is clearly visible in the screen shots for what it's worth).

Finally, the terminal emulation has been flawless for me so far. Emacs runs well (and is quite usable with Prompt's modifier key placement, unlike in other iOS ssh clients I'd tried; remember, folks, ESC is Meta, Meta is ESC). touchTerm had some screen lag/partial refresh issues for me with Emacs -- these may have been fixed in later versions, though, as I gave up at some point.

In short, I like it.

obDisclaimer: I don't know the developers. I don't have a dog in this fight. I do like the app. I'm not the only one who likes it though -- see the daringfireball take here:


points by ceejayoz 1 hour ago replies      
Anyone got any ideas on getting a .pem key from EC2 into this? I got my non-EC2 id_rsa into it just fine, but .pems seem to disappear.
points by doron 28 minutes ago replies      
Slightly off-topic
I couldn't find a free SSH Ios app.

not that 4-5$ is a big deal, but there are myriad of ssh clients for free on android, no doubt due to different clientele, and different developer sentiments.

points by ezy 1 hour ago replies      
iSSH is still better. Especially on an iPad, and most definitely if you use a bluetooth keyboard once in a while [1]

Furthermore, I don't understand how someone could even make the remark that an ssh client reminds them of linux, and mean it negatively. :-) I certainly don't want an the equivalent IOS SSH client to the one Apple makes for OSX -- Terminal is rather shitty.

[1] BT keyboards on ipads don't support CTRL as a modifier for all keys, you have to create a workaround. So far, iSSH is the only app I've seen that has done it (By using option and remapping).

EDIT: Ga! Read the wrong tab.. was responding to Prompt and Grubers comments on it (re: linux), but this link points to the actual app. Just clearing that up... sorry.

points by rauljara 1 hour ago replies      
Anyone downloaded this? Anyone have a sense of how it compares to any other ssh clients?

I was super excited when I downloaded TouchTerm, but it was just so difficult to do anything with it on my iPhone that I gave up, despite thinking it was very clever. The Prompt screenshots at least look a lot cleaner than TouchTerm's, but I don't want to go through another fit of excitement only to never actually use it.

points by justinxreese 1 hour ago replies      
It's definitely one of the prettier SSH apps, but it faces the same barrier I've seen with other SSH apps - typing is too slow! It will sometimes be a full 2 seconds between when i finish a word and it starts appearing on the screen.

Has anyone else noticed this or found one that doesn't do this? I'm doing this on iPhone, maybe the iPad is better, but I don't see any reason why.

points by krosaen 56 minutes ago replies      
A bit off topic but anyone know how they do the fancy header with the pushpin notes swaying forward as you hover over them?
points by plusbryan 1 hour ago replies      
Hey @panic - I love the shiny new products, but could you pretty please update Coda a bit? It used to be my favorite editor, but the lack of git is killing me!
points by twism 1 hour ago replies      
Has anyone tried GNU Screen on this yet (before I plunk down the $5)? Thanks.
points by lordlarm 1 hour ago replies      
The problem with ssh on iOS and then especially iPhone is the size of the screen. When I'm eg. connect to irssi I can maybe see one line of previous conversations.

And regarding 'Prompt', I dont see how Panic has made something 'new' here. There are several ssh clients with favorites, shortcuts and nice designs. Some are even free.

points by alexg0 1 hour ago replies      
How is this different/better then iSSH?
points by askedrelic 1 hour ago replies      
Does this have 256 color support? Anyone know of an iOS Terminal with 256 color support? Too much to ask for in a terminal? Heh.
points by chmike 59 minutes ago replies      
This looks like spam to me.
Rent the country of Liechtenstein for $70,000 a night with Airbnb airbnb.com
26 points by jamesjyu 54 minutes ago   4 comments top 4
points by furyg3 0 minutes ago replies      
Ahh Liechtenstein. Random trivia: the Swiss once invaded the country by accident and nobody noticed.


points by joejohnson 18 minutes ago replies      
Liechtenstein is 160 km². That's only $437.5 per square kilometer!
points by gwern 4 minutes ago replies      
April 14th. Hm.
points by joejohnson 21 minutes ago replies      
It always strikes me how nice their website is. I really like the design and layout.
VMware CloudFoundry: Ruby powered PaaS igvita.com
27 points by igrigorik 1 hour ago   1 comment top
points by timf 1 minute ago replies      
Also see here for more on the emerging architecture details: http://blog.rightscale.com/2011/04/14/cloud-foundry-architec...
North Carolina cities mobilize against anti-muni broadband bill arstechnica.com
71 points by evo_9 3 hours ago   1 comment top
points by _delirium 3 hours ago replies      
One of the more active previous discussions of this controversy, fwiw: http://news.ycombinator.com/item?id=2328116
Simple algorithms openmymind.net
80 points by taylorbuley 4 hours ago   21 comments top 10
points by baddox 4 minutes ago replies      
I like the writing style and simplicity of presentation. I would recommend putting some thought into the order and organization of the articles. Perhaps you should have "main" articles about data structures, then sub-articles about the algorithms that are relevant to them (e.g. "binary search" could be under "arrays," "heapsort" and "priority queue" operations under "heaps," etc.). Obviously, it's a challenge to choose the order and organization of topics and subtopics"it's essentially the task of developing a curriculum.

If you plan on doing some tree/graph algorithms, perhaps you could have a brief introduction to the topic by talking about trees and graphs in general, then proceeding by discussing heaps, simple binary trees (which can branch off into more advanced topics like the various balanced binary trees), and so forth.

As a side note, I think binary trees are a great visual way to introduce the concept of asymptotic running time in a more accessible/pragmatic (albeit less rigorous) way, by showing that the more balanced a binary tree is, the fewer steps it will take on average to find an element (approaching the best-case of log base 2 of n). You can show how a worst-case unbalanced binary tree degrades to a linked list.

points by jcampbell1 1 hour ago replies      
A fantastic set of articles. I do think that tutorials in general obsess over sorting too much. Rather than expanding this with dozens of sorting algos, it would be nice to see a treatment of trees and graphs. Lay the foundation for someone to understand search trees, huffman coding, A*, etc.
points by tszming 2 hours ago replies      
points by nikolaplejic 3 hours ago replies      
I really like the simplicity and the choice of language. I think a comments / discussion section would be useful, for people to ask questions, talk about the ways to make the articles even better and perhaps translate the code to other languages.

All in all - I hope you keep up the good work, solid tutorials like these make it more compelling to keep up with the basics and learn new things from the "CS 101" department.

points by mfonda 1 hour ago replies      
I really like this idea, thanks for putting this up. I think it would be nice to additionally split it up into a data structures section and an algorithms section.
points by nickconfer 2 hours ago replies      
I had this same idea the other day. Glad to see someone made this. I hope you add more content in the future.

From a teaching perspective I think it would be great to see the math behind this as well to get the worst case scenerio.

points by thurn 55 minutes ago replies      
Is this open source? Nice visuals.
points by giltotherescue 2 hours ago replies      
Yes! Thank you for putting this together.
points by olragon 34 minutes ago replies      
you are my sunshine
points by Rickasaurus 1 hour ago replies      
Oh look, it's 200 level CS algorithms.
Voyager 1 is on the edge, and so is he latimes.com
85 points by edw519 4 hours ago   11 comments top 6
points by wingo 2 hours ago replies      
What wonderful writing this was. I have been really pleased with some of these more pop articles recently: the Dan Plan one and the Snider huffpost piece being two examples. Very pleasant to read between hacks.
points by vbtemp 3 hours ago replies      
I once found a manual describing the machine code for the Voyager computers (can't find it anymore). Recently there was a flipped bit in memory which caused serious problems:

One of the things I find most fascinating about the mission are the RF systems -- the antenna on voyager is approx 40 watts yet the ground systems can read something on the order of kilobits per second -- and they are billions of miles away.

This is an interesting read too:

points by meric 2 hours ago replies      
We're where we are today because of pioneers of previous generations. My eyes always water a bit when I think about how much they would have wanted to see what we have now.

Beautiful story.

points by checker 3 hours ago replies      
What a magnificent story. It's amazing to me how he he and the spacecraft are so connected. The radios go off, he works on other projects, until one day they are reunited as both reach the ends of their journey, transitioning to whatever lies beyond.

I'm saying that last part without hyperbole, which is beautiful to me.

points by markbnine 34 minutes ago replies      
Here's a tool to check out some of the raw imagery from Voyager (if you can figure out how to use it. . . the tool is meant for planetary scientists). Check out Jupiter/Saturn. The next dump should get have more stuff from Uranus/Neptune. http://pilot.wr.usgs.gov
points by iwwr 2 hours ago replies      
While the Sun's radiation extends up to 100au or so, its gravitational influence goes to over 100K au (2 light years). In this deep freezing wasteland we may find many more orbiting bodies than in the solar system proper.
Garry mods Garry's Mod to catch pirates joystiq.com
20 points by iughugh 1 hour ago   3 comments top
points by jrockway 18 minutes ago replies      
So, it's a good idea to punish people that are trying out your software by banning them from purchasing it? Sounds like the A/B testing phase was skipped on that one...
Scala + Mozart/Oz ac.be
25 points by swannodette 2 hours ago   7 comments top 6
points by lars 2 minutes ago replies      
Oz is a pretty damn cool language. But I think another thing that's hampering, apart from syntax, its adoption is the lack of available libraries.

I also have to add to the praise of Concepts, Techniques, and Models of Computer Programming. I remember in particular an exercise from that book where you were asked to write a simple parser in a particular paradigm. It basically asked you to write the procedure parse(code, syntax_tree), where you'd give the code as the first parameter, and where syntax_tree was the output parameter. Once you had made that, the last sub-exercise said "Oh BTW, try calling your function with an empty code parameter, and an existing syntax tree." Lo and behold, it worked! My parser could run in reverse, and it spat out all strings that would produce the given syntax tree. This was a mind blowing experience at the time.

points by jimwise 1 hour ago replies      
For more on the "Functional Patters are Concurrency Patterns" idea, see Bob Harper's recent post on the subject:


(The whole blog is worth reading; Harper is basically giving a blow-by-blow as CMU rolls out it's new functional-programming-first CS curriculum.)

points by rdtsc 43 minutes ago replies      
points by doublec 31 minutes ago replies      
Wasn't functional programming with a Mozart/OZ Engine basically tried with AliceML:


I'm not sure syntax is the main reason for slow uptake of Mozart. In fact the syntax can offer some advantages. See this post by pvr about a tail recursive append without helper functions:


Chris Rathman followed up with an AliceML version and compares the syntax:


I like Mozart/Oz but the two main things missing for me are unicode support and 64 bit support.

points by andrewcooke 42 minutes ago replies      
pvr doesn't give up, does he? :o) i hope this one finally hits the big-time (but i worry he's underestimating the complexity of scala and over-estimating its popularity).
points by Arrgh 40 minutes ago replies      
Jonas Bonér has been a big fan of Mozart/Oz for a long time, so naturally he rolled some dataflow concurrency support into Akka: http://doc.akka.io/dataflow-scala
XOR patent ended CD32, and Commodore-Amiga xcssa.org
116 points by robin_reala 6 hours ago   35 comments top 6
points by rbanffy 4 hours ago replies      
Erm... I see XOR for blinking cursors is obvious and should never be patented, but to blame only that for Amiga's demise is... a bit of an exaggeration. Management deserves most - in fact, almost all - of the credit for driving Commodore into the ground.

It's such a shame. Commodore made great computers.

The Amiga had sort of an identity problem. It was born a videogame console and NTSC timing was pervasive throughout the system and that made the design more complex as the machines evolved. They should have gotten rid of that as soon as they launched their second-generation machines.

points by jefffoster 4 hours ago replies      
points by Sodaware 5 hours ago replies      
Here's the patent in question: http://www.google.com/patents?vid=USPAT4197590
points by al_james 5 hours ago replies      
Although I used to be a huge Amiga fan, I think that Amigas of the time lagged behind other systems for gaming. PC's having chunky display formats (as opposed to plainer (planer?) on the Amiga) meant that PC's could do far more 3D graphics than a equivalently powered Amiga, and Amigas were often underpowered anyway. The Amiga kicked ass at 2D games, but could not hack it in 3D.

