Hacker News new | past | comments | ask | show | jobs | submit login
Stack Overflow reducing headcount by 20% (twitter.com/andrewbrobston)
171 points by cjCamel on Nov 2, 2017 | hide | past | favorite | 199 comments

It's undeniable that their main developer Q&A site is a really great thing that saved me many hours of work. However, it was somewhat hostile to its more export users from the very beginning: the reputation system favors quickfire replies to grab the first upvotes, and they seem to put a strong emphasis on "cookbook" answers where many of the more nuanced discussions were closed as "opinion based". I think there are some missed opportunities here to make the Q&A site even more useful.

However, perhaps that wasn't their focus. My impression is the Q&A site was supposed to be mostly a gateway to their other services, and for that to work, the Q&A part simply had to be "good enough". It's an interesting strategy that perhaps didn't work out quite the way they had hoped it would.

Their Q&A site was vastly better than what existed at the time, and that's how they won the market. Now, though, it is full of people looking for quick points (as you suggest) -- people that think that the original poster was asking for someone to search Google for 'em -- and I'm convinced that the truly helpful people have moved just given up.

One thing I've noticed, that drives me crazy, is people demanding more information in response to a question, without any intention of using that information to help find an answer. Their whole goal is to score points for asking more details. I don't mean to suggest that more details aren't useful, only that it isn't helpful in any way to ask for more information if you don't intend to follow up. (Perhaps that could be improved by allowing those sorts of requests to be sent privately, and without any chance of generating points?)

I suspect there's a mismatch between what you expect Stack Overflow to be and what the other users expect it to be. From the tour ( https://stackoverflow.com/tour ):

> With your help, we're working together to build a library of detailed answers to every question about programming.

The "build a library" being the key part. Vague questions aren't that useful in building the library in particular because lacking the necessary information, they can have too many unhelpful answers. Thus the focus on the problem with example demonstrating the problem (so the next person with the problem can verify that it is their problem or not) that gets answers.

Furthermore, you appear to be under a misunderstanding of how the stack overflow reputation system works.

> Their whole goal is to score points for asking more details.

No one gets points for comments. At all. The voting on comments is purely for visibility ranking of helpful and constructive aspects to the post they are on... and nothing else. No points.

The reason that people ask those clarifying questions is to make sure that they are answering the right question and aren't wasting their time on trying to answer something that turns out to not be the problem. Say someone asks how to do XYZ in C++... and the person gives an answer for C++17... and then the OP comes back and says "sorry, I'm using turbo C++ and that doesn't work." Instead a comment asking "what version of C++ are you using" up front to make sure that the person asking the question gets the correct answer rather than something that is unhelpful.

There is no private messaging on Stack Overflow, probably for the same reason there's none on HN. It increases the moderation workload and possibility of bullying in private. I'd hate to be Jon Skeet if there was private massing. Furthermore, having it be a comment presents the information in public (including the clarification) so that if someone else has the problem they can read the comments and be aware that something may not work in their situation even if parts of the questions match. ("You didn't include a version of C++ in this and it appears to be a very old version, could you update the question with the version you are using?")

> Vague questions aren't that useful in building the library in particular because lacking the necessary information, they can have too many unhelpful answers.

If that is truly their goal (and I don't doubt that it's at least a hope) then they should go ahead and remove all the vague questions from their database, or at least flag them so they don't show up in web searches. After all, who's using Google to find unsolved, vague problems?

> No one gets points for comments. At all. The voting on comments is purely for visibility ranking of helpful and constructive aspects to the post they are on... and nothing else. No points.

I see. I didn't realize that. They do get the number but I guess it doesn't "persist" throughout the site. Good to know. Though, then I guess I don't understand what their motivation is to respond with useless comments.

> Instead a comment asking "what version of C++ are you using" up front to make sure that the person asking the question gets the correct answer rather than something that is unhelpful.

That's OK, I guess, but it seems to be counter to the idea of building a library. Rather than have a set of answers that may address the question for various versions of C++, you have a question and zero answers (because the person asking for details disappeared). That's the reality, anyway.

In any case, it wouldn't even have to be a private message. Simply having a "needs more detail" checkbox that hides the question (from Google et al) until the detail is added would be enough to improve the SO experience.

There have been efforts to remove vague questions. There is a lot of them. Stack Overflow corporate tends to leave that to the community moderation... which is a bit under powered. Until recently, the elected moderator team tended to be on the preservationist side. As to removing from search? https://meta.stackoverflow.com/questions/316881/ - remember that Stack Overflow monetization is driven by views. Deleting those vague questions that people search for (they do) that happen to have a lot of views and are locked ends up as a - on the balance sheet.

Reputation changes happen from: * Upvote on a question you asked (+5) * Upvote on an answer your provided (+10) * Accepted answer you provided (+15) * Accepted edit (+2) * Down vote on a question you asked (-2) * Down vote on an answer you provided (-2) * Down vote you gave on an answer (-1); yes down voting other peoples answers costs you points. * Receiving a bounty (+varies) * Giving a bounty (-varies)

Thats it. Nothing about comments. Nothing about closing questions.

The comments aren't necessary useless. They are often trying to help the person asking the question write a better question that can get a better answer. If the comment truly is useless, it should be flagged to be removed (and that won't give or cost anyone any points).

Guessing at the answer isn't that helpful. You've got no idea if it solved the problem or not. The next person to find it with search won't know if it solved the problem or not either. So instead of one question with an answer that did, you've got a question that has a dozen guesses... and you've got to go through and figure out which one works for you in your environment. Might as well have searched a forum and paged through all the guesses there.

Questions that don't have any answers eventually get removed by the system. https://meta.stackexchange.com/a/92006 describes the criteria.

> If the question is more than 30 days old, and ... > has −1 or lower score > has no answers > is not locked

> If the question is more than 365 days old, and ... > has a score of 0 or less, or a score of 1 or less in case of deleted owner > has no answers > is not locked > has view count <= the age of the question in days times 1.5 > has 1 or 0 comments

The "needs more detail" is often part of the close reason. The question is closed not as a "you did bad" but rather as a "don't try to guess at an answer until this is cleared up"

> Questions seeking debugging help ("why isn't this code working?") must include the desired behavior, a specific problem or error, and the shortest code necessary to reproduce it in the question itself. Questions without a clear problem statement are not useful to other readers. See: How to create a Minimal, Complete, and Verifiable example.

> Please clarify your specific problem or add additional details to highlight exactly what you need. As it's currently written, it’s hard to tell exactly what you're asking. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.

The hides from google, again, is the "this is how SO makes its advertising impression money." The reason given often is the "breaks the internet." If people linked to the closed question from outside that link becomes a 404. I'm skeptical on that being the only reason as I noted the advertising impression dollars there. I personally believe that maintaining poor quality material, no matter how many people link to it, is damaging to the brand and sets a poor example for what people asking questions expect. And thus, when they do catch the eye of someone who has an Atwoodian (quality above all else, delete the stuff that doesn't contribute) philosophy of quality on the site, it becomes a poor user experience for everyone involved. And as Jeff isn't there and Joel is CEO, the emphasis is different.

I say Atwoodian there - thats following the call of:

> It is by programmers, for programmers, with the ultimate intent of collectively increasing the sum total of good programming knowledge in the world. No matter what programming language you use, or what operating system you call home. Better programming is our goal.