At the end of the day, had the Amiga been a viable business, $10M in patent fines would not have killed them off.

points by brlewis 4 hours ago replies      
It's easy for HN to see that the XOR patent is obvious, and not just in hindsight. Convincing a court is not so easy. The solution is not to get rid of quote-unquote bad software patents, but to adhere to the Supreme Court precedents that make software per se non-statutory. Any patent whose novelty and non-obviousness is claimed only in software should rightly be invalid.
points by EdwardMSmith 4 hours ago replies      
Hmm, I wonder about the accuracy of this, specifically about CD32s not being imported.

I worked, at that time, for the largest Amiga dealer in the US, and we built a couple hundred multimedia display kiosks using CD32s and Paravision SX-1s. We had no trouble getting them (well, other then the usual Commodore supply issues of the time).

Oh, wow, found an old email about this:



Google Translate, Now With Voice Input googlesystem.blogspot.com
14 points by Kylekramer 1 hour ago   3 comments top
points by dstein 33 minutes ago replies      
I wonder why Google is doing it like this. Sending an audio file of your voice over the wire seems like unnecessary overhead. From an API perspective I'd much rather have client-side speech-to-text built into the browser (accessible via a JavaScript API). I think they're already doing STT client-side in Android, so what's the hold-up with embedding it in Chrome and letting web developers go nuts with voice-enabling their web apps?
Collection of documents that startups commonly need: Privacy Policy, NDA, etc... pearwords.com
112 points by x03 6 hours ago   21 comments top 5
points by duck 5 hours ago replies      
I like the idea, but I just don't see startups needing or using an email disclaimer. Plus they are pretty much worthless legal wise - http://www.economist.com/node/18529895.
points by woodall 4 hours ago replies      
I know you are not offering legal advice, which IMO is good, but has the language in any of these documents been looked over by a profession/practicing lawyer? Other than that, these are great; I'll be using the Privacy Policy and NDA.
points by blhack 2 hours ago replies      
This is excellent, thank you to whoever is putting it together.
points by jamiecurle 4 hours ago replies      
The privacy policy states

  PearWords does not...
Place "cookies" (small text files) on your system for any reason.

I would beg to differ, it may be Google Analytics dropping the cookies on pearwords' behalf, but they're still there. (http://d.pr/D3oy)

Portland Seed Fund seeks inspiration from Y Combinator grads oregonlive.com
9 points by turoczy 40 minutes ago   discuss
Pictures of the first GUIs from Xerox digibarn.com
97 points by coliveira 3 hours ago   14 comments top 10
points by thought_alarm 1 hour ago replies      
There are a few demos of the Xerox Star floating around on YouTube. Seeing it in action gives you a much more complete picture of the system.


The "look" was obviously hugely influential, the "feel" not so much. It is quite an odd beast, as the mouse is actually only used for selecting objects.

The Lisa GUI prototypes are also and interesting bit of GUI trivia. It's easy to see the Star's influence on the final shipping version of the Lisa GUI compared to its prototypes.


points by mcritz 36 minutes ago replies      
This is amazing! It's hard to imagine these images are real given how complex and beautiful the UI design is.

· Multiple fonts.

· Multiple, simultaneous languages.

· Dithered graphics to simulate value, tone, and shading.

· Rounded buttons.

· Use of line-weight to simulate dimension.

points by jgrahamc 2 hours ago replies      
I'm old enough to have used these machines and what's interesting is that I remember well how much of a 'wow' it was to get to use a lovely user interface like that. Since then I haven't seen a real step change in user interface until we moved recently to sensitive touch screen devices.
points by rbanffy 2 hours ago replies      
I'm impressed by the utmost attention to details. I remember similar care when working with educational software for Apple IIs - we limited our 50% checkerboard pattern to 279x191 (instead of the 280x192 maximum) in order to be able to do exactly the same rounded corners on all four corners of the screen.

At times, the guy who came up with these ideas infuriated me, but, in hindsight, I am very glad I surrounded myself with such perfectionists.

points by bostonpete 2 hours ago replies      
Very impressive. One inconsistency stands out given the careful attention to detail. In this image, they make a point of emphasizing that background pixels should be split to form a cleaner edge:


...but this image shows that most of the top of a folder icon (not including the tab) did not split the background pixel:


points by epenn 2 hours ago replies      
Actually the first GUIs from Xerox were on the Alto, which the Star is based on (see: http://en.wikipedia.org/wiki/Xerox_Alto). Although this is certainly from the first that was available commercially.

Nonetheless, these screenshots are amazing and show how groundbreaking the GUI concept was at the time.

points by Luyt 1 hour ago replies      
From the same site, stories from back then:

"My Cajun country upbringing had never taken me any further west than Dallas. And since I wanted to make a good first impression on my new California friends, I purchased a spanking new three-piece navy blue polyester suit, super-wide ‘70s tie, platform shoes and the finest imitation naugahyde briefcase I could find and made my first reservations at Rickey's Hyatt House.

I arrived at the lobby of PARC, resplendent in polyester and cheap Old English cologne, and was met by Charles Irby… ponytail, scruffy beard, tie-dyed t-shirt, khaki shorts and Birkenstock sandals. He welcomed me warmly, and then took me around to meet the eclectic cast of colorful characters and future luminaries that made up the Star development team. As we toured the offices, and the more folks I met, and the more beanbag chairs I saw, the more conspicuous, foreign and puritanical I began to feel… a penguin in the company of parrots. And yet, I was embraced and welcomed into this cadre of characters. It would not take me long to assimilate."

points by rplacd 3 hours ago replies      
The typography demos are oddly tasteful. http://www.digibarn.com/collections/screenshots/xerox-star-8... seems like a dig at the Lisa - it does seem woefully inadequate there for DTP.
points by rsbrown 2 hours ago replies      
Wow, I see the Mac, Amiga and other next gen UIs that came soon afterwards in these. I'm also struck by how much more appealing these screenshots appear than the early versions of Microsoft Windows.
points by elliottcarlson 2 hours ago replies      
Seems to be down for me - gotta love CoralCDN: http://www.digibarn.com.nyud.net/collections/screenshots/xer...
IGDA about the Amazon Android Appstore: just say no. pastebin.com
111 points by swombat 6 hours ago   52 comments top 16
points by msy 4 hours ago replies      
Say hello to the new boss, same as the old boss.

Amazon, Apple et al are becoming new middle men of the digital marketplace. Want to sell content online? If you want mass market exposure we'll have 30-70% of the revenue from the sweat of your brow and you better thank us for the privilege.

I wonder how long until they band together and form a lobby group called the Digital Content Providers Association of America.

points by jcl 3 hours ago replies      
I'm surprised the IGDA didn't also call out Amazon's apparent requirement that a developer must allow Amazon to distribute all the software the developer is distributing through other channels. In other words, there's no way to try out Amazon's service with just one title -- it's all or nothing.
points by _delirium 5 hours ago replies      
Sounds like Amazon is trying to extend to apps the weird pricing racket that's standard in the bookselling industry. A lot of books are cheaper via Amazon than via the publisher's or author's own site precisely because of these kinds of terms, which require you to sell to the bookseller at a steep discount over list price, and simultaneously prohibit you from selling below "list price" yourself, outside the bookseller channels.
points by aristidb 5 hours ago replies      
Amazon does however have an incentive to maximize revenue and there is a lower bound to what a developer gets (a percentage of the list price), even if Amazon promotes it for free, so Amazon loses money on that.

I think Amazon is better at optimizing pricing than most game developers are, so I expect this to improve revenue for both Amazon and the game developers. But only time will tell for sure. Panicking is not the right response.

points by allwein 3 hours ago replies      
I'd need to read the Amazon agreement much more closely, but what's to stop developers from gaming this by taking all their apps released on Amazon's app store and making them special "Amazon Editions"?

Then the argument is there that the minimum list price on Google's store doesn't apply, because the Google Edition and the Amazon Edition are two different apps.

points by dazzla 2 hours ago replies      
From my limited experience (a friend's and my own app in the Amazon store from the first week) the exposure is pretty negligible anyway. I imagine the requirement to enable side loading (not possible with AT&T), downloading another store app, enabling 1 click, etc is really hindering adoption. So I don't think developers need to loose any sleep over this right now.
points by gacba 2 hours ago replies      
Three words people: In App Purchases! If you have a free app, and then use In App Purchasing for your real revenue stream, you won't be affected by these Orwellian terms and Amazon's whims.

It's relatively new on Android, but a number of iOS apps are raking it in this way and giving away the app for free...


points by alvarosm 4 hours ago replies      
points by yawn 5 hours ago replies      
It is curious that there haven't been any prominent developer blog posts about how well an app has done on Amazon's store. Where are the developer comments?
points by shareme 5 hours ago replies      
Just wait folks..remember most device users have do one extra step to get these apps..if we pressure Amazon enough these terms will change..

In my own case I am going with:

1. Android Market
2. SlideMe
3 GetJar

..when Amazon changes the terms as it should be pricing only
at their market pricing at other markets should not bear on pricing at theirs..its a anti-trust potential violation via the collusion implications of anti-trust law..

points by joelackner 5 hours ago replies      
does anyone know how the free app-a-day promotion is being handled? does amazon pay the developer 20% of the price for each free download?
points by fomojola 3 hours ago replies      
Until Amazon fixes that, there is quite frankly no incentive to use their store for anything that isn't free. I had thought that cross-promotion for software that works with hardware Amazon sells would be win, but until Amazon fixes their pricing policies I've no interest at all in their store. 30% off the top is bad enough but at least you got to determine the price. 20% of list price on a 99-cent app? Really?

FAIL. Call me back when you get your mind right.

points by kodisha 4 hours ago replies      
All i see is that most of the top selling apps are already on Amazon Store :/
points by Shorel 5 hours ago replies      
Any Indie developer with experience on Steam sales to light up this issue?

Seems to me that the arguments of IGDA sound very close to the arguments of the RIAA and real numbers will show a very different outcome, specially for the smaller devs.

points by tybris 3 hours ago replies      
or... be an entrepreneur and take a chance...
points by neutronicus 5 hours ago replies      
Kind of paints the Apple / Amazon spat in a new light.
Listing "customer personas" instead of features improved our CTR by 200% brandregard.com
55 points by geirfreysson 4 hours ago   19 comments top 11
points by arnorhs 2 hours ago replies      
Awesome to see an Icelandic startup on the HN frontpage!

This post is a great example of how applying typical established marketing knowledge to a product presentation can improve things a lot.

It's well known that presenting benefits, use cases and personas is a more successful approach than presenting features and screenshots. - For most users - I thought every A/B test in the world had already proven that.

And yet we have all these startups out there presenting nothing but features. I think presenting features might probably work better for techies and more tech-savy people, but for the great majority stuff like this will always work better.

Great job. Iceland ftw!

points by Zak 2 hours ago replies      
I see this sort of thing a lot on sites for products trying to sell to large businesses, sometimes to the point that it's hard to find an actual product description or list of features. I have, several times wanted to yell "but what does your product actually do?" at the person responsible.

I am obviously not the target market here, but why do you think this works?

points by capstone 1 hour ago replies      
EasyDNS does this beautifully:


points by geirfreysson 4 hours ago replies      
For anyone interested we used A/B tests, powered by the seven_minute_abs Rails plug-in, to measure the results.

I'd be interested to hear about other people's experiments both with A/B tests and customer segmentation.

points by toddh 3 hours ago replies      
I like the idea, but you have to pretty sure you know exactly who your customers are don't you? I would move on if I wasn't on the list. Features are more universal, which I guess is why they are also less effective.
points by Darraghb 3 hours ago replies      
Nicely done.

You will also get a much better indication on the type of audience your site is attracting, enabling you to tailor content and better position your offering.

points by pkamb 1 hour ago replies      
I wonder if any of it has to do with better results for someone typing "XXXX for creative directors" into Google.
points by delinka 2 hours ago replies      
I was almost amazed at the correct math. A "200% improvement" is definitely "three times what it was previously." But I'm not entirely convinced +200% is a "threefold increase" (I'm a math guy, not a 'human language is fuzzy and ambiguous' guy.)


points by aginn 2 hours ago replies      
This is kind of what Apple does. Apple sells the vision and value of the product, not the features and functionality. Apple wants people to buy into the vision
points by chopsueyar 2 hours ago replies      
This is very interesting. In personal selling (face-to-face), emphasis tends to be on features and benefits.

The personas work better. Proof is in the numbers.

points by kfkristjansson 3 hours ago replies      
Love the system
Interview with Rich Aberman of WePay founderly.com
20 points by viking 1 hour ago   1 comment top
points by viking 1 hour ago replies      
“Be arrogant enough to get in the ring and naïve enough to think you can win after you feel that first punch.” (Go to 1:53 in the video) - Rich Aberman, WePay Founder
Analysis Of Apple's A5 crunchgear.com
16 points by flippyhead 1 hour ago   1 comment top
points by jonknee 41 minutes ago replies      
This is blog spam. The interesting stuff is here:


Is Scheme Lisp? c2.com
6 points by gnosis 24 minutes ago   discuss
Game makers says Amazon's Android appstore terms are greedy venturebeat.com
29 points by shawndumas 3 hours ago   8 comments top 4
points by statictype 46 minutes ago replies      
Something I'm missing here:

Amazon will either pay the developer 70% of the actual purchase price (which is set by Amazon) or 20% of the listed price (set by the developer), whichever is higher.

So what's to stop a developer from listing the game at $100 and letting Amazon figure out for itself, what a more sensible price should be (and then collecting $20 which would be higher than 0.7*whatever_price_amazon_finally_sets)?

Will Amazon reject apps on the basis of the listed price?

points by samgro 1 hour ago replies      
It would be interesting if app stores worked like retail stores. Developers sell apps to the store at their desired price, and the app store sells it to a customer, usually with an added markup of 30%. If Amazon wants to sell a popular app as a loss leader Walmart style, they are free to do that without screwing the developer.
points by neuroelectronic 1 hour ago replies      
Build your own App store?
points by greyfade 2 hours ago replies      
Given how well the Humble indie Bundles have done in the past (and how the current one is doing), I think their concerns over pricing are a bit overblown.
Does the CIA own part of Facebook? stackexchange.com
5 points by rjstelling 20 minutes ago   1 comment top
Joel Spolsky on allocating ownership in your startup onstartups.com
317 points by _pius 14 hours ago   61 comments top 17
points by ghshephard 4 hours ago replies      
Joel's article is pretty good as a starting point, but, I think there is a lot of variation on what the first set of employees get.

I've been a first 10 employee (As an infrastructure contributor, not core engineer) twice in companies that eventually were valued at greater than $1 Billion. The first time I received 0.03% Equity (Before Dilution) - the second time I received 0.1% Equity (Before lots of dilution).

For one of those companies, I know that some of the core engineers received 3-4x what I did, so - extracting to all of the six core engineers in Layer 1, Plus the Administrative crew - comes around 6 * .4% + 3 * .1% = about 2.7% for the first nine employees. We had our series A before anybody came on board, as an employee.

There is probably a different allocation method for teams comprised of "Serial Entrepreneurs" - Your risk in joining that team is much less, so your equity is typically much less. Also, the approach usually goes like the following:

Step 1: Two - Four Founders create a company. Roughly sharing the equity, though, if there is a "Named" founder that will Garner Press/Financing/Customers, they take a bigger chunk.

Step 2: Founders brainstorm for month or two, commit to working together for a minimum of four-five years, and then go pitch their preferred VCs. VCs give them a valuation of $5-$10mm (pre-money) and invest $1mm-$2mm.

Step 3: First 5-10 Employees are hired, with a stock pool of 3%-10% - Sr. Employees with a great track record who currently have great jobs at Google/Facebook/etc.. will require a larger equity share. Out of work contributors who have a solid, if not exceptional track record will receive significantly less. The team now has a clock ticking, and has to demonstrate some traction within six-nine months to get their next round before the money runs out.

points by agmiklas 11 hours ago replies      
I didn't think his IOU solution for founders that either don't take a salary or contribute property made much sense. For that to be fair, you'd have to set a super high interest rate on the loan.

At the same time, I see the difficulty with assigning a concrete value to the shares early on. The angel investment world solves this exact problem using convertible debt. Why not take the same approach with investments-in-kind made by the founders?

If a founder forgoes a salary, why not agree to convert the pay difference relative to the other founders into stock at the time of the first equity financing at the share price negotiated with the VCs?

points by gyardley 34 minutes ago replies      
Holy good lord, that's a lot of equity for employees.

I can see this causing all kinds of problems. You're not going to allocate an option pool for employee layers one through five all at once, prior to your seed round, because that'd be massively dilutive to you in the event of an early sale. (The unused options go away, but the premoney the VC invested at takes the unused options into account.) But creating such hefty option pools down the road is going to cause issues with your existing investors, who at that point would be diluted.

The conflicting interests of founders and earlier investors (who don't want to be diluted by a large new option pool) and later, new investors (who want to make sure the company has a lot of options to incent new employees) will get you to an 'industry-standard pool' pretty naturally. Unless the market's changed dramatically recently, that standard pool is a hell of a lot smaller than what Joel's suggesting.

points by ookblah 11 hours ago replies      
I come from perhaps the small subset of a being good friends w/ my co-founder, having a 50/50 split, and being the technical one ....so that sets up context for my thoughts.

I resonate w/ this article a lot because to me, the appearance of fairness trumps everything. The 50/50 split lets me know that I don't have to worry about who does what exactly or who is working harder, but sets it up so we are both "all in". I feel like if you're debating equity split at that stage (provided you're both at the same point, quitting your job, etc) you're already setting up a rocky relationship. Either that or you're not really finding a co-founder, more of a dedicated employee.

I guess I'm just a little unclear on how you can define clearly what a "60/40" workload split looks like when they might not even be the same type of work.

points by alain94040 13 hours ago replies      
1. Use http://foundrs.com to split equity early on, before your project gets traction. It has vesting built in. And it forces co-founders to have that oh-so-feared discussion early.

2. I respectfully disagree with Joel on certain aspects. He is very unclear about how to split equity among a few founders. He seems to advocate 50/50, which I strongly advise against. Fairness is one thing. But my litmus test is: if you quit, would the project die instantly? Then you are the CEO and you should get more.

I have advised tens of founders on those issues, including convincing some to fire useless co-founders. It's painful, but usually pretty clear when an outsider (like me) listens to all sides.

points by limist 13 hours ago replies      
A 50/50 split can work for the (very) small subset of cases where two people of comparable skill and commitment start building something from scratch (nothin' but a half-baked idea) together, with no prior investment/work/IP, no domain expertise, no key contacts, no customer channels, nor any major capital infusions. Oh, and both parties have a clear record of making good decisions together and resolve disputes effectively.

But for most other cases in the real world, these two resources offer a more rational and open/honest approach:



points by Murkin 6 hours ago replies      
10% for the first 4 employees (paid?).

Anyone has example of a startup that actually did that ?

AFAIK the total employee pool is rarely beyond 15% total. And that number is for a few layers ahead.

points by Chocobean 14 hours ago replies      
His answer for "What happens if not all the early employees need to take a salary? " makes sense, but leaves the question of "why don't I get paid now instead of getting paid later if it means I get just as much (or less due to inflation)." Presumably, you'd have an understand co-founder who understands that cash in the company now is a little more important. Failing that, I think it might be fair to add interest to that IOU.
points by ozataman 3 hours ago replies      
Good article, but leaves a few points open in my mind. The biggest issue that most people seem to be missing is control. When you have a 50/50 setup, you better make sure that you are comfortable with all the decision-making dynamics that partnership structure will bring.

Consider when you have a technical and a semi or non-technical founder. Let's say the technical founder is the visionary for the product you are building, understands what the customer needs (which means he has some/good business savvy) and give direction to all engineering related activities - from technology stack selection to what features and how they will be built. He/she is the one who will take the lead in defining the product and its -hopefully- many iterations going forward. A quick high profile example that comes to mind is Mr. Zuckerberg - he did the programming and he continues to give direction to the product.

Now you also have the semi/non-technical founder, who is obviously there because he/she is talented, smart and will have large impact going forward in building customer relationships and contribute to higher level discussion on where the products should go. There is a good chance the company won't go anywhere without this guy either.

How would you now do the split? A 50/50 arrangement would mean both parties get the same say/leadership over where the company/technology needs to go. Is that right? This is not about money, as both parties will be in good shape as long as a reasonable arrangement is chosen. It is about what is fair/right/sensible regarding what the company is going to be about and how it is run.

points by bravura 10 hours ago replies      
What do you do about disputes in the case of 50-50 ownership? In particular, what do you do if one founder wants to fire the other founder?

I've heard of a so-called "shotgun clause". It's analogous to the problem of fairly cutting a cake. One person cuts, the other person gets to pick a piece. IIRC, in the shotgun clause, one founder can demand that the other founder leave, and names a price for the company. If the other founder wants to stay, he can buy out the first founder for the named price. This sounds reasonable, except that founders might not actually have the money to buy out the other founder.

points by acangiano 14 hours ago replies      
> Otherwise your co-founder is going to quit after three weeks and show up, 7 years later, claiming he owns 25% of the company.

Or half the company for a $1000 investment.

points by ry0ohki 6 hours ago replies      
The general advice I've heard is you don't ever want 50/50 splits because if there are important decisions to be made, you can often end up in deadlock, and no one is truly in charge of making a final call or being responsible. Since I don't have enough karma on onstartups to ask Joel this, I'm curious what his response would be.
points by mkramlich 13 hours ago replies      
Sounds like reasonable advice. And I'm reminded of how good of a writer Joel is when in peak form. I'm also a fan of Inc. magazine and it's been great to see both him and Jason Fried contributing in print there as well.
points by gatlin 14 hours ago replies      
I'm working on starting a small worker co-operative. The advantages are normally considered for a large group (say, to increase buying power for interested consumers) but in the case of a lean startup, the law simplifies these questions. After bills and other fixed costs, you apportion net savings to the members proportionate to their contribution. Additionally, in a small group democratic (maybe even consensus) voting allows everyone to be equally in control of what is a joint partnership. I know Texas has laws covering "cooperative associations," can't speak for other states. Thoughts?
points by kchodorow 5 hours ago replies      
A lot of comments seem to be squabbling over details, but your startup is almost certainly going to fail, and the longer you haggle over splitting proceeds, the more likely failure is. Just split it and start working already!

If your company is a success, great, but is it really going to matter if you're worth 50 million vs. 60 million in the infinitesimal chance that it pops?

points by fedd 9 hours ago replies      
what about the advisors? some suggest give them stake.

Zuckerbergs need Sean Parkers so that VC would invest faster

points by tomjen3 11 hours ago replies      
This is great advice, except that he puts a yearly cap before the first vesting which means that the company is better of fireing you the day before you earn your shares than keeping you employed.

Neither you nor your employees need that kind of perverse incentives.

Anatomy of Google analytics cookies eduhub.nl
12 points by wasigh 1 hour ago   3 comments top 2
points by dude_abides 28 minutes ago replies      
Just wondering: Why doesn't a webserver encrypt the cookie value with a server-side key? This way, the information in the cookie (timestamps, pageviews, number of sessions etc.) can be safe from wireshark sniffers.
points by toddh 36 minutes ago replies      
Nice, I always wondered how they worked.
Poll: Display points on comments?
351 points by pg 16 hours ago   278 comments top 158
points by samdk 16 hours ago replies      
I think that not having points has some nice qualities, but it also feels like I'm being denied information that I find useful in reading/skimming a thread. I've noticed I find reading HN a lot harder while this has been in effect. (I've also noticed that I tend to unconsciously give numbers in usernames some weight when reading a comment. The same applies for the time it was posted.) One possible compromise would be to display either a number or simple graphic that approximates point totals instead of displaying them explicitly.