The Spolskyians are following the call of:

> What kind of questions are appropriate? Well, thanks to the tagging system, we can be rather broad with that. As long as questions are appropriately tagged, I think it’s okay to be off topic as long as what you’re asking about is of interest to people who make software. But it does have to be a question. Stack Overflow isn’t a good place for imponderables, or public service announcements, or vague complaints, or storytelling.

Its shifted a bit from those original calls to people to contribute... but you can see a profound difference in the emphasis between those different calls for contributors. Then give https://stackoverflow.blog/2011/02/22/are-some-questions-too... a read and look at the editor history of https://stackoverflow.com/posts/1003841/revisions

The thing I'm trying to say there is the "why don't these questions get hidden from google" gets an "its complicated" and goes to the top of the company and its founder about what is appropriate for the site and not. Its not something easily done or decided.

> Questions without a clear problem statement are not useful to other readers.

I guess this is fundamentally where I disagree with SO's staff and/or moderation team. I occasionally find those questions and their answers very useful, when there are answers, and again more useful than the perhaps ideal, specific, and ignored questions that fill the site. When questions get too specific, the only person that might benefit from the reply is the person that posted the question in the first place. At that point it might as well not be part of a public forum.

My favorite SO pages are those that generate multiple replies with different perspectives, and often different languages or at least libraries. Those questions are admittedly usually vague but I think that given the result (many interesting answers from different people) that is a net positive. You might say that the answers given look like guesses, and that'd be completely fair, but there's still value. Certainly more value than an unanswered question abandoned by both the asker and the one person who expressed a small amount of interest.

I do see the difference you're describing (between Atwoodian and Spolskyians) and I can see the merits in both sides. I'm not satisfied with either, but that's more on me than on SO. In any case, I doubt I'm making any unique or new arguments here. I'm mostly just frustrated with SO and Google in general because I almost never find a solution to my problems.

As a sort of aside:

> The key distinction to make here, in my mind, is that all questions are ultimately in service of the people answering them. That is the audience you need to satisfy if you want to have any hope of creating and sustaining a community of peers learning from each other.

That seems to be how you make a site by and for narcissists, or people otherwise interested in hearing themselves talk. It'd be a different story if the answers were paid, I guess, because at least then the motivation would be something more tangible.

To be honest, I probably would have written SO off entirely if I'd seen this comment earlier.

Other sites have optimized for other things. Slant.co does a great job (in my opinion) of the "pros and cons" for different things. Quora has a more anecdotal spin to its thing. You can ask a question, get lots of viewpoints, but may not get any real answers for the underlying problem. Trying to make one site that does everything... I'm not sure it exists.

That last comment is more elaborately written in https://stackoverflow.blog/2011/06/13/optimizing-for-pearls-... . Read it all, but it concludes with:

> We feel that the world is awash in questions, but not answers. Answers are the real unit of work in any Q&A; system. Therefore, the only logical thing to do is to maximize the happiness and enjoyment of answerers. If this means aggressively downvoting or closing unworthy and uninteresting questions, so be it. Without a community of people willing to answer questions, it really doesn’t matter if there are questions at all, does it?

This was written in the early days of Stack Overflow. I believe the philosophy there was "its easy to get questions." I'm not sure I agree with that... or at least, it is certainly not easy to get well asked questions and separating the well asked form poorly asked questions is an immense undertaking. I don't think the system that SO has has scaled well. That you have trouble finding answers that may exist show that there is too much noise in the system and unless you know exactly what you are looking for it becomes difficult. Try removing too much of the noise and people start complaining about losing internet points.

Aside from the Atwoodians and Spolskyians... there are the Zuckerbergites.

> No, Zuckerberg didn't have anything to do with Stack Overflow. The philsophy of :+1: however, is one that can be seen on Stack Overflow. The web is a social place, and Stack Overflow is too. People you interact with, you up vote. It makes them feel good and you feel good knowing they feel good. No one likes to get a dislike or unfriend on Facebook - and no one likes to get down voted or have their questions closed on Stack Overflow.

And the LinkedIn

> The world is a hard enough place to get a good job. Especially when there are scores of other people trying to that single entry level coding position that opened up and you just got out of college. So, how do you set yourself apart from all the others? You put Stack Overflow on your resume. You provided 200 answers on Stack Overflow! Opps, that one just got deleted.

And the freELance

> Did you know there is a site out there where you can pay money for people to do some work? I haven't gotten job as a programmer yet, but I took a bid on eLance for writing a facebook clone for $200 for some beer money. I know some JavaScript and php... At worst, I don't get it done but have some great material for when I get out of college (or maybe I'll drop out and become self taught on the weekends while answering phones)

All of these are competing for how the community on the site is run. Without direction you get rules and squabbles. The results of this show up in reviews (first post - everyone gets an up vote no matter how poorly written) and close / delete "wars" where people try to remove something seen as cruft and other people try undoing that... and it ends up undeleted and locked (and contributing to the noise that makes it harder for you to get an answer with a google search).

The site that you get when you don't try to have the great answers that arise from good questions, well... https://answers.yahoo.com/dir/index?sid=396545663&link=list

As to the narcissists? Its more of a "the people answering are trying to answer the question once." Its not to hear themselves, but so that programming problems are resolved by searching rather than by asking. Similar to the driving goal of Wikipedia - all knowledge easily available.

If answers were paid... that goes down an interesting path for philosophy and motivation. A difficult path for a web site (money is hard to handle) and motivation (you're only putting $0.05 on this question, we'll, I'll give you $0.05 of my time. "This can't be processed by a regular expression." - poor answers to poorly paid questions)

>many of the more nuanced discussions were closed as "opinion based"

This is also something I don't like, but I wonder if they went on, if the website would become another place for flamewars - something that happens a LOT when you have low moderation.

The thing that separates SO from forums is the focus on quality answers. There are more than enough locked examples of the opinion bickering if one wants to see what happens.

They may be interesting to read. They may be popular. But most of the time by page 2 and even worse at page 5, it’s crap. And that crap requires an excessive amount of community moderation work to keep clean and try to avoid having more people add answers to the bottom of page 7 that are already stated 6 times on previous pages.

If the moderation tools are less powerful than the popularity keeping it there, there isn’t much reason to moderate.

And thus the culture by those who shovel crap daily to try to shut down opinion based questions early - so they don’t have to do more work later.

Every attempt I’ve seen at an opinion oriented SE alternative has failed to get sufficient expert answering. It’s not fun as a ${Lang} expert to have a newbie argue with you regularly about some feature or design choice.

I couldn't agree less. First off, they're not going out of business, just letting 20% go and one would assume refocusing. I think the site is set up to get good answers and not just the first upvotes as answers stick around and users can vote answers up or down as they see fit. The first answers on older questions tend to be very long and in-depth and give a lot of good information about the pros and cons of the approach they discuss. (If you want something really in-depth, check out https://codereview.stackexchange.com/.) The ones that get closed as opinion based are almost entirely asking for software recommendations or the like. I've rarely seen an objectively answerable question closed as being opinion-based.

For many of the supposedly "opinion-based" questions, reasonable people will at least agree on the relevant trade-offs, even if they come to different conclusions overall. Thus, when a question asks about "pros and cons of technology X", expert insight into the trade-offs involved could be really useful to help evaluate your own specific situation. But that kind of more nuanced question doesn't really fit the simple "one correct answer" model that Stack Overflow seems to be going for and is thus very likely to be closed. IMHO, there should be a distinction between "it depends on the circumstances, and here's why" and "purely opinion based", and Stack Overflow would be more useful for me personally if the former were allowed.