Also, I've been planning to write a longer blog post on the following, but given that I've had no time lately and am not likely to have any soon, I'll just float the idea here.

One idea I've had that I think might be interesting is dealing with upvotes or points in terms of logarithmic scales. That is, it takes one upvote/point to get a comment from 1-10, 2 upvotes/point to get from 11-20, etc. (Exact numbers would have to be scaled, of course.) I find that going into a thread an hour or two old and seeing comments with 50-100 points is a major disincentive to commenting, even if I have something to say. That comment or couple of comments and their resulting threads are going to make sure very few people ever read what I've written. An appropriately scaled log-scale system might make it so that really really good comments still get really really high scores, but so that others (which might have simply come too late in the discussion to be competitive on a raw-point scale) still get a chance at being seen.

(One related idea would be to make the point-approximating graphic log-scale even though the points themselves remain the same underneath.)

points by patio11 12 hours ago replies      
I'm conflicted. As a (heavy) consumer of HN, not having the best comments called out to me has made it more difficult to parse, and not being able to sort searchyc results by points will eventually diminish its value as external memory for me.

As a producer on HN, I have noticed two things: one, the subjective self-assessed quality of the comments I have been writing since the change has been far higher than it was in the few months prior to the change. I've been quipping less and writing meatier, substantive, useful things (I hadn't stopped writing those, but there were periods of weeks where I had no comment longer than a paragraph and very few actionable bits in those comments). I do not know why this is -- it could be phases of the moon, totally unrelated to the interface change, for all I know.

I also note that my per-comment scores for meaty comments are higher than they've ever been, which may or may not be desirable. I don't care about karma, but to the extent anyone else cares about their karma relative to folks on the leaderboard, my anecdotal single-point observation is that winners seem to be winning at the moment.

points by edw519 16 hours ago replies      
For years I listened to users complain whenever I removed something they were used to for the improvement of the whole endeavor. I have rarely encountered a user that was happier with less information.

Now I know how they feel.

points by jerf 16 hours ago replies      
I think it's too soon to actually judge, as we're still in the "ick! change!" phase. Ask again in another week or two. Same for any other experiment you run in the future; unless it obviously and immediately fails, give it some simmer time. (IMHO, of course.)
points by tokenadult 6 hours ago replies      
I went back to the post where pg asked for advice on how to prevent decline of HN:


He wrote, "The problem has several components: comments that are (a) mean and/or (b) dumb that (c) get massively upvoted."

To help along the thinking process here, as we digest our own personal experiences (each from a different subset of threads, I suppose, unless several participants here read HN exhaustively), let's think about those issues:

a) After the change, are mean comments less likely to be upvoted?

b) After the change, are dumb comments less likely to be upvoted?

c) What is the general character of highly upvoted comments after the change? Are comments with the highest number of upvotes after the change usually helpful, thoughtful comments, or flippant comments that don't gratify intellectual curiosity?

Over the next few days, it should be possible to look at some highly upvoted examples. The bestcomments view of HN content


still shows highly upvoted comments, although right now it shows them without explicit comment scores. How do those comments look to you?

P.S. There is a lot of speculation in this thread about how comments are weighted, how users gain karma, etc. As far as I know, except for possible details of the current experiment, the source code for this site in ARC,


which was mentioned in an HN thread a while back,


tells any code-literate user most of the story about how karma is allocated and how users gain karma. (Please note that I am NOT a code-literate user here, not in ARC, and I have never attempted to reverse-engineer any aspect of the karma system here. I simply empirically observe what happens to my own submissions and comments after I submit them, which I can still do during this experiment by viewing my own followed threads.)

points by keyist 15 hours ago replies      
"My goal in not showing points on comments was to prevent the sort of contentious exchanges where people (in this case literally) try to score points off one another."

I think a suitable compromise would be to hide for X days until most voting activity is over (reusing the threshold where downvotes are no longer available but upvotes still are might work). This would still meet your goal without the cost of information loss.

EDIT: I'd hate to give up lists like http://top.searchyc.com/comments_by_points and http://top.searchyc.com/users_by_average_points_per_comment .

points by Terry_B 15 hours ago replies      
I thought it was working well without points until this morning.

With this ASK HN post (http://news.ycombinator.com/item?id=2444709
) about the best credit card payment method I would typically immediately go and look at the comment scores.

Because a vote for a comment is typically a vote for the thing being recommended.

It's really a "poll-like" question except it can't be a poll because the OP doesn't know the options in advance.

Other than that one problem though,I found I was judging people's comments more thoughtfully without the scores.

points by michaelochurch 16 hours ago replies      
I think the scores should appear to a user after (A) that person votes, or (B) 24 hours. Keeping the karma hidden to a user until he or she votes is, IMO, a good idea.

I'd actually argue for keeping the text color the same at 0 or -1, for the same reason.

I don't know that these "contentious exchanges" happen in practice. I remember that a year or two ago, there was the discussion of whether people with high average ratings deserve to have their names appear in orange, and my thought then was that making a fetish of karma averages discourages nested discussions, since root-set comments can end up at 20+, but 6th-level comments, no matter how insightful, are unlikely to get more than a single upvote.

points by blhack 14 hours ago replies      
Points should stay:

If a post has a lot of points on it, it's telling people that they should be paying attention. For instance, (and this may have been after the points disappeared, but the example still works) the other day when ioerror came into the thread that was talking about him getting harassed at airports. I don't know who ioerror is, but points allow people to call my attention to his posts, more so than just voting them to the top of the hierarchy.

I've learned a lot reading comments here over the last 3 years. Lack of showing points makes it harder to discern what I should pay attention to. A good comparison might be book reviews. If I get on amazon and search "iOS4 development", I'll get tons of results, but when a book has 200 5 star reviews, it helps me decide that that is the one I should read.

This is true even if the books are all free.

Don't get lost in the idea that everybody here is a seasoned veteran who knows everything about everything and can easily judge a post's merit based on its content. It's less true for me now than it was three years ago, but it was (and still is) helpful to be able to look to the community to help me know what I should be paying attention to.

points by skennedy 16 hours ago replies      
Without. The lack of points make me look closer at the content rather than group think of a post. Even if briefly, a brand new comment is at the top of a thread. Unless I read it there is no way of knowing the quality in comparison to other comments. Would be interesting to know if the average karma per post is going up/down for those that have been around for a while.
points by staunch 16 hours ago replies      
The problem is I feel no feedback for voting, so I've stopped doing it.

If I could see the points after I vote I would probably vote even more than before.

points by mcav 16 hours ago replies      
Maybe making it a coarser indicator would give the best of both worlds... maybe a colored indicator (or grey dot) with color/darkness indicating how many upvotes.
points by m0nastic 16 hours ago replies      
Not to try and "King Solomon" it, but couldn't you make it a user profile option to display comment points?

Personally, I think it's better now that they are not displayed, but if it reverts back to displaying them, I'd be happy to just not have to see them myself, regardless if other's want to.

points by entangld 16 hours ago replies      
HN isn't only opinions. It also gives advice. How would you HNers who don't like points get advice in an area you know nothing about? What indicators would you use? Whatever sounds the best? Seeing a cumulative score of the opinions of intelligent users helps me.

Perhaps some of you know much more than me and don't need to learn anything. I'm not in that boat. HN helps me learn how other entrepreneurs think and what they think about areas that I'm moving into. This isn't reddit. This is a serious forum and seeing which advice gets the most upvotes helps me tremendously.

points by ElbertF 16 hours ago replies      
• I'd like to at least get some feedback when I vote, now my vote seems to just disappear into a black hole. Perhaps make the arrow orange after I click it (or simply a tick, " voted)?

• It feels pointless to upvote the only comment in a thread, it doesn't affect anything unless more comments are posted (again, lack of feedback).

• I often can't tell if a comment is any good (e.g. an answer to a scientific question), points really helped here.

points by anigbrowl 16 hours ago replies      
With. [EDIT: oops, deleted half my post] If something stupid only has a couple of points in a busy thread, then I know there's no need to respond with a passionate denunciation of the obvious. Likewise, if something I disagree with has a ton of upvotes, perhaps it's me that's stupid and I should think carefully before starting an argument. Sure, it's flawed, but so's every other approach.

If you want to get real results, I think you need to start doing randomized trials of different users, showing karma to some and not to others to see whether it results in a change of behavior. Of course, you probably need to warn people about this in advance.

BTW, there seems to be a little bug wherein clicking on a comment or poll option no longer updates. I had to refresh to see whether my vote had taken or not, although the - correctly disappears.

points by makmanalp 16 hours ago replies      
What is the use case for seeing the points on a comment anyway?

All you need to know is that the comments on the toppish are the best ones, and as you scroll down you can stop reading whenever you feel like it's gotten too bad.

I find that without points I'm definitely more focused on the content and are less likely to consciously / subconsciously groupthink.

points by lukeschlather 4 hours ago replies      
I think HN would do well to look at how Slashdot scores its comments.

Especially, I think having a few pre-defined tags that people can give is very useful, and in many ways a lot more useful than votes. Votes, as many people have noted, are often off the cuff, and I'll give a +1 without too much thought. Calling someone insightful? A troll? I'm going to think about that a little harder.

So the big thing is increasing the average amount of time people spend before voting. Some AJAX trickery would be a little strong-armed. I think focusing on finding ways of filtering content that are more descriptive is a better direction to go. (Being more descriptive, I think, is the only way to force people to be more discerning.)

Though I don't mean to suggest precisely Slashdot's system. The tags should be carefully chosen. "factually incorrect" would be a nice one to have. (And is a little less insulting than "troll," making no assumptions about the intentions of the author.)

points by india 9 hours ago replies      
This is a nightmare. Please end this. There are 201 comments in this thread at this moment. I am not interested in reading 201 comments about this experiment. I am however very much interested in seeing what the top few most insightful comments in this thread are. That is a very important signal. Right now I am feeling blind as a bat...
points by sosuke 16 hours ago replies      
I seem to prefer HN without points because it forces me to actually look closer at the content instead of just skimming for the big numbers.
points by joshhart 16 hours ago replies      
Maybe you could have a color range indicating roughly how good the comment was? I like the feedback a lot the numbers gave but I think hiding the actual numbers would be a good thing so people don't play the numbers game too much.
points by Tangaroa 30 minutes ago replies      
My problem with showing the point score is that it can lead to a share of the score being due to popularity rather than quality. Some people will uprate comments that they might otherwise consider too marginal to uprate, if they see that everyone else is uprating that comment. More commonly, moderators are also readers and will only want to read the highest valued comments if they are short on time, and then these become the only comments that they issue any ratings for.

Even without point scores, a comment's relative score can still be inferred from the system floating high-scoring comments to the top while comments with negative scores are displayed in a shade of gray. Readers are given hints to what the community currently considers valuable, but moderators have to read the comment to be sure.

points by tokenadult 16 hours ago replies      
I think, based on the small subset of threads I have sampled by a convenience procedure, that not displaying minute-by-minute comment scores avoids the cognitive illusion human beings suffer from called anchoring bias



and helps readers focus on the inherent worth of a comment. That's my general observation from seeing which comments are floating up or down in threads and which comments are graying out.

points by codeup 1 hour ago replies      
I agree with the goal in not showing points. I think it works and the benefits outweigh the costs mentioned by others.

The quality of comments on HN has increasingly suffered from people behaving like in hives. Let the content of a comment decide if you agree, not the number of others who agree with the comment.

Similarly, let the content of comments in this thread have a say in the future of not showing points on comments, not just the number of votes for or against it.

points by dkersten 15 hours ago replies      
I've been getting more upvotes since the scores were hidden. Maybe my comments were better, or maybe groupthink was holding people back, but either way, my karma has increased wince the scores were hidden, so.. I'm all for keeping it how it is. :-D

Seriously, though, I kinda like it without scores, because I don't get inadvertently suckered into voting to go along with the crowd (I try not to anyway, but sometimes it happens without thinking about it), while now I only vote if I feel the comment needs it (ie, the comment is very relevant and informative (up vote) or off topic/rude/irrelevant (down vote)).

points by SkyMarshal 51 minutes ago replies      
I like it better without. I prefer not having the herd effect/anchoring effect of points pre-affecting my judgement of someone's content before I even read it.
points by hammock 1 hour ago replies      
One benefit of points displayed that is totally lost is the ability to compare the points of a comment with the /replies/ to it. This lets me get a general idea at a glance if a commenter was mistaken or incorrect or misinformed or whatever (this happens for example when a thread has one comment and one reply, and the reply has 100x more points than the comment)
points by cookiecaper 11 hours ago replies      
I'm late to the party so I doubt this will be read by many. I just want to say that I've found the experience in the last few days without a displayed score really interesting, both as a reader and a commenter.

My voting decisions in the past were based on the metric "How many points does this deserve?" If I found a good comment with 0-3 points, I was sure to vote up, and if I found a bad comment with 4+ I was sure to vote down. Normal comments or comments that had scores that were roughly where I thought they should be would be ignored.

I'm generally relatively sparse with the votes, but I've been voting a bit more actively since the counters were removed, and as someone whose voting habits are dictated by the score, it's been really interesting to go "blind" for a while. At the very least, I'd like to see this continue for a while longer, and perhaps finally wind up as a configuration option that each user can toggle based on his/her preference.

points by thekevan 4 hours ago replies      
I prefer points on comments. I commented about this elsewhere:


A couple reasons. If I am not totally familiar with whatever the original post is talking about, often the top rated couple of comments give me some good insight or jumping off points to look into it further. Again if this is something that is new to me, it gives me a hint that the poster(s) of these comments know what they are talking about as opposed to a comment rated at -3 which seems ordinary to me but maybe has a hidden agenda I have not seen. (An example of this could be a comment saying--and let's pretend I know nothing about domain registrars--"I recommend GoDaddy because they respect wildlife on a corporate level and have family friendly advertising." I know that is a crock but if it were about a Rails hosting company or a feature of node.js I would not.)

I respect the HN community and have learned a lot here. I generally trust their judgement and I have found if a comment is rated highly, it most likely adds a lot of value to the discussion. (Unlike on reddit where the top couple comments may be a clever joke or inside reference to the community--but that is okay because I read the two sites for different purposes.)

Sometimes I disagree with the highest rated comment(s) and sometimes those comments have a bunch of replies from other HNers stating their agreement. I then see my opinion is in a minority and maybe I re-examine it or stand firm and make a comment to the contrary.
Basically it is a nice guide in my perusal of this site. It is not absolute but I like to use it as a reference.

points by antirez 10 hours ago replies      
I was for not displaying points initially, but after a few days of using the site in this way I'm for displaying the points again. The strongest reasons are:

1) It is harder to scan for good content.
2) Even if I want to read everything, when my knowledge about something is too weak to evaluate the real value of a comment, I can no longer use the wisdom of the crowd to make an idea.
3) Sometimes to resonate with a comment like a question like "Please can you point me to the source code?" the best thing to do is to upvote, so that it is easy for people willing to reply to evaluate if the effort is forth it. There are many instances of this case.

So all in all it was better with points IMHO even if there are advantages without points.

points by jonmc12 15 hours ago replies      
Ideally, I'd like the voting history - both upvotes and downvotes in a little sparkline (or at least the %). That can tell me more about the comment than upvotes alone.

If there will be no points on comments, it would be very beneficial to have some kind of indicator on the comment that it has 2-3 std deviations of votes above the average comment on the thread.

points by greendestiny 16 hours ago replies      
I'm not sure. I definitely prefer to read with points on. I find it hard to skim a comment thread now, and I have no idea what the community thinks of particular points of view.

As for what it does for the community there hasn't really been time to judge. I've seen a lot less of the usual suspects coming in saying the same the things they always say because they know it'll be highly upvoted.

Could there be a keep it for now option?

points by gommm 15 hours ago replies      
I think it would be a good idea to keep it for another week or two so that we can all get used to it first.

The advantage I can see from not having points is that I spend more time actually reading the comments and not be influenced by the majority here.

The disadvantage is that I lose a bit of information when there's a comment with scientific content in which case I like to look at votes to see if it's correct.

I agree with some people to show the points in topics that are more than 2 days old...

points by kmfrk 16 hours ago replies      
I am glad that pg managed to remove the karma display while making it clear what comments are below 1 karma, so you don't downvote someone beyond what is reasonable.

I also get the impression that it's affected the discussions for the better.

points by logjam 32 minutes ago replies      
Why not allow the user to choose?
points by notJim 16 hours ago replies      
One thing I don't understand (I'm relatively new, I guess): are comments in a thread sorted according to upvotes?

If they are, my vote is for without points, because then I can still skim, but I won't have these numbers to focus on.

points by sliverstorm 14 hours ago replies      
I kind of like the feel of the site since, but I don't love how hard it is to wade through comments now. Have you considered collapsible comment threads, ala reddit or slashdot? I know this place doesn't want to turn into either, but as the comment volume goes up it seems like a better and better idea.
points by anigbrowl 10 hours ago replies      
I prefer HN with points displayed on comments.
1168 points

I prefer HN without points displayed on comments. 989 points

Ah, welcome my creatures of the night. You have brought victory after a close contest. Seriously, it is interesting that early evening USA saw an almost perfect 50-50 split (384-382 at 9pm PST), but as daytime moves west across Asia towards Europe, there's been a strong trend towards points.

Paul, this might be worth exploring in more depth, breaking out by age or timezone for example.

points by anateus 12 hours ago replies      
I think a poll isn't the best way to decide this issue. After all the problem you're trying to fix is the greater number of low-quality comment, i.e. those who provide low-quality comments outnumber those who provide high-quality comments. Thus, it is not unreasonable that the low-quality commenters would select the option that enables them the most.
points by rsbrown 4 hours ago replies      
Beware the results of polls such as this. Any time change is introduced, we know that some people will naturally react negatively. When asked what they prefer, people will often give an explicit response even though their "espoused theory" may differ greatly from their "theory in use": http://www.lopn.net/TheoryofAction.html

Not to say these polls aren't valuable, just take them with the requisite serving of salt.

EDIT: In the interest of full disclosure, I strongly support the new "no display" points.

points by dpcan 1 hour ago replies      
Show me comment points after I've reached a karma threshold. If I have 2000 karma points, then comment points show up for me.
points by larsberg 4 hours ago replies      
On the upside, I spend a lot less time looking at the HN comments. I used to scroll down, looking for insightful comments within otherwise low-ranked threads or in a sea of trite comments. Now, I can't distinguish them, so I don't even look past the top-ranked reply and the top comment or two.

Not having points displayed has probably given me ten minutes a day of my life back :-) Though I can't say that I prefer it that way.

points by 3pt14159 2 hours ago replies      
No one will probably read this, but please, please, please bring back the points score. I've been working on a hacker news crawler that rolls up facts about comments (user centric trends, etc) and without the points it will be very much limited.
points by random42 13 hours ago replies      
I like HN without comment scores, because it avoids the bandwagon effect, also username biases become more important that the actual content, BUT comment score do solve an important problem, to highlight the quality of a comment, in a threaded discussion.

Solution - May be instead of say "37 points", mark it in a range, (say "20-50" points), to give some idea of the comment quality.


Even better, Show the percentile range for the comment, within the thread. (Say "80-90"%ile comment, would be better than a "40-50"%ile comment.)


Grade them ("A" - "E") for the thread comments.

points by DanielStraight 6 hours ago replies      
Asking people what they want isn't necessarily the best way to give them what they want. Remember the Henry Ford line, "If I had asked people what they wanted, they would've said faster horses." PG has mentioned this before as well in response to requests for UI improvement on HN. I don't remember the exact words, but people were saying things like, "If there was any other site that had the quality of HN and better UI, I would go there," and PG's reply was that the reason the was no site that had the same quality and better UI is because he spends time worrying about quality instead of worrying about UI.
points by Evgeny 15 hours ago replies      
Often, when the discussion is related to the field I'm not knowledgeable in, the points are very useful to me.

Simplified example: if I know nothing about security, but I can see that the comment "Emailing passwords in plaintext is wrong" is heavily upvoted, and "Emailing passwords in plaintext is not an issue" is downvoted, it helps me learn things.

points by barista 16 hours ago replies      
Stop groupthink. Please don't show points on comments.
points by philfreo 16 hours ago replies      
"With", simply because I'm often too busy or lazy to read entire threads and just want to see the "best" comments.

I'd be okay with hiding the numbers but having some other indicator which could accomplish the same goal.

points by mrjbq7 16 hours ago replies      
I'd recommend adding an "Undecided: let the experiment continue" option.
points by sixtofour 1 hour ago replies      
When I'm skimming comments on a front page post, name and points help me zero in on things I'm more likely to find interesting or useful.
points by jarin 15 hours ago replies      
I think it would be useful to see the number of upvotes and the number of downvotes instead of the total score (or possibly a percentage instead of a hard number). That way, you can see whether a comment is at -2 points because it's bad, or if it's at -2 points because it's controversial.
points by allwein 3 hours ago replies      
The consensus points I'm reading are:

I miss knowing whether a comment will be good and worthwhile to read or not.

I miss the points because I use them to search/filter/etc.

I'm being more thoughtful in reading comments because I don't have the points anchoring or influencing my opinion.

I'm being more thoughtful in the comments I write.

On the whole, it's sounding like the change is bad for individual users, but good for the HN ecosystem in general.

points by spicyj 16 hours ago replies      
Ironically, I'm actually voting much more than before; perhaps because I'm reading more comments in depth, perhaps because I can't feel like a comment already has its "proper" score. Either way, I like the new way. (Perhaps show points after a few days or weeks, though.)
points by Goladus 4 hours ago replies      
I thought I would hate not seeing points on comments, but I find it's been very refreshing. It's much rarer that I find myself annoyed at the number of people who have upvoted something stupid or ignored something insightful. I still see the points on my own comments, which is the important part.

My vote: definitely continue hiding the points. It takes focus away from the content.

points by Locke1689 16 hours ago replies      
I think no comment points is fine, but in the case where there are no points I think you need to have a more accurate comment ranking system on the page. While having gravity to pull comments down is perfectly reasonable when points are displayed (because users can still look at the points and pick out old but valuable comments), without points the most important indication of comment quality is based on the ordering on comments.

Also, perhaps it would be useful to still display comment points in single-view? It would prevent eye-balling a comment, but would still allow you to view a comment score with a little extra effort.

I am aware this could provide preferential treatment to comments already on the top of the page but with the influx of new users the most important thing to prevent is a deluge of mediocre comments, not necessarily the dearth of exceptional comments.

points by 6ren 12 hours ago replies      
For the case where scores are used to measure agreement: a special kind of vote comment (like polls for submissions), to express agreement/disagreement.

A checkbox for whether you want your comment to be voted on, or evaluated. This is displayed, so HNers know how to treat it.
A "poll" comment doesn't count towards karma

points by nickconfer 2 hours ago replies      
I voted for with points, but I think there really should have been a third option.

The main reason I like points is it helps tell me what comments are worthy of me spending time to read when a story has 50+ comments.

So without points, I think you still need a way to show what is valuable and worth reading for the most popular stories. There are a couple ways to do this.

One option would be to personalize what comments are deemed important to me. For instance if I go and up vote pg comments, then when I go to another story and theres a pg comment it should give me a mark to let me know this comment will probably be of value to me. Another user might visit the same story and down-rate your comment because they don't like it. Thats fine, in the future, your comments probably won't be displayed prominently for them in this case.

This moves away from the group-think mentality of what is valuable to an individuals preference, which I think is more valuable and provides a less contentious exchange platform.

points by fname 7 hours ago replies      
I'm sure it's been brought up before, but what about only displaying comments either above or below a certain threshold. For example, hide the points on a comment once it's above 4.

Or show points after someone has upvoted or downvoted a comment?

points by ern 6 hours ago replies      
I think that visible point counts make it clearer what the norms of the community are to newcomers and reinforce them for other users.