Edit: Just to check if moderation policies have changed since I last looked, I did a quick search for questions containing the phrase "pros and cons". As unfortunately expected, practically all of the questions are closed: https://stackoverflow.com/search?q=pros+and+cons

While "it depends on the circumstances and here's why" is the ideal answer, such questions rarely get those answers.

On one of the sister sites to SO, there's a post about pros and cons on their meta - https://softwareengineering.meta.stackexchange.com/questions... - such questions are often a moderation headache when you get the inevitable spam and the poor quality answers of "I can't believe anyone didn't mention XYZ".

Look at https://stackoverflow.com/questions/15142/what-are-the-pros-... and start reading through and consider the one that starts out with "@Keith" responding to a different answer or "A SQL stored proc doesn't increase the performance of the query" being the entirety of the answer.

Such questions, when left open can get awful answers that are difficult to clean up.

Have you considered that there are sites with as laser like of a focus on pros and cons as SO maintains on Q&A? https://www.slant.co is one such site. https://www.slant.co/topics/607/~best-java-ides https://www.slant.co/topics/440/~best-nosql-databases-for-we... and so on.

> Opinion based

I wish Stackoverflow would buy https://slant.co and just point people to that site for opinions.

Why would you fire people on such a short notice? No need to hand over things? No interest in providing at least some time to look for a new job?

EDIT: To clarify that a bit, I am interested in what the upsides are except for possibly saving some money. An orderly shutdown usually seems preferable to me over quickly killing the process, for both sides.

Depends on country too. In EU our work laws regulate how your employee may be laid off. My country (poland) says that employee has to be notified about termination:

- week ahead if he or she was employed for less than month - month ahead if he or she was employed for less than three years - three months ahead if he or she was employed for more than three years

This goes both ways however, as employee has to notify his employeer on same basis that he or she quits. This introduces amount of games to recruitment process, where company frequently not only has to make bet by deciding to hire somebody, but also make sure that they'll wont be outbid during the three months that have to pass for person to change employeer

All that stuff doesn't really matter... If you have 3 months notice, it only means they'll come up to you and offer you to pay you those three months and you don't have to show up at all. And you sign a contract that prevents you from suing.

Most worker rights, in the end, become a matter of how much money they'll pay you to leave if they want you to leave.

Yes, one could--theoretically--fight back and not accept the money, but that only means you'd be a pariah in the company. What's the point of staying if the company wants you to leave and offers you a package? Take the money and get a new job in a couple weeks...

Well, where is the disadvantage here? As far as I can tell, instant layoffs are bad because you without any income at short notice. If you have three months where you still receive pay to search for a new job, you have gained a lot of security.

Besides, I doubt Poland has arbitration clauses, so the workers would be free to sue. It would not be pleasant to stay in that job either way as you correctly pointed out though.

The disadvantage is morale. It's much easier for the company to regain momentum if the chord is cut cleanly and swiftly.

That said, I'm not sure if it's true or effective, but having worked at Stack Overflow for years I'm pretty sure this was done to spare overall misery.

Same in France and the majority of EU countries.

Three months seems like way too long to know you're being laid off but I guess maybe the purpose is that it gives you time to find new work?

In the US, it's just assumed that the ex-employee will be as malicious as possible and so it's best to get rid of them fast as possible.

If the employer is worried for security reasons, they can send the employee home but must pay his salary for 3 months (in France at least, I don't know for other EU countries).

It’s called gardening leave in the UK

The same assumptions are made in the EU, the only difference is that the employee is let go, but still continues to receive his salary for some months.

Think of it more like a severance package and it makes more sense.

This is supposed to protect both the employee and the employer.

The employee to give him some time to find a new job (he is entitled to some days off during these 3 months to interview).

The employer to (at least theoretically) force the employee to still work to prepare a handover, docs, etc.

What happens in practice depends on the case : these 3 months may be shortened via a mutual agreement, the employer may force the employee not to come to work (but still pay him) and other arrangements are possible.

This only applies to permanent workers though. Contractors or temporary staff may be employed on much shorter terms.

I'm a student assistant at SAP in Germany. Even I (as a part time hourly-paid employee) have a one month notice period.

And you have to do it on the first of the month right? Because the one month doesn't become affective until the first of the month so if you quit mid-month then you actually have a month and a half left at the company. At least this is what I was told by a friend who lives and works in Berlin as a tech recruiter.

I’m quite surprised to hear one must give their employer notice. Doesn’t Poland or the EU have laws against forced labor? In the US, for example, all employment by private corporations is “at will.”

Anyone can quit at a moments notice, for any or no reason at all. For the sake of one’s colleagues and reputation it is widely considered good practice to give no less than 2 weeks notice, though this is merely custom.

It's very similar in most EU countries I know.

Responsibilities go both ways.

No one can "force" you to continue working somewhere, of course, but you may lose compensation for extra vacation time and could be liable for financial damage you cause for your employer. But I've never actually heard about that being an issue.

It's very common to just mutually agree to part ways right away.

And of course there are provisions both ways that allow immediate termination, like gross negligence from the employee or the employer not paying out your salary.

It's not "forced labour", it's just a contractual term. You certainly _can_ quit on zero notice but you've breached contract and it will cost you.

If you look closely at US startups you'll usually find the founders and senior staff are on "notice required" contracts of several months.

In the Netherlands (EU example) as en employee:

- You can do it in mutual agreement (see below).

- You can quit directly but you generally forgo any company benefits (like build up bonuses) and you might/most likely not have rights to state-supported unemployment benefits (70% of your last paycheck).

As an employer:

- You can either in mutual agreement (in which case the employee usually also keeps his/her rights since the contract does not end abruptly).

- By court order but this usually requires extensive documentation on malpractice of the employee.

Notice time usually is at least 1 full month in most cases, can be shorter if the employment is not time-fixed or by agreement. All notices are for both employer and employee.

I do not know exactly but if you have build up "vacation days" you usually cannot use those in your last month and forgo those, unless you agree with your employer in some sort of scheme. "Going in sick" will get you reported in which case a investigator will check you out and you most likely will forgo your sickness benefits. (I've had this once already, but not for contract finalization but reporting it really late. The guy was very surprised to see someone actually sick when he checked it out so I guess abuse happens often.)

There are some contract termination clauses (money mostly) in some contracts I've seen but as far as I know they are not enforceable in any way (since that equals servitude by power). Thinks like immediately handing back lease-cars and laptops under force of fine are though.

> If you look closely at US startups you'll usually find the founders and senior staff are on "notice required" contracts of several months.

Yes, but at least California courts have found these "garden leave" provisions in contracts are not enforceable. [1]

As for the rest of the United States, I'm not sure, but it seems like somewhat unsettled law as such notice requirements are quite rare in the US.

1. " And it turns out that companies also cannot require their employees to provide any specific length of notice, even when they offer to compensate them at 100% of their salary and benefits for the duration of the notice period. California courts have found these mandatory provisions to be an unenforceable restraint, as well." https://www.huffingtonpost.com/entry/understanding-californi...

If you make the breach cost infinitely high, how is that not a modern form of slavery?

It's not infinitely high? In practice the worst it's likely to be is "actual damages" - costs directly incurred by you quitting. http://employeradvice.org.uk/what-to-do-when-staff-quit-with...

Modern slavery is things like confiscating the passports of your immigrant workers so they can't escape.

Because the contract is voluntarily entered and the penalties for breaching it are purely financial.

All Europe operates in a similar way, the difference is only the notice period, which is shorter (few weeks) in some countries like UK or Netherlands, and longer (~3 months+ once you've worked for a while) in other countries like France.

Of course many companies find that it's not flexible enough for them, so they abuse the system and illegally hire low qualified workforce as "independent contractors" on civil contracts instead of work contracts. Search for "zero hour contracts" for UK, but it's rampant also in Poland, Italy and other countries where unions are not strong and employees' rights guaranteed by law are not enforced.

Living in the UK, my last job required three months notice to leave on paper. In reality they accepted a month which is standard enough.

Its because employment laws descend from those governing Masters and servants as does the USA's

A three months notice before quitting seems pretty excessive IMHO. I can kindof understand it from the perspective of the company that would need to find a replacement, but for the worker, it seems ridiculous.

Note: I am American, and most states are aren’t “right-to-work” meaning you can quit at a moment’s notice, but it’s common courtesy to give 2 weeks notice.

> most states are aren’t “right-to-work” meaning you can quit at a moment’s notice, but it’s common courtesy to give 2 weeks notice

The term "right to work" has nothing to do with employment status. It's political branding for laws that forbid mandatory union membership. It's a pretty effective union busting tool and has contributed to the decline of organized labor as a political force.

I think you're confusing it with "at will" employment.


> It's a pretty effective union busting tool and has contributed to the decline of organized labor as a political force.

I think it's fair to say teachers have the right to be teachers without joining a teachers' union. I don't think enforcing that right "contributes to the decline" of unions. Organizational models should be able to handle that sort of freedom to choose.

To me, the question is why unions don't innovate in their organizational models and policies so that giving employees freedom to choose is moot?

Same reason that people who live in the US don't have the freedom to choose whether they want to be governed by the US government. Social contract

But on a more concrete level, union fees go to campaigning, lobbying, and donations to candidates the union members may not support.

I don't think it's part of an American social contract that you would be forced to tangibly support candidates who have (by some accounts) immoral positions on abortion or immigration (to pick to polarizing issues).

For the worker it's actually great. It means that if the company wants to fire you they still need to pay your salary for at least the next three months.

Often times, especially in IT, when people are let go, it's often important for security reasons just to have them stop working immediately, and that worker is still going to get his salary.

When employees quit, it's not 3 months mandated, that's just the limit. You just work out your transition with your employer. Most job changes usually happen within just a few weeks.

Its normal for professional jobs and most eu residents someone only having 2 weeks notice would probably think oh so your a janitor or toilet cleaner

This probably depends a lot on the industry but my mom works at a grocery store and the reason they give short notices (as short as allowed by law) is because of the work moral. In Germany, there is no such thing as sick days, you can take off as long as you need if you are sick and have a doctors notice (which is mostly free or very inexpensive because of public healthcare). Most of the time someone is laid off, this person will suddenly become sick for most of their remaining time. When they come to work, they often procrastinate a lot more and get little to no work done. This means, once they have been given notice, they barely work and the employer can do very little. They cannot get fired (because they already are) and have to keep paying them for that duration. The same also often happens when the employee gives notice.

While the employer can take legal action and sue them, proving that someone is sick is rather difficult (and frowned upon by most judges). There are many cases where my mother's employer could have easily proven that (because witnesses same him partying or there are even pictures online where he partied on that day) but it is generally not worth it because a trail is more expensive then paying one month's pay and also not worth the time and overhead.

Not everyone takes work very seriously, unfortunately, which makes the process suck for everybody. When my mother's boss has an opportunity to fire them eithout notice (because they are late or caught steeling), he generally does that, which is super unfair for the employees who would have worked until their very last.

Likely security. It's not mutually exclusive to helping people with their future prospects.

We know nothing about severance, job assistance, etc. You can't really infer anything about this subject from that tweet.

Maybe the pressure was slowly building up from investors/board. They wanted the returns, profit, path to IPO , VC's carry for the Nth largest site of the internet.

From the https://www.crunchbase.com/organization/stack-overflow, the first round was from 2010 - 7 years ago. USV will like to get their exit soon.

I don't know, I figure that SO's investors made billions via value SO created for their other portfolio company. Investing in SO is like building a well or a power plant.

The transition is being handled with as much dignity as possible - The company is also offering to help place people through USV.

It wasn't possible to give more notice? Do you know why?

I've never heard of a company giving advanced notice of layoffs. Its always just an email the morning of.

Usually you will get at least a 4 week notice in Germany and this increases to up to 7 months when you worked for more than 20 years for the company. That you get fired effective immediately usually only happen if you really fucked something up.

"Layoff" and "firing" are separate concepts here. The central example of a true layoff is a trade work shortage: the layoff happens with very little notice, because it reflects the volatile nature of the demand for the trade. If there's suddenly, unexpectedly nothing for anyone to do, then there's suddenly, unexpectedly no benefit in paying anyone to come in for the day to do it.

Does that kind of situation still confer protections in Germany? If so, how do German manufacturing companies manage to not go bankrupt? Do all the companies subscribe to huge insurance pools?

As far as I know the barriers are even higher for layoffs. You can not, for example, arbitrarily pick which workers to layoff but have to minimize the social impact when you have to choose between several employees doing the same work, for example take age, spousal support, and things like that into consideration. But I am totally not an expert and I am sure it is a vastly more complex topic than I can imagine.

This seems backwards to me. Because my spouse has a better job than Mike's, I'm more likely to get laid off? Because I'm younger than Bill and can theoretically make up the financial impact sooner I'm more likely to get laid off?

If there are ten employees you need to cut, shouldn't they be cut based on either their recent performance, or a completely random method?

If there are ten employees you need to cut, shouldn't they be cut based on either their recent performance

No, because layoffs are position-based. If it's a performance related cull then that's not the same as redundancies. If you want to get rid of people like that then you need ironclad documentation and/or to provide generous packages.

In a hard time it tries to do the thing with the less overall impact.

Maybe it's putting lipstick on a pig, maybe it's fairness.

Layoffs (mass dismissal) are very hard, because usually it means the whole region has economic problems, and those that get laid off are especially vulnerable - because they almost by definition work in a sort of dying industry/trade.

Welcome to the world of progressive taxation. The more secure you are, the more you pay.

.. but there isn't suddenly, unexpectly, no need for the employees to pay their rent, mortgages and bills.

When I was made redundant a decade ago in the UK I got:

- week's "consultation period" during which the ranking process for who would be made redundant was explained

- pay in lieu of month's notice

- (optional on behalf of the employer) three month's pay for signing away my right to sue for wrongful dismissal

The only situation in which employees can really lose out is if the company goes bankrupt in which case they are at risk of losing their last paycheque.

Traditionally, large scale layoffs are hard and avoided, especially in industries where unions are strong. If required they are usually negotiated with the union. Still, there are often ways to avoid a layoff. The union sometimes agrees on a temporary pay cut across the board. There are some legal instruments that help a company in a crisis, Kurzarbeit is one of the most german one: The company can under some circumstances cut work time (and salaries) for a specified group of employees and the state will cover for a certain percentage of the wage loss. This allows the company to retain staff in an unexpected slow in business and be ready and fully staffed when orders start rolling in. In effect a giant insurance pool.

The US has much weaker employee protections. Aside from workers in unions and certain protected scenarios, layoffs usually occur with zero notice. (And, often, when associated with public bad financial returns, shortly after management denies any layoffs are coming.)