Comments that fail to attract points (but are not downvoted) send a clear signal that something is not right. +1 would be an appropriate HN score for borderline snark, for example, that adds some insight, but doesn't aid the tone of the discussion. This sort of subtle signalling is lost when points are hidden.

points by tt 5 hours ago replies      
The problem HN is facing is a human-interaction problem of which hiding or showing points on comments won't solve. Sooner or later, problems with user experience will come about one way or the other.

I saw some other post about the Whirlpool Moderation forum (http://whirlpool.net.au/wiki/wp_modfaq) and thought the transparent moderation policy they have is pretty amazing. PG et al should at least research it and consider implementing something similar.

points by davi 16 hours ago replies      
I immediately liked the change. I realize now that in addition to evaluating the comments I was reading, I was also evaluating the readership's collective response to the comments. It makes for better reading to only evaluate one thing at a time.
points by vacri 14 hours ago replies      
The solution is simple: display capped votes.

- 'good' results get shown for the cap, and from experience on other such sites, upvotes slow down once it hits the cap anyway
- 'bad' results show as well, as someone already said, sometimes a comment needs to be downmodded to oblivion, other times it just needs that slight rebuke and you don't want to downmod more.

Caps of +/-10 are enough to point out 'this comment is worthy of more attention', and will help avoid that problem where a popular opinion draws a hundred upvotes where a mere insightful opinion draws only a score.

points by richardw 15 hours ago replies      
I think it's great. It makes people slow down and consider the discussion instead of being guided by those who have seen it earlier. That reduces the impact of the first-to-vote.

Essentially, it makes each vote more independent.

points by natch 15 hours ago replies      
If points count, then I want to see them.

But there's a bigger problem, as long as points matter for anything: Early comments, even fairly inane ones, have an overwhelming point advantage over late comments, even fairly good ones.

points by wslh 5 hours ago replies      
If I were you, In a few cases (10%) I would include random fake points to see what happen. Someone who has 2000 points showing as 20000 or 200 to see how the HN crowd react.
points by afhof 16 hours ago replies      
I think the bigger problem is that highly rated comments below the first comment are pushed out by replies to the first.
points by johnfn 12 hours ago replies      
I've been noticing that the poll tilts in favor of the highest comment on the page. 'Without' used to be slightly ahead when edw519 (for example) was on top; now that a post more in favor of the points is on top 'with points' is nearly 100 points ahead.

I guess that in some ways it's hard to be completely unbiased.

As for my own opinion, I want to echo that no points are fine as long as we see some sort of gradation of quality. Maybe just >10, >50, >100 is necessary. It doesn't have to be too complicated.

points by msg 11 hours ago replies      
It's been said a few times on the thread but I'm not willing to read the entire thing just to upvote rather than comment. Here is the problem:

It's been said a few times on the thread but I'm not willing to read the entire thing just to find the nuggets among the blather. Here is the problem:

It's been said a few times on the thread but I'm not an expert on everything and I'm not willing to halt my day and gain the background to fairly judge the veracity of every comment on the threads I'm interested in. In fact, I'm fairly likely to miss a counterintuitive point or chance to think deeper.

Was this comment too verbose? I don't know because I only skimmed it while I was reading this enormous thread.

points by dave1619 14 hours ago replies      
I think a bigger question is how do you create order when a thread gets to 100+ comments like this one. I personally get lost and stop reading comments on long threads. But there's a wealth of good conversation going. It just gets harder and harder to pinpoint them as comments increase.
points by mediaman 11 hours ago replies      
The majority will likely state they prefer to see points, because people will systematically underestimate the (negative) influence the data have on their decision making, much like people cannot accurately estimate their own ability to drive cars with above average skill.

Paul, in making the judgment, if even a third of voters think points shouldn't be displayed, it is a strong indicator that that's the right thing to do, because there is no countering cognitive bias that would cause people to state that they want less information.

points by johnrob 11 hours ago replies      
I feel like part of a firehose when commenting on this thread, but I'll quickly add my 2 cents:

How about showing points for top level comments (those whose parent is an actual posting). Any derivative comment has no visible score. This would help me parse the high level topics while still deflating most of the flame wars.

points by jasonlotito 7 hours ago replies      
I like without points, but would still prefer some way to know the quality of the comment. Maybe some visual indication (both color and a symbol maybe) would help. Then we would know a comment has been rated higher than normal. Set a cutoff small enough to be useful in smaller threads (like 5 or 10 points), and leave it at that.
points by godDLL 11 hours ago replies      
As current poll results indicate we're torn.

On one hand, skimming the comments to see which are worth it to read is now denied. On the other hand, discussions seem to have improved, and more focus is on a comment's content and less on community appreciation of it, which makes an individual voice stand out more and not be drowned.

I like HN without comments, I just wish there was a way to tell which comments the community appreciates more, at a glance.

points by stuhacking 6 hours ago replies      
Why not stop rendering HN pages in tables with inline formatting and render the page as a semantically structured document with a default style sheet. This will allow others to come along and restyle the site as they see fit. Don't like comment points? hidden. Post score is less than -4? hidden.

I personally don't understand the reasoning behind using tables as formatting and inline styles on a site whose content generally includes articles about web development.

points by OoTheNigerian 9 hours ago replies      
I blogeed about it http://oonwoye.com/2010/09/25/angelgate-dave-mcclure-and-re-...

summary is below.

    Vote count should show up only after the discussion is no longer ‘hot' (say after a day). So people are not unnecessarily influenced by the upvotes by others

No down voting: So people will not lose karma just for having an opposing view. Not gaining, is not as bad as losing

points by 3am 14 hours ago replies      
I'm not in a YC startup, and I don't seek to start one. So take it with a grain of salt... the tone here has been better in the short time without points.

I think some of the personality types that YC attracts do poorly in a civil discussions where there is a (public) competitive element.

points by fuzzythinker 15 hours ago replies      
I mostly like no points displayed, but with one exception - comments that provides information you are unsure of trusting or not, and there hasn't been replies to it yet. (eg. scythe's recommendation on domain registrar [1]). In order to have better information on the post's trustworthiness, I had to click on the user's link. I guess we'll need to wait to see how much other nuances for this no points display.

[1] http://news.ycombinator.com/item?id=2445061

points by Kisil 4 hours ago replies      
How about separating "quality" and "agreement" into separate scores?

While it complicates the interface, it would be much more useful for both authors and readers. It naturally leads to measuring controversy - the most interesting comments would be those with high quality and mixed agreement. Having a sense of HN's opinion would direct posts towards areas of maximum contribution. It would also re-direct some of the pile-on upvoting from low quality but obvious/funny/mean/etc. comments, so they could drop down the page despite having some appeal.

Some example cases:
a) I may agree with the snarky comment that calls out someone's obvious mistake, but I can simultaneously downvote for snark.
b) I may disagree with someone's analysis of the iOS vs Android market battle, but I can now register my disagreement while acknowledging that the poster made a good point.
c) I made a comment that came off as means-pirited, though that wasn't my intent. The quality downvotes are an unambiguous message, which I can't just write off with "I guess HN disagrees."

points by duck 16 hours ago replies      
Mini-poll: Has everyone noticed how polls don't get very many upvotes? Many of them tend to get a ton of actual poll "votes", but not upvotes. I have always felt like if you are going to vote for a item you should give the poll a vote as well.
points by thenduks 6 hours ago replies      
I use HN to hear about news and read interesting articles, and also to occasionally engage in conversation. Points on comments have absolutely no bearing on being able to do that, so I didn't even notice at first that they were gone.

It is handy to know the points of your own comments, which can help you adjust your behavior to fit in with community expectations -- but we haven't lost that ability, so I'm not even sure what 'cost' this poll is referring to.

points by grandalf 14 hours ago replies      
It's a nice improvement. There is nothing worse than writing a thoughtful comment and then seeing someone write a snarky reply that got more points than the thoughtful comment. It just causes bad energy.

I like how points are shown only for each user's own comments. This way it's a competition only with one's self.

points by marklabedz 4 hours ago replies      
While I appreciate the idea of logarithmic scales and other advanced formulas for accumulating points, how about a simple threshold for displaying the points on each comment? For instance, no points are displayed until a comment is +5 or -2?
points by danenania 15 hours ago replies      
I really miss the points. I know it's superficial, but it added a form of engagement that is gone now.

What I don't understand is anyone getting worked up about the points one way or the other. To me it's just a meaningless, harmless, and somewhat fun diversion. Yet so many seem to take them seriously and get bent out of shape if they don't agree with people's votes. What difference does it make? Oh no, my post is light gray now! Oh no, a less than brilliant comment has a high number next it! Really? Just take a deep breath. It doesn't matter.

Despite all the grumbling, HN is still one of the best places to find interesting and intelligent discussions on the web, and despite the occasional flukes, votes are generally a good indicator of which posts in a long thread are most worth reading when there's limited time. I say don't fix what isn't broken!

points by leon_ 9 hours ago replies      
I'd prefer no visible "karma" at all. Karma should define a comment's placing in the thread/a posts' placing on the main page - nothing more. No user karma.

Also no user nicknames - if we had a truly anonymous discussion culture there would be more interesting discussions and people wouldn't be afraid of posting controversial opinions.

points by jmatt 15 hours ago replies      
I generally support not displaying points on comments. But I did run into a case where it would be useful to have some sort of feedback on a series of comments (even if it's just dots or some other non numerical indicator).

I was reading a recent post on the open-sourcing of some of Greplin's Lucene and Bloom filters code. sigil made a comment asking about performance based on how they implemented prefix matching. Then nostrademons responded[1]. When points were displayed it was trivial to get a consensus of whether the community agreed on the performance assessment or whether it was dubious/questionable. Since points are not visible I can't rely on the consensus of the community to help verify or make a decision. Instead I just have the posts and order of the posts. Which, of course, isn't necessarily bad. But has taken away my ability to use the HN community's consensus as verification.

[1] http://news.ycombinator.com/item?id=2444585

points by sambe 5 hours ago replies      
HN doesn't want either comments-solely-for- or voting for agreement or disagreement. Yet that is what a lot of people want to express. Perhaps it is more useful to have agree & disagree buttons, and the comments interest value is the total votes in either direction. Offensive/troll comments could still be flagged to separate them from comments which are largely disagreed with.
points by Ruudjah 7 hours ago replies      
In 2000, having a guestbook on a website where you can post a message anonymously was the norm, spambots did not exist.

in 2005, the ability the reply using an account was the norm, where replies could be rated 1-5 (slashdot and the rest).

In 2010, upvoting messages by giving them points was the norm, which Digg started and Reddit perfected.

In 2015, labelling posts I think will be the norm. Users will get the ability to associate posts with labels, such as insightfull, funny, interesting, firstpost, nonsense, et cetera. Every post will get it's own tagcloud, the labels being associated a lot being bigger.

The problem with upovoting is that you only classify posts using a numeric value. A numeric value can only express a limited classification. It does not say anything more specific then "bad rated" (<1), "not rated" (1), "somewhat rated" (>1 <6), "good rated" (>5 <20), and "awesome" (>20). Other then that, it does not say anything useful.

points by nitrogen 16 hours ago replies      
I think some of my posts have received more upvotes without points than they would have with points displayed. I wonder if in general there has been more total upvoting without points, as people are simply expressing appreciation or agreement rather than trying to move posts toward a "fair" score (something I've done in the past).

It's difficult to say whether that's a good thing. Since the poll numbers look pretty close so far (though it's only been ~15 minutes as of this writing), maybe it should be an option.

points by teyc 16 hours ago replies      
I felt the way you have put the comment in various shades better. The algorithm already sorts out what comments should float to the top. If people feel an answer is incorrectly placed, they can vote on it to fix it.

Point scoring: yes, it is a problem. HN "feels" somewhat nicer.

Further, there may be different reasons a comment may have been voted up:

1. The comment voted up is considered relevant

2. The reader agreed with the sentiment.

3. The comment was irrelevant, but was humorous.

Similarly, a comment may have been voted down because:

1. The comment was irrelevant.

2. The comment was relevant, but the reader disagreed with the point of view.

3. A polarising comment may end up with a net score of 0, but it is actually very relevant.

4. The comment was relevant, but was delivered in a brusque manner.

points by CallMeV 11 hours ago replies      
Perhaps this is a good thing. I consider it like this.