The WARN act is part of US labor law. https://en.wikipedia.org/wiki/Worker_Adjustment_and_Retraini...

For larger companies with layoffs beyond a threshold, they employer must give 60 days notice. I think they can instead give 60 days severance.

Yeah, but the threshold is quite large; 500+ workers at a single site, or both 50-499 workers and over a third of the site workforce.

The Stack Overflow layoffs are 60 people and 20% of their workforce, and so wouldn't (even at a single site) be subject to the WARN Act’s 60-day notice requirement.

True it could be stronger. Some states like California have more stringent versions.

These layoffs weren't in Germany, so I'm not sure that German law is relevant.


> Per Chapter 4, Part 4, Sections 1400-1408 of the Labor Code, WARN protects employees, their families, and communities by requiring that employers give a 60-day notice to the affected employees and both state and local representatives prior to a plant closing or mass layoff. Advance

The layoff that I was part of in '09 in California (Netapp, ~500 employees let go, about 6% of staff) had two different groups in the IT side of the house:

A. Out the door now. B. Pick your brain for some time.

The out the door now were let go immediately, though they were on payroll for 60 days. They weren't allowed in the building, but they were technically still employees. Severance package followed the 60 days.

The pick your brain group which I was part of were still allowed in and we worked. We had 30 days to sign a "increase pay and pick brain from Feb until July" or out the door with 30 days left on the WARN (if I remember it correctly). Come July, the 60 day window kicked in and then the severance package.

I am of the understanding that this approach isn't unusual with tech companies.

Interesting. Company I used to work for reduced their LA office to a handful of people after laying off 2/3 of its workers to move operations to Vancouver. Everyone was considered 'contract' despite being on W2s so not sure. I think we did get a notice but not months to look for new work. I had a week.

Companies benefit from terminating with zero notice, and subsequently providing money and services to facilitate the employees transitions.

Labor seeks regulations that makes it difficult to be suddenly terminated. The primary downside is that companies are reluctant to hire employees.

I'm OK being "instantly" terminated as long as you pay me for the duration of the notice period.

I agree with you, but many people do not share our view.

The general counter argument is that if you make it difficult for companies to terminate employees, then companies will search for alternate mechanisms to sustain profitability. For example, instead of terminating an employee, the company might reduce executive pay, or pass the increased expenses on to customers.

Again, I agree with you.

> instead of terminating an employee, the company might reduce executive pay,

Well that won't happen.

Apart from a few cases where executives agree to take all their compensation in equity to rescue a failing business, they tend to pay themselves first, sometimes at the cost of the long term health of the business. Philip Green / BHS is the example that comes to mind.

To be fair, your example is not a counterpoint because the company also collapsed.

I think you’d have to show examples where a company:

  1. Was not allowed to fire employees,
  2. Executives did not take a pay cut,
  3. The company continued to do well.


> it had found that “the main purpose of the sale of BHS was to postpone BHS’s insolvency to prevent a liability to the schemes falling due while it was part of the Taveta group of companies ultimately owned by the Green family, and/or that the effect of the sale was materially detrimental to the schemes.”

IE, Green underfunded the pension scheme in order to pay himself several hundred million pounds, and then sold the company off to escape liability for it. He "voluntarily" paid back £363 million in order to end an investigation that might have resulted in him being jailed.

No need to hand over things?

Correct. Remember ROLES are made redundant, not people. If the role no longer exists then by definition there's nothing to handover.

If there is a handover then there better also be a big chunk of cash to the departing worker in return for waiving the right to sue, because if there isn't, it's tribunal time.

That's a very simplistic view of the way work is distributed in a small company .

That would be for a tribunal to decide. But at the point that your employment is terminated, why would you do it?

Let's say that the law requires a 90 day notice. Maybe you can fire them today--as long as you pay them for 90 days while they sit at home. The spirit of the law is that you have sometime to look for another job so it's all covered.

My point is that we don't know the details, a tweet is way too short.

Why would you give notice? Leaving people dangling for weeks while you figure out who will be cut is worse than making the decision, telling them it's effective immediately and giving a fair severance package so they can immediately start looking.

The idea is to give notice after you figure out who will actually be cut, so that you can have an orderly handover of responsibility and knowledge.

That's how it's done in sane countries because notice periods are proscribed by law. And no, it does not lead to people vandalizing their workplace.

You're conflating notice and severance. It might be desirable to give enough notice so people can do more work, but it might not. I actually think it's more dignified to let people off the hook immediately. When you lay them off/fire them, you've typically failed in some way as a company, and expecting people to fulfill some responsibility knowing their time's up is pretty harsh.

In "sane countries" you can have people stop working immediately as long as you pay them. That's really all the notice period is for, it has nothing to do with continued responsibilities.

Tech company severance packages are usually pretty ok, I would bet stack overflow is giving packages that would be just fine in other countries.

The question isn't so much notice as it is severance. As long as the severance package is good, notice isn't that important.

You decide who needs to get cut and notify only those people. Then they have a few weeks to search for a new job and get their things in order.

Yes, but you forget "water cooler talk". Having worked at a large hedge fund that went through a massive layoff after a sub par year (we still made 10%) that resulted in 2/3s of all IT employees, roughly 600 people, being terminated. The announcement was made over email. First, all consultants/contractors were immediately shown the door. Those we wanted to keep were converted to salaried employees. This was the week before Thanksgiving in the US. Volunteers to be laid off from employees were asked for the Monday before Thanksgiving. Long of it was morale tanked. I informally asked my director to be let go and was told not to formally ask for it as they weren't going to cut me and it wouldn't look good in my official record. I saw a lot of friends and great developers walked out the door. The layoff was bad enough in of itself, but watching it dragged out over a few weeks was brutal.

The short notice is if the work is going away. It's an independent decision from how well they'll treat people going away. My most recent fund was similar - once they decided someone should go, they preferred to pull the trigger very quickly even if the severance was generous. (There's a toll on the org for having people stick around with negative attitude)

It's start up life, iterating off VC money. Fail fast also means fire fast.

Is it even possible to call Stack Exchange a startup anymore?

It definitely is, they dont have a solid monetization plan, consistently pivot/iterate product ideas and take quite a bit of VC money. It's something that has gone from MVP and is trying to transform into a real business.

Sometimes you just Robin Hood your VCs to build something wonderful to give to the world.

I tried to hire through them recently. It was a joke of a process. More like the way uaed cars are sold. Don't contact us, we'll contact you. No price displayed, we negotiate that depending on how much you can afford. The sign up system simply didn't work, saying I had already signed up but then the next button simply didn't work. They did get back to me a week later. This completely change my view. I went from wanting to use them to advertise for people to work here with like minds to a detractor.

I had the exact same experience recently. I was credit-card-in-hand ready to pay $500+ for a job posting on their site, but there was no way to do so. Then when they finally called me back, they gave me this story of how they were "didn't want to be thought of as a job postings site as they were far more than that" and said the only thing they would sell is subscriptions of multiple job postings a year. I ended up posting on some competing sites - I'm surprised they're willing to lose out on willing customers like this!

Indeed is doing a good job at hiring in the tech field. Their Prime product is great (if you live in one of the cities they are running it in).

(Disclaimer: I work @ Indeed, but I've seen friends quickly hired using Prime.)

It took me a while to understand that by "Indeed" you refer to the company.

Surprised no one has mentioned StackOverflow Docs, which was recently shut down as a failure. I unfortunately panned it at launch but it seems they devoted significant resources to the dead product.

StackOverflow Sunsetting Documentation:


Why I think they targeted the wrong market:



- it doesn't make sense to donate open-source docs to an offsite corporate service that might shut down

- also, the very common complaint that SO has poor moderation, groupthink doesn't work for docs like it does for one-topic answers

I remember reaching out to them a bunch of years ago (2012 or so) in order to ask them to host a stack overflow sub-site for our product (from a sizeable company). We were willing to pay quite a lot (because they had the premier UX for this kind of thing), but they weren't at all interested. They just told us to try to grow things "organically", because they had decided to focus 100% on open stuff.

We weren't really interested in their google juice, all we wanted was their actual functionality. Some non-stackoverflow domain would have been fine.

It just seemed like a missed opportunity on their end. I don't think we were the only company asking for this kind of service...

If I understand correctly, they offer this now, your private instance for your own product(s) & team(s):


Edit: while interesting, this (private internal instance for your team) is not what johansch is after (separate public instance dedicated to your product, for your users)

I looked into using their Enterprise product for my employer about 6 months ago. It seemed to be a nice product but was shockingly expensive compared to its counterparts in the space. Also, when they say "enterprise" they meant it. They are targeting large companies with 10s of thousands of employees; our 300 user company was really on the low side. Of course this is a pretty new product and this type of thing tells me that it's not quite mature yet and probably difficult to support.

this is very intriguing. Right now we use Confluence for something in this direction. I'd much rather us have a private stack overflow page for a subset of what we do on Confluence.

(Edit) This appears to be for non-public stuff only, though. So an internal stackoverflow.

Still no solution for a company wanting to host a public stackoverflow instance for some product.

Ah, yes, right, sorry. I understand now what you were after. Agree there would have been (is?) an interesting potential there too.

Considering there are companies that are completely based around selling private StackOverflow clones, I agree.

Maybe the Channels are what's you're after? Granted it's not out yet, but it seems a good fit for your use case.


I'm doing project in a similar vein right now, but in a totally different space. We call it the "Spotify model" where you host your stuff intermixed with ours, with the option to query just ours or just yours.

Good times - having to write a huge authentication / middleware layer to make sure everything is correct.

see Get Satisfaction, actually a little older than stackoverflow

Get Satifaction is far simpler in features than SO. I don't think you can compare the two at all. SO is a full proper Q&A site.

I've never personally seen a thriving Get Satisfaction instance. So maybe I'm being a little too hard on them. But I must've come across a hundred of them. I don't believe it's a well designed product.

It's been interesting watching SO try to monetize.

I wonder: How would people feel if they went the Wikipedia way? It's obviously a very beneficial site, but not as widely applicable as Wikipedia. I personally prefer the Wikipedia model of being ad-free and having no additional product and doing a fundraising drive every so often. PBS as well.

That said, I certainly think given the audience of SO that there are several opportunities for them, so it'll be interesting to see what works.

I remember seeing an article here these past few weeks talking about the business model of Wikipedia and how it's supposedly inherently flawed. Wikipedia doesn't bring in a lot of cash, but it rather sets a sad tone for itself how unfortunate that may seem.

Personally I'd be more interested in seeing StackOverflow branch out. What have they done these past few years? It seems like StackOverflow has remained stagnant, yet they have so much potential.

I'm not sure where you got the idea that Wikipedia doesn't bring in much. Their annual plan from last year [1] says that they expected to raise over $60M during the year. Unless anything's changed recently, they also try to keep over a year's worth of operating expenses in reserve, which means they're likely also sitting on almost $100M.

The complaints I see are usually related more to thinking that they really shouldn't need to spend so much (the FY17-18 plan for expenses is $76.8M [2]), and that their fundraising drives always make it seem like they're on the brink of death when they actually have a huge amount of money in the bank.

[1]: https://meta.wikimedia.org/wiki/Wikimedia_Foundation_Annual_...

[2]: https://meta.wikimedia.org/wiki/Special:MyLanguage/Wikimedia...

Wikipedia's primary purpose is not to bring in cash, it's a non-profit, so the money is going towards mission-related activities and supporting the site. Generally, the current level is enough to support the site (check also [1] which is meant to ensure at least basic level of support more or less forever). If donations dry up (which they don't seem to be) the activities may be reduced, but "bringing in a lot of cash" by itself is not really the purpose, the cash is the means to get things done. It's a bit different for for-profit establishments when usually getting profit on the investment is at least one of the goals of the whole deal.

[1] https://meta.wikimedia.org/wiki/Wikimedia_Endowment

They've done a lot, but apparently it hasn't been particularly successful. Their new "Documentation" platform didn't really ever gain traction: https://meta.stackoverflow.com/questions/354217/sunsetting-d...

I don't think I ever saw a Documentation page in a search result. Instead, I saw (and see) dozens of copies of the same SO Q&A content spread across multiple sites. I wonder if they would have had better success if they adjusted their SEO strategies and went after the rehosting sites (more?) aggressively.

Google was never pointed at Documentation, so you wouldn't find it there. There were constant concerns within the meta SO community about the quality of the content and how embarrassing it would be (not a small fraction of it was worse than W3Schools of old - https://meta.stackoverflow.com/questions/334638/ ). It could have been rather damaging to the SO brand to have such poor material shown as examples of what is desired.

Most of SO's operations are something I would not like to fund. I'd be happy to donate for their core service, but that's really only 10% or less of their headcount. Who funds the other 90% and why would they want to?

SO is too big to go for a Wikipedia style model unless they strip off just the core SO part and relevant engineers and make only that part donation based. And I'd guess the other 90% would end up in layoffs anyway.

> Most of SO's operations are something I would not like to fund.

Can you elaborate? Do you mean the other stack exchange sites like Photography, Gardening, Software Engineering, etc.? Or the careers site, or something else entirely? I'm only aware of the Stack Exchange network and careers.

Keep in mind that the Wikimedia Foundation also spends about 85% of the money they raise on things other than operational costs [1] (servers + engineering).

[1] http://wikimediafoundation.org/w/index.php?title=File%3A2014...

Plus another 20% for product development, and some other fuzzy categories like outreach

Very sorry to hear SO is financially in trouble. Also sorry for the people who lost their jobs. SO offers a valuable service. Sad to hear they can't monetize enough.

Edit: Reminds me of the situation of SoundCloud (company offering a service loved by their consumers, still can't monetize enough to satisfy investors, let alone cover the costs (huge headcount))

How many people work at StackOverflow?

Seeing this post also made me realize I have no clue how the company makes money.

I'm sure they make TONS of money from ad revenue. Probably they tried to generate other revenue streams but it just didn't work out and hence the layoff.

Here's a crazy idea: what if Stackoverflow developed a search engine for developers? Usually I get to Stackoverflow posts through Google but perhaps Stackoverflow can provide a better experience by doing code specific web crawling.

>I'm sure they make TONS of money from ad revenue.

An old news article said they only get ~33% from banner ads. Most revenue comes from from job listings:


Right, although I personally would consider job listings ads

True enough, but the details are different. And surely SO will have different internal departments and processes for managing each.

Inherent conflict here - programmers don't click on ads, fullstop. I can't imagine a group I'd want to advertise to less.

Clicks are not the only use of ads, and not always the point. They are easy to measure and so everybody tracks them, but it isn't the point. The point is to get people to buy.