I well understand the idea of downvoting to express a response towards ugly, trolling and irrelevant comments. By downvoting, responders show trolls that their attitudes are not welcome.

However newbies such as I are often too scared to ask our questions, or to comment on something of interest, or indeed to contribute in any way, for fear of someone more experienced coming along and downvoting us.

As a system of control, to weed out the unhelpful while promoting helpful discussion, karma points were initially a useful tool. High karma scores indicated a valued contributor to the site.

However once the mindset of points farming settles in, as I suspect it has done here, the contribution ceases to matter in the rush to acquire more points, leading to a drop in quality and this competitive points scoring obsession. Delusion replaces reason and desire to contribute.

Doing without the points scoring could go some way to redirecting the focus on the items and the discussions stemming therefrom, rather than on the karma farming.

And that is my opinion.

points by brg 15 hours ago replies      
I prefer without comments, but would like some mechanism by which comment scores affect layout. It is often now difficult to process a large discussion, as high signal and low signal replies are indistinguishable.

For instance, having predefined threshold for collapsing to a title (5, 10, 50) or deep shading for >= + 10 comments would be very helpful in processing a large discussion.

points by faramarz 16 hours ago replies      
I think it would be useful if the points were shown after a time delay. Throughout the day, I visit the same thread a dozen times and follow the discussions. It would be nice to gauge what the general consensus is by seeing the comment points upon the 3rd visit or so..

But to answer your poll, I voted for them to be hidden.

points by smokinn 16 hours ago replies      
What I used the scores for most was to know when to stop scrolling down when a story was highly commented. I would figure that if there are 100+ comments chances are any threads that start with an initial post < 10 are safe to skip.

Now though, I have very little info as to where I can "safely" stop. (Though I know I've gone too far if the font starts fading!)

points by benreesman 10 hours ago replies      
for years i've been very impressed with how well the original approach worked, and I tend to get downvoted a lot, being a bit of a troll now and again. I don't oppose a new approach, but I do think it is on the hook for compelling evidence of improvement. hacker news has aged well in the same way that the constitution has, considering Internet time.
points by bguthrie 15 hours ago replies      
For what it's worth, I think it's been a big improvement. It forces me to focus more on the content of individual posts, it's made me more careful with my own allocation of points, and--I know this this fuzzy and subjective--I think it's improved the quality of discourse. Which ultimately is the only metric that really matters here.

Maybe that's just a temporary thing; people change their behavior when they know they're being observed. But I'd urge you to keep it for a little while longer.

I like the idea of a more coarse-grained system, perhaps one that uses font weight or shades of grey to emphasize and de-emphasize certain areas of the conversation, weighted perhaps towards the dialog around a particular comment rather than the comment itself. Perhaps over time that would come to suffer from the same problems as the last system. But hiding the algorithm allows you to readjust it before people have the chance to reverse-engineer and game it. Numerical point systems are useful sometimes, but they tend to activate humans' incentive structures in some really negative ways.

points by dr_ 15 hours ago replies      
Oddly the point system went off when I hit a karma of 250, making me think I was being punished in some way. Great to see that isn't the case.

Honestly, there are sometimes WAY too many comments to wade through, and I feel that comments with higher points have been vetted by the community, so I tend to focus on them. This isn't the case when there are just 10-12 comments, but with over 70 comments, the point system can be helpful.

To remove the system is to imply that the community can no longer be trusted to handle the sometimes contentious exchanges on their own. I hope that hasn't become the case.

points by crystalis 11 hours ago replies      
Without points, I mostly just read posts from users whose comments used to show a lot of points. Without points, that list only gets smaller for me.
points by Groxx 15 hours ago replies      
I'm going to toot my own horn: https://chrome.google.com/extensions/detail/hahhhdmfdgfiehpg...

I prefer them hidden until I don't want them to be hidden. It makes sense to hide them prior to voting, but I do like to see them sometimes - either after I vote, or on the comment's "link" page. But getting rid of the first, default display is nice once you get used to it.

points by ecuzzillo 16 hours ago replies      
I'd be most interested to see whether hiding the scores changed voting behavior appreciably. Do stupid comments get upvoted more or less? It seems like that's the ultimate test, much more than whether people prefer it after trying it for a few days.
points by shawndumas 6 hours ago replies      
Having a bit of aspergers I am seldom able to interpret feedback.

HN has been such a shelter of safety for me because of the clear (in comparison) and immediate feedback.

points by swah 15 hours ago replies      
One thing about points its I can pinpoint the "best" or most controversial comment quickly - if you can do that other way, for me its fine.
points by gojomo 14 hours ago replies      
I'd prefer an experiment in 2-axis voting. The up/down karma points need not be displayed per comment, but the new right/left agree/disagree totals would be intended-to-be-displayed-per-comment. More details here:


points by user24 9 hours ago replies      
One thing I'll say is that if 'without points' is kept, please change the arrow UI. It's now impossible to tell if you accidentally downvoted someone.

I like it sans-points over-all though.

points by petercooper 15 hours ago replies      
I voted up both as a sort of "abstain". At first, I hated it. Now, I think I might like it. It's stopped me from voting comments up and down so much but I think, perhaps, it discourages the wrong sort of people from voting but the right sort of people keep on doing it anyway. If that's true, it could be a big win.
points by jamesbkel 13 hours ago replies      
I quickly scanned for another reference to this, but didn't see anything - pardon if a duplicate.

It seems that you can vote for both options. After I voted, the other choice still has the 'upvote arrow'. I didn't try to see if it worked since I don't want to vote for the other option.

Is this intended? I never noticed on any other polls, but maybe I just wasn't paying attention.

points by razerbeans 3 hours ago replies      
I personally find that seeing the points by comments allows me to see interesting comments while I am skimming over a thread. I wouldn't mind having points hidden if there was some way that the most commented on threads or most popular threads appeared closer to the top of the story.
points by yurifury 16 hours ago replies      
I'd like some sort of indication of which posts are exceptionally high scoring. This would let people skim the comments if they aren't going to read them in-depth anyway, by signalling that maybe this one comment thread is actually a worthwhile read.
points by jamesbkel 13 hours ago replies      
Hopefully not adding to the noise, but wanted to specify my opinion beyond the simple Y/N of the poll. I like not knowing the precise point value, but I do agree - especially for large threads - that it is nice to have a way to quickly scan. I think that some sort of color coded could work, as several of the other posts suggest.
points by skm 13 hours ago replies      
I vote to remedy imbalances:

  valuable && low points => upvote
valuable && high points => read, no need to upvote

Does anyone else do this?

points by w1ntermute 16 hours ago replies      
More importantly than hiding points, I think names should be visible on mouseover only. There's truly no need to see the username before reading the contents; rather, you should look to see who wrote a particularly good comment after having read it.
points by hboon 15 hours ago replies      
Give it another week or two. Then poll again. It's too soon to conclude, both for people who may take time to get used to it, as well as people who don't visit HN that often (but still regularly).
points by latch 15 hours ago replies      
I want points gone, but I want something more generic to take their place to filter out useless comments. The ordering doesn't work because 1 - it often doesn't differentiate between a 50 point comment and a 2 point comment, and it doesn't handle useless comments which is a child of a popular comment.
points by markokocic 7 hours ago replies      
I would prefer if HN would let me configure if I want to see points or not.
points by Tiomaidh 16 hours ago replies      
My gut reaction is to keep them, but I think that's just because I don't like change.

If you do decide to drop them, it might be nice if something different was done for the time-displaying. I keep thinking the minute count in the comment header (pg 22 minutes ago | link) refers to points (pg 22 points [random text] | link). I still feel like jerf's comment (9 minutes ago) has more points than skennedy's (4 minutes ago), even though I know it's not actually true.

Although I'd probably get used to it within a month--it's just an annoying thing during the transition.

points by burke 15 hours ago replies      
Without. I don't want to have to read the whole comment thread to get the HN Zeitgeist.

What if it was an option? Or better yet, something like "Hold Ctrl to see comment scores", or "hover here to see the score for this comment", etc.?

Seeing comment scores is massively valuable to the way I consume HN -- skimming the article, then reading the top-rated 2-5 comments.

points by gbelote 16 hours ago replies      
I like comments without, I find myself focusing more on what's being said than skipping over folks with significantly less comments.

I've read criticisms about not being able to quickly/easily filter out top comments (which I care about, too). But there might be other (better?) ways of achieving that without reverting back to scored comments. For example, coloring/annotating comments that cross a certain threshold (either an absolute number, or relative to the other comments in the post). The advantage of color is that it becomes easier to visually filter out highly voted comments.

points by Vivtek 16 hours ago replies      
I'm starting to think I like it better without comments. Although if without, then I'd prefer those dots that showed up for a couple of hours on frequently upvoted comments. I really do like a skim option.
points by rooshdi 15 hours ago replies      
Hmmm, both options do seem to have their pros and cons in the current layout of the comment system. You may want to keep displaying comments without points but also try splitting the comment section into two parallel parts: A "Top Comments" column on the left side and a "New Comments" column on the right. This will highlight newer comments at the top longer and may encourage more readers to comment and reply to those while still showcasing the top comments of a story on the opposing left side for those who wish to skim real quick.
points by entangld 14 hours ago replies      
It seems people who don't want points displayed are trying to evoke a feeling.

I want points for one reason and one reason only. So I can tell which piece of advice people agree with.

points by tonystubblebine 15 hours ago replies      
I notice that my eye gravitates to longer comments now. That's probably a good thing.
points by samstokes 15 hours ago replies      
For the same reasons (presumably) as leaving the scores off comments, might it have made for a more representative poll to hide the scores on the poll options?
points by BasilAwad 15 hours ago replies      
How about showing comment points after a set period of time? I can see the value of not showing points on current submissions, but I was frustrated when I was trying to skim older topics.
points by dkeskar 15 hours ago replies      
Unless I am mistaken, there are two changes: a) no points display and b) no downvote arrow for many of us. Both are good.

There used to be many responses triggered by consternation over downvotes. This way, most people can only upvote iff they resonate with the comment, without the possibility of jumping on a bandwagon (or trying to derail one)

Suggestion: Moving the comment box to the bottom would nudge people to scan the thread before responding, duplicating, etc.

points by Jebdm 16 hours ago replies      
A better way to do this experiment would be to have half of threads show comment scores, and half not.
points by clark-kent 11 hours ago replies      
Show HN and Ask HN are two good reasons why we need comment points. Comment points provides raw data to know how many people agree/disagree, like/hate a technology or topic.

In a post with 100+ comments, the points makes it easy to parse the various threads and conversations going on.

points by tedkimble 14 hours ago replies      
My biggest complaint of the recent changes is that it appears to have broken the hckr news Chrome extension, which added new comment highlighting. For me this is one of those features that I never knew I wanted, but now that it is gone I can't remember how I browsed HN without it.

As for the point displays, I want them to be there, but I think HN will be better off without them displayed.

points by palehose 14 hours ago replies      
Upvoting a comment should signify that I agree with what the author is writing, not that I want that comment to have more karma points. I don't care how many upvotes a comment gets if the people who are upvoting it are not people I agree with.

It would be more helpful if there was a lingering effect to agreeing with someone instead of just giving someone a karma point increment. Lets say that I upvoted zzyzx's comment, and based on that upvote, I am able to see a thumbs up on other comments that zzyzx upvoted, indicating that people who I agree with, agreed with that comment.

points by huhtenberg 15 hours ago replies      
Perhaps make this a configurable option and keep them hidden by default?
points by brendano 15 hours ago replies      
How about, display a discretized form of points? -- e.g. high, medium, and low, with unknown thresholds (that may be changed through time or whatever).

Can still use them as a heuristic for filtering, but the details and possibly divisive aspects of worrying about minute differences are hidden.

points by eam 13 hours ago replies      
How about if points are hidden for the first 2 hours or so after a story is submitted? This way initially there wont be any bias on the comments, and eventually we can see the comments that are the most meaningful.
points by tuhin 9 hours ago replies      

-Display points after one has upvoted/downvoted a comment

- Have a quota of votes in any thread so a user has say 5 votes to spend in a new post (may increase with karma)

points by goblin89 14 hours ago replies      
Seems like it could be made either
a) easier to read / skim through comments ---with points shown, or
b) easier to cast a fair vote, and thus contribute to HN ---with points hidden.