Now it is understood that when I click on an ad for a widget and buy the widget it is easy to tell. However if I see an ad for a widget, and next week buy that widget the ad probably worked even though it is muck harder find a cause/effect relationship.

Thus we need to keep reminding people that clicks are rarely the point of ads. The point is the sale. Radio, TV, newspapers, magazines, park benches, bill boards (and many more) all have advertisements. The people who place those ads believe they drive sales long term - and they have statistical data to back that up. All of that data came without a single click.

StackOverflow ads are not the same as typical ads that technical users tend to block - they are typically just simple text links to job offers. Additionally, SO has fantastic analytics on logged in users which would drive the cost per impression up - if you've ever filled out their developer survey, you've given them everything they need to sell employers very expensive targeted ads directed at you.

I had to turn my ad blocker off to even see the ad you're referring to. Do you know of programmers who don't run ad blockers?

Of course. Are you saying all programmers run ad blockers? Just visit any [usually divisive] ad blocking Hacker News post and you'll see people who aren't blocking ads. Chances of none of them being developers is slim. Also I don't block ads. So that's one.

Of course. I know even more who whitelist sites with ads that aren't overbearing.

I have a hard time believing anybody actually goes "lemme turn off ads on this site to see how they are."

It's just something people say they do online to feel good.

The thing about ads is that you don’t need to click on them to work. You can’t click on billboards or newspaper ads either. Keep the product in mind is already most of what you want to achieve.

From a monetary aspect, impressions are basically worthless. You need clicks to make any money.

I haven’t advertised on SO, but the copy text on their advertising page doesn’t sound like they’re selling slots on a CTR basis. They seem to be well aware that clicks will be rare and are upfront about that. https://www.stackoverflowbusiness.com/advertise

That's true for generic keyword/retargeting ads, not premium placements.

Not true, I've clicked a few on accident.

Ah, fair point. Maybe I should have said "don't click on ads intentionally"

They do click on job ads.

Not the ones we don't see.

To be fair, all the ads on SO are vetted (I think they did a blog post about it) to be relevant to the target audience, so it's always about tech.

Do you have any numbers or references on that claim?

Numbers: https://pagefair.com/downloads/2017/01/PageFair-2017-Adblock...

"Adblock users in the US are 1.5x as likely to have a bachelor’s degree than the average American adult, increasing to 3x as likely among 18-24 year olds. Pronounced adblock usage among college-age respondents points to campuses as a major vector for adblock adoption."

Personally, I think being online without ad-blockers, VPN, and AV is like having public, unprotected sex with strangers.

So, I don't have a formal study saying that techies are the most likely to ad-block/least likely to click ads. What we know:

- Ad-blocker users skew young and wealthy.[1] The developer community as a whole also skews young and wealthy.

- If you are aware of the existence of ad-blocking technology, you probably use it.[2]

- Via comparison with other sites, you can also make some extrapolations. Take IGN - not a perfect proxy, but a reasonable one, with a core audience that is probably fairly demographically similar to the core audience for most developer-oriented sites. Approximately 40% of their traffic was using ad-blockers in 2015.[3] A Wired statement posted in 2016 has 20% of their traffic using ad-blockers.[4]

- Anecdotal evidence: every dev I've worked with has at least one ad-blocker installed. The vast majority of dev-adjacent people I've worked with - PMs, technical writers, designers - have ad-blockers installed.

Put it together and I think it makes a fairly compelling case that techies are the last audience on earth you'd want to orient your online marketing towards. I used to hope that something like The Deck[5], which was explicitly targeted towards "web, design & creative professionals", would be a good solution to this, and I even permitted their ads on Metafilter, but they shut down last year, presumably because they just weren't making enough money. They did everything that people claimed they wanted: the ads were unobtrusive, mostly text and optimized images, they didn't engage in tracking, they didn't sell user data (as far as I know), and they still couldn't make it work.

1. https://marketingland.com/ad-blocker-usage-highest-among-key... people-and-high-earners-143546

2. https://digiday.com/media/survey-80-percent-know-ad-blocking...

3. http://adage.com/article/digital/websites-hit-hardest-ad-blo...

4. https://www.wired.com/how-wired-is-going-to-handle-ad-blocki...

5. http://decknetwork.net/

People can claim all they want about wanting The Deck sort of stuff. But at the end of the day, those ads are still blocked most of the time. Possibly out of laziness to completely whitelist it, but the end result is the same. The ads are blocked like any other. I don't feel as optimistic that people blocking ads even want The Deck sort of ads. But without a study I'm okay not being sure.

Yeah, that's what I was trying to get across. I think the revealed preferences of most people using adblockers is that, while they say they would be ok with unobtrusive ads, they really don't want any ads at all. And to be honest, that's not a ludicrous view, because I've found that even on sites that have ostensibly "unobtrusive ads", all it takes is one bad actor and one slip-up in the ad network's verification process and unobtrusive ads become obtrusive and/or start invading your privacy. But we're going to have to find some other way to pay for content at some point.

Also, thanks for fixing that link.

Maybe, but a lot of sites only use Carbon/Deck or Adsense and some of those click bait native ad networks. I'm not sure if any of the click bait native ad networks have ever done too much with the issues you listed. I personally believe it's a bit too privileged to say no to Adsense like ads and native ads. You're likely to ever get a virus from them. At worst they are mildly annoying. That's all.

I find the one bad actor excuse to be just that. Another rationale or excuse.

So yeah I think the people just don't want any ads at all. Just like if I showed people how to block Spotify ads on their desktop and on how to do it on rooted android or iPhones, many people would stop paying Spotify. I don't tell my friends about these things for moral reasons, but it's just adding on to the point. People by and large will go with what's convenient. A simple app that can block Spotify ads will do for them. While jailbreaking would be too inconvenient.

Nice post with citations.

Your first link messes up though. Seems to be the correct link, but got cut off somehow. In case you don't or can't correct it anymore, it's: https://marketingland.com/ad-blocker-usage-highest-among-key...

I find their ads are way to expensive to trial. Minimum $500 spend for selve serve and minimum $5000 if you want to talk to their representatives. Most other sites allow you to dabble.

No doubt they could provide a better experience; the problem is probably monetizing that search engine. I have a hard time imagining an ad-based model working well.

Google is there main referral site. Probably not a good idea.

i doubt it. i would guess a much larger proportion of their audience uses ad blockers than the normal consumer media site.

I think they've been primarily monetizing with their careers section. The listings, last I checked, are entirely paid, and they have (at least had?) a sales team dedicated to them.

Their website says 250+

Do they not make money connecting people to jobs? (Take money from company taking people

Of all questions I've asked on Stack Overflow, 50% were eventually answered by myself (and some became pretty popular after that) and 5% was answered by someone else. 45% is still open.

Good to see them back to a focus. I guess hiring could be a good cash cow but all of the sub communities are a bit much.

Your first paragraph says more about you than about the site.

It's a bit of a humblebrag... but I think he's TRYING to say that meaningful StackOverflow activity seems to have stagnated in recent years, and he hopes to see an uptick.

It could also be the "normal" stuff has been asked many times over and he might be into a very niche/obscure topic.

I have a question that I was my own answer on. It was pretty obscure (pageant/cygwin specific) and when I finally found the answer I cited the 2 sources I used hoping it would save others time in the future with an easier google than what I went through

Right. Can't count the hours stack overflow saved for me although I'm disappointed by the lack of response.