Thinking of it this way, maybe we should have two different modes--- ‘quick reading', with points shown and voting disabled, and ‘contribution', with points hidden and voting enabled.

Wait, I know it complicates things a bit, but it makes sense. Look:

- If you do read everything carefully and don't rely on crowd's opinion, then you don't need points anyway and you can well contribute to HN by voting justly. Use ‘contribution' mode.

- If you usually read HN in a hurry, thus you shouldn't be tempted to vote. Use ‘quick reading' mode.

The weakest point here is changing mode. Don't know how frequently people switch between ‘skimming' and ‘contribution' mode while reading HN… If not often, then maybe it makes sense to place a switch on the user's settings page. (Maybe with contribution mode disabled for noobs.)

points by thurn 6 hours ago replies      
Sort of unrelated feature request (is there a better place feor this?): a small piece of JavaScript to warn before you submit a title that starts with a number between 3 and 15. There are about five articles on the front page right now that break the "remove useless numbers" rule.
points by wowamit 8 hours ago replies      
The points would make sense if I could sort by upvotes received for a comment. As long as that is not the case, I prefer it with comments displayed without points.
points by volida 14 hours ago replies      
As long as there are points, and comments are sorted by popularity, not showing the points, I don't see how it can mitigate the problem of people trying to game the system.

If it is somehow measurable who is trying to do so, in obvious way, in some automated way and if this activity is at a level that affects the rest, then the user could be somehow penalized off that karma gained. Nevertheless this is democracy, so I would only expect this behaviour being penalized only if it really affects other in some measurable way, so that the community could approve penalizing users who are trying to advantage of lack of automated moderation.

The contrary could be done i.e. instead of penalizing users who are suspectable in such behaviour, preferably award recursively to each comment in the same thread more points, therefore the commenters who are 'legit' are awarded. Therefore, add a category of comment points named 'HN ponts' which are awarded automatically by the system.

points by VB6_Foreverr 8 hours ago replies      
What would be an interesting experiment would be no points and no name displayed. Let the comment rise or fall purely on its own merits rather than the rep of its author. It seems to me that some people need only to cough to get karma
points by gcb 9 hours ago replies      
after removing the points on the comments, my comments bashing HN or Apple stopped being a consistent -1 to 3~7 (i post when the thread is already late)
points by mdg 15 hours ago replies      
There are quite a few people here who are suggesting pg make it an option. While that would make everyone happy, it wouldn't solve the problem; people who vote based off points will continue to do so.
points by nhangen 15 hours ago replies      
I hate not seeing it, which is why I think it's important to keep hiding them for a while. I find that I actually judge a comment by how I feel about it, rather than what the community feels about it, which is nice.
points by adrianwaj 9 hours ago replies      
Well, if you could at least do a "display:none" on the points, at least hackerbrain would show the points. I'm sure those points are used by derivative sites during parsing. Can you please put them in the html?
points by maddalab 15 hours ago replies      
I skim through the points, both on stories and discussions to decide which ones to read. As the size of the community has grown, I believe, probably mistakenly, that the points are more closely related to domain names on submissions and user names on comments, rather than the quality of the comment.

I tend to read some comments based on user names, irrespective of the points, for a few users, raganwald being one, primarily because I have read pretty much everything he has written and find it interesting more often than not. Even in cases where his comments has garnered few points, I have found the comment worth while. So I can understand the tendency to up vote comments based on user names.

Having said that, I vote for display of points and the hiding of user names. I would like comments to stand up for themselves rather than for the user posting the comments.

points by kschua 11 hours ago replies      
I use the points on the comments to sift make a quick decision whether I should read a particular comment.

As a compromise between your goal of people scoring points, maybe a scale might be good?

For example,

0 - 100

100 - 300


Edited: for formatting

points by niels 10 hours ago replies      
I prefer HN with points initially hidden, but displayed after a while. This way ratings are not as susceptible to group think.
points by instcode 14 hours ago replies      
I think a user content generated community where huge number of people participate needs an explicit feedback system. We all want to learn from the good sources and filter out what aren't worthy and we need a way to distinguish them. If you think the point rating system isn't good, it doesn't necessarily mean we don't need it.

Here is my suggestion: How about we use the absolute points to rank a comment in a different scale, such as if a comment collects enough a number of point, it will be ranked as 1 star, then 2 stars... and so on. In this way of rating, we not only get what we want but also not to follow "group thinking" as someone said in this thread.

points by blantonl 16 hours ago replies      
What is the algorithm that determines where a comment ranks in the hierarchy from top to bottom? It appears that there is more to it than just points.
points by joubert 15 hours ago replies      
I find this interesting given the discussion yesterday about transparency about salaries.
points by breathesalt 15 hours ago replies      
Even if the difference only amounts in cosmetics, I still would prefer HN without points displayed on comments. Hopefully though, it will place an emphasis on the message--you can always intuit its popularity by its position.
points by koenigdavidmj 15 hours ago replies      
Maybe hide them on a story view and show them when you click the permalink to that post?
points by mshron 15 hours ago replies      
What about an experiment? PG could take a few dozen people he wants to be exemplars for the community, pick some semi-popular articles before and after the change, and ask the test subjects to rate which article had better comments. For best results, only pair people with articles they haven't read.

Myself, I didn't even realize how strong the group-think effect was until the comments were turned off. It feels like there's more substance now, but I might just be reading more closely in the absence of loud communal signals.

points by Devilboy 15 hours ago replies      
50/50 so far. I guess it's up to you Paul!
points by mcantor 14 hours ago replies      
It would have been funny if no one could see the point total for the options on this poll itself.
points by skm 13 hours ago replies      
I love seeing what other people in the community find interesting. Points let me see not just the comment, but how others felt about it.
points by eiji 15 hours ago replies      
Only display points when logged in. What about that?
points by mdg 15 hours ago replies      
Why not let the person who submits the article decide whether or not comments are visible for that thread?
points by chicagobob 15 hours ago replies      
Hey! If you're going to turn off points in comments, please make it a preference. I enjoy seeing them.
points by sushumna 11 hours ago replies      
For some who wants to quickly go through the comments, it would be good to go through those comments which has more points. They are the most valuable comments and worth reading.
points by kbd 15 hours ago replies      
Wow, vote even at 500 to 500.
points by KevBurnsJr 14 hours ago replies      
I just voted for the poll option with the fewest points. I like rooting for the underdog.
How Much Does a Software Developer in Portland Earn? dyepot-teapot.com
6 points by turoczy 35 minutes ago   1 comment top
points by siculars 0 minutes ago replies      
I would say the same as you would get anywhere else +/-10%.

There have been a lot of threads recently around salary. I'm pretty sure it is entirely based on your ability to negotiate to the best of your abilities. Thankfully we work in a very mobile and fluid industry. I know people who live in Oklahoma and work for well known startups that get paid a very competitive wage even for NYC. Location is becoming less of a factor in the war for talent.

Concepts The Emacs Newbie Should Master benjisimon.blogspot.com
102 points by pdelgallego 9 hours ago   38 comments top 11
points by ekidd 7 hours ago replies      
If you're using Emacs and git, you might want to look into magit, which allows you to touch-type through common git operations. If you commit a lot, this will speed you up significantly.


I use custom Elisp to combine magit, git-blame, and other commands into a unified git environment:


It's especially handy to be able to pull up git-blame information for the current buffer with just a few keystrokes. This turns your git commit messages into a layer of hidden code comments.

points by silentbicycle 3 hours ago replies      
The post has many good suggestions! To them, I would add:

The mark ring. Whenever you place a mark (highlighting a region), its place is also added to a ring, and you can jump between them by pressing "C-u space" (or "C-x space" for pop-global-mark, which will jump between buffers). Many commands that move a long way add an implicit mark, so this works quite well as a "where was I..." command.

For "dynamic abbreviations", I recommend pabbrev (http://homepages.cs.ncl.ac.uk/phillip.lord/download/emacs/pa...) - when Emacs is otherwise idle, it starts building a lookup trie of prefixes to frequently used words, so it will suggest completions under the cursor as you type. I love it.

I much prefer iswitchb to the default buffer switching system. YMMV.

Don't forget about learning Emacs lisp, either! It adds quite a bit of flexibility to Emacs. It's kind of a crappy, backward Lisp dialect, but still tremendously useful.

Finally, learn to use the help system. Emacs's naming is really idiosyncratic, but it gives you several powerful ways to look things up, so you don't have to memorize everything.

points by davnola 6 hours ago replies      
2 concepts I think are missing from this list:

* Tramp for remote/sudo editing: http://www.emacswiki.org/cgi-bin/wiki/TrampMode

* emacsclient if you want emacs everywhere all the time: http://www.emacswiki.org/emacs/EmacsClient

points by program 5 hours ago replies      
Missing examples:

C-x z, the command that repeat the last command for every 'z' you type. This is in my opinion one of the most important feature of Emacs

M-s o, the 'occur' command that show all the lines that match a specified regexp

File variables: http://www.gnu.org/software/emacs/manual/html_node/emacs/Spe...

points by gcv 4 hours ago replies      
This thread is hilarious.

The Universal Principles of Design (among other places) talks about the 80/20 rule " but Emacs is so powerful and customizable that no two people ever agree on which 20% is the most important!

points by reedlaw 2 hours ago replies      
As far as buffer switching, I never keep hundreds of files open because when I switch git branches it's hard to remember which buffers need to be reverted. Instead I usually close everything, change branches, and then open up files as needed. I find I can open files more quickly using IDO-mode than I could by switching buffers (plus I often forget the filenames).
points by neutronicus 5 hours ago replies      
Is there a way to kill something without putting it in the kill ring? It's probably my biggest pet peeve about emacs that there's no way to kill text without emacs assuming that you want it somewhere else.
points by ndaugherty18 2 hours ago replies      
For the ruby people out there you guys should try xiki (xiki.org). It extends emacs with a bunch of cool commands via el4r.
points by calpaterson 7 hours ago replies      
Didn't know about dynamic-abrevs, and haven't learned how the kill-ring works yet. Thanks
points by limmeau 8 hours ago replies      
I trade registers and the help system (especially hyper-apropos) for narrowing and spellchecking.
points by JohnFritzen 7 hours ago replies      
The Geek's Guide to Finding the Best Startups bostinnovation.com
14 points by kmccarth 2 hours ago   1 comment top
points by gerner 46 minutes ago replies      
tl;dr build your personal brand via social media, blogging, Q&A sites, and open-source. Communicate what problems and technologies you're interested. Get other people (recruiters, blog readers, followers) to look for new opportunities you'll be interested in and they'll reach out to you.
Cash Flows in Silicon Valley, but This Isn't 1999 wsj.com
3 points by pelle 8 minutes ago   discuss
Compiling to Java as a target language might.net
34 points by fogus 5 hours ago   8 comments top 2
points by larsberg 1 hour ago replies      
I disagree with one statement:

> performance will go up, but implementation code size and
> complexity will go up by about a factor of two

Moving from basic compiler optimizations (CSE, let-floating, unused variable elimination) to static-analysis driven optimizations (arity raising, useless variable elimination, super-beta, CFA-driven closure conversion, transducer identification and optimization for concurrency) may only double code size, but the implementation complexity is pretty intense. Those optimizations often are performing global reasoning against a CPS-style intermediate representation, and debugging issues with them can take a while. We do all of the parenthesized operations listed earlier (in Manticore, http://manticore.cs.uchicago.edu), and I would claim order of magnitude more time/thought spent on the them.

That said, I think I'm supporting Matt's argument of avoiding moving up the complexity order unless you have to!

points by markokocic 5 hours ago replies      
Why would anyone compile to Java when it is much easier to compile directly to JVM bytecode?

A lot of languages are doing exactly that, like Mirah, Clojure, ABCL, Scala and many more.

       cached 14 April 2011 18:02:01 GMT