The most popular answers and questions tend to be new things in popular SDK's. The unanswered things are deeper problems in existing API's. You either get an answer in a very short term by multiple people or nothing at all.

I feel SO must really be run as a non-profit like wikipedia.

I think an interesting path for stackoverflow would be to provide live channels for technology support and conversation. A bit like public slacks.

I remember a few technologies that officially said something like "support is provided through custom stackoverflow tags", but it didn't bring more functionalities.

A company trying to launch its new technology (i can imagine asp.net core for example) could have an associated community channel to talk about it, with automatic links from tags. That would be a nice addition.

Do you mean something like https://chat.stackoverflow.com/

I was hoping nobody would point me to something like that :D So either my idea sucks, or they didn't do a good job advertising the feature...

That's pretty much Gittr, and perhaps a few similar clones.

Good. In the past year they have become very preachy about exact which political opinions the tech community is allowed to have, and how they are to express them.

My limit was the "Time to take a stand incident" when Joel effectively dictated that the developer community of StackOverflow must agree with the statement.

> Carving up the world into ... nations ... is both morally repugnant and frankly stupid

The follow up of mods keeping the post open, backing it up and enforcing that idea on other questions really hammered home the idea that StackOverflow belongs to them, regardless of whatever they might say.

There was no room for nuance, just an American-centric political orthodoxy you must follow or aren't welcome.


Whatever your political views, the ellipses in the quote are deeply misleading, and in fact make his quote seem to say something Joel emphatically did not say.

Here is the full quote:

>It’s impossible not to see the parallel: the only way to build a successful world today is to allow the contributions of everyone. Carving up the world into us vs. them, building walls, and demonizing religions, nations, and refugees is both morally repugnant and counterproductive, and it goes so much against the spirit of Stack Overflow that as a community we must speak out.

That doesn't make it better in my opinion. The politics of nations, strong border controls and vetting for refugees are nuanced there are pros and cons to all angles in that debate.

Whereas, Joel dictated that if you supported "carving up the world" you were against the sprit of Stack Overflow, no matter how that might impact you as a non-American.

I'm happy that I was too busy at the time doing an handover to see that. Seeing the CEO get a pass on all the "Be-Nice" rules and use his site as a political platform...

The problem is more how, after creating a website, saying it's open and run by user, profiting (financially) from the work of the community, then saying "I'm above all the rules you users have to follow".

If SO and its related network sites die, what's going to happen to the millions of questions and answers their users contributed?

Is anyone archiving them and making the archives available in any useful way?

Nothing is dying. Our data dumps are live at https://archive.org/details/stackexchange as they've always been.

I wish your older snapshots weren't deleted. Any chance to get them back?

I had no idea this existed. This is awesome. Thanks for the link.

I don't know, but you guys remember the dark times before SO?...omg.

Now let's see.. we had Usenet news groups, mailing lists, IRC channels, and eventually web forums on every programming language, operating system, and software package under the sun. You could ask questions and get them answered. What was so cringeworthy about this exactly?

That's not to say that SO doesn't bring value. It has nice tagging and discoverability through search engines, good moderation, and an achievement and reputation system that encourages people to make quality contributions. But the days before it existed weren't so dark.

There was something called experts exchange. Remember that? It was horrible.

Funny you brought that up. I think it was Jeff Atwood who joked about how the domain name read like Expert-Sex-change

StackOverflow was such a welcome relief to this site and Atwood also wrote about expertsexchange in a blog post.

> I never appreciated how easy Experts-Exchange makes it for us. They are almost universally loathed. We don't just have a rival, we have a larger than life moustache-twirling, cape-wearing villain to contrast ourselves with.

Source: [ 2009 ] https://blog.codinghorror.com/whos-your-arch-enemy/

i remember that expert-sex-change was blocked by our schools filter. thinking about it triggers a vague angry feeling, no specific reason other than, possibly, often repeated disappointment the answer wasn't readable.

Yes. I was there. There wasn't even a google. I would stumble along on deja.com in the late 90s until it died a slow painful death and got scooped up by Google [1]

[1] Source: [2002] https://www.cnet.com/news/google-buys-remaining-deja-com-bus...

Bahahaha. That wasn't even the worst of it. Imagine this scenario, you have to install linux on a pc. There's no stackoverflow, there's no google. The linux distro you're using doesn't even have a package manager. And currently it doesn't have a working ethernet driver either. Also, it's the only computer that exists in your entire house.

The installation CD for SuSE back in the 1990s had a big how-to PDF, plus another documenting all the kernel compilation option flags ( we didn't have enough RAM then to run fat-and-happy kernels with everything included ).

I printed-out both of those over a long period on the work printers, using the reverse side of single-sided pages recovered from the paper recycling bins. And then settled in for a weekend...

I still remember choosing to include AX.25 in the kernel because it seemed like a cool thing with which I'd want to experiment at some point. Never got around to it!

Connecting to whatever internet connection I had at the time was always the hardest thing (if all of the other basic hardware was supported).

Maybe I was better off in that I was getting my distro (Slackware) along with the giant 'Linux for dummies' book in which it was included. It still never got me far enough to be able to remove my Windows partition.

I remember installing an early version of Red Hat on a machine that only had a Winmodem, not a real hardware modem that would just work in Linux.

Getting a Winmodem to work in Linux was possibly, but not exactly easy. I couldn't afford a better modem at the time, so I just kept hacking away at it and booting back into Windows and searching the web to try to figure it out. This was 1996 or so; there was far less information on the web than there is today, and search engines weren't exactly great at finding what you needed.

I think I finally got online from Linux after a week or two. :)

I have not-so-fond memories of trying to troubleshoot PPPoE issues.

I bought a book and had some floppies with SlackWare on it. The book didn't have install instructions, but it explained the basics of Unix and Linux. And SlackWare was pretty easy to install and worked well.

Also I was a teenage nerd with lots of time for home computing.

Copied ethernet driver through floppy disks from another machine.

For values of "another machine" possibly equal to "a system in the campus computer lab at school".

My first compiler was on a disk and I bought it from a video game store.

On the flip side, back then most of the programs I was writing could fit on a few screens and had almost no external dependencies besides libc and the C book was the SO for libc.

bruh, put a trigger warning before bringing that up! faints

It's not dying, I think their recent expansions in job/talent advertising isn't paying off as they originally thought.

Yeah, and it wasn't too long ago that SO Documentation was shut down as well. Hopefully Channels works out better for them.

I don't see why this person's post is downvoted. That was a good question.

I'll hasard a guess as to why it happened:

-The failure of the documentation (which failed for various reasons)

-VCs who want their return on their precious dollars.

Given some of the pettiness of my interactions with their mods, I kind of felt they might be a bit overstaffed alright ...

No notice? That seems a bit weird, no?


(╯°□°)╯︵ ┻━┻

So sorry to hear this. My thoughts to everyone, it is a very difficult day.

I would advise people to stop comparing US workers rights to their countries. Just because you think your country gives you extra leverage and protection it does not mean it is a universally applicable law

Compared to the rest of the developed world, the U.S. does have very poor working conditions. I'd also say a majority of U.S. workers doesn't have a clue and has little experience to make a comparison anyhow.

I actually like people outside my country reminding my fellow citizens that we do have it pretty bad and it is in our power to change it if we decide to do that one day. The message might eventually soak in after some decades have passed.

A large portion of the laid off workers were from the EU, under EU law.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact