I have a hard time feeling sorry for Stack Overflow. They had most developers in the world visiting frequently. They have detailed information on exactly which languages, libraries, toolchains, and platforms those developers use and their pain points. All of their content is given to them for free by volunteers. All of their moderation is done for them for free by volunteers. The only thing they have to serve is text. How can you fail to build a sustainable business on top of that‽
Having a lean and mean company doesn't make for happy investors.
The company was finished in 2020 when they raised $85M in a Series E. The fall after that is inevitable. Even the $40M they took in 2015 is a questionable decision for the very reasons you detail. What did they need tens of millions in investments for?
This happens so, so often; companies raising too much money at too high a valuation, based on unrealistic “moonshot” expectations, which then fails and they’ll have to pivot and refocus and they’ll be a mere shell of their former self.
Raising at too high a valuation is not a position you want to be in. Unless you’re planning to cash in sooner than the ship will start burning.
Pretty much nobody would use that on a standalone keyboard either. They had to know the design isn't practical for the stated purpose.
However, it does make a nice little trophy/trinket. It's merit based, so getting one could become a goal for contributors. It's branded. It fits on a desk.
If they made it twice as big and called it a trophy, it'd probably make more sense. But then costs would be higher and it'd be more formal.
Look up "artisan keycaps". Not only do people use them, they collect them, and certain specific rare and desirable artisan keycaps can be worth a lot of money. The high quality ones tend to be cast out of resin in small batches (sometimes literally just 1 in a given color/sculpt combination) and sold in short-duration raffles. People tend to put them on less-used keys.
I'm a senior dev. The only thing I found useful in the last years was the job market on SO. They closed that down recently. Now I don't even go there at all.
It still boggles my mind that they closed this down. To date, it was the best job board for software engineers, and had filters you just don't see anywhere else - like filtering by companies that sponsor visas.
Finding work there was a dream, and while there weren't many roles available, they were usually great quality roles. Finding candidates was also great, and all of them were solid candidates that met the basic bar of "can you write some basic code in a language of your choosing".
To be honest, it felt like SO went downhill when Jeff Atwood left.
I'm a senior dev. Which means my knowledge and experience is often on higher level stuff. Hence I often need to look up details I don't bother to remember. I'm not ashamed to say I search for lots of stuff, and often get results from SO.
If anything, pretending to be completely knowledgeable "since one's a senior" is quite junior behavior.
Also: sometimes I need to work with stuff I'm not so familiar with; no one is "senior" across the board. Even though I did Ruby full-time for a few years, it's been a while and I forgot a lot of basic boring stuff so I will end up searching things like "Ruby remove whitespace from string" (trim? strip? chop? chomp? delete? What were the differences again?)
And even inside my main expertise, I will sometimes search for an answer just to verify that my thinking is correct and that I didn't miss anything (which is not a strong guarantee that I didn't, but better than nothing).
That's where GPT-4 ate their lunch. I've never really administered servers myself before, but I was able to set up multi-container CI/CD on bare servers and monitoring with ELK, filebeat and metricbeat and public-faced services through nginx with SSL certificates and basic password authentication all in a single day.
Before, I would use stack overflow and it would take me at least a week.
I work in cloud consulting and I had a project where I needed to create a CI/CD proof of concept for the customer to use that involved deploying a Docker container to ECS (AWS’s Docker orchestrator) and show how they could integrate automated testing with the pipeline.
They use Java - a language I literally haven’t used since maybe a year after it was introduced.
I used ChatGPT to create the source code for the sample based on my specifications including Junit tests, the build with Maven, the Dockerfile - everything.
Of course it got a few things wrong and missed steps and I had to keep pasting in error messages until it got it right. But it was still faster than trolling the internet.
Of course I was very up front with the client that the Java pieces were all generated by ChatGPT.
Even though the sibling comment said I was technically not wrong, I think “trawling” is the better term.
And I assume trawling vs trolling is one of those things that you feel the urge to point out just like I hate when someone says “jive” instead of “jibe”.
That last sentence was not meant to be passive aggressive at all. It was just an observation.
To each their own. I haven't had the necessity to use it, and when I searched for problems the answers were mostly not applicable. All in all it wasn't worth the effort. I just read the source code of the lib I use instead. This tends to give me more for my time.
I understand that everybody has their own unique situation and my statement wasn't meant to be generalized anyways.
Stack Overflow job ads were the best part of the site. They were affordable enough for small companies to run them, and they were the only place that actually showed compensation info front and center. They didn't have the most job ads, but they had really good ones.
All they had to do was leave it alone. Maybe leave a team to do moderation and keep things running operationally, and done.
Heck, spin it off as a sub company and resell the stack overflow brand and do nothing.
They actively went out of their way to get rid of it. If I didn't know any better I'd say this was backroom dealing where job ad company owners somehow pressured the CEO to stop competing with them because they were just too good.
Sounds like you could do with learning something new!
My 30+ years experience doesn't matter for shit when I say... want to fix an annoyance in an open source tool using an unfamiliar language. I'm back to "how do you deserialise json" level query. Offical docs are typically either useless auto-generated placeholders or over-detailed rabbit warren not to mention there are usually five ways to do anything and I need to know the blessed approach not just any approach. I want a few lines of sample code and some confidence that it's the approapriate method and not 10 years out of date. It's what a QA site should excel at and exceed ChatRoulleteGPT answers given social proof from real people.
Why.. why did they closed this down? I remember this was my single most successful place to find good software developers / run job ads. We spent thousands and thousands on these ads, and I always felt this was their chicken that laid golden eggs.
> Exiting this space allows us to refocus on products that build on our core strengths: knowledge reuse and building communities at scale.
A year has passed and I'm not sure what progress have we seen on the "knowledge reuse" and "building communities" front.
The post has this quote from the CEO:
> We are realigning the Talent business to focus more on customer employer branding and company awareness needs, and moving away from job slots and direct hiring. This will tie the product closer to what we offer through Stack Overflow Advertising [...]
The article linked by OP (related to generative AI) has this quoute, on the other hand:
> There's definitely a question around how we leverage [generative AI] technology to deliver on our mission of helping build technology through collective knowledge. This intersection between the power of community on one side and AI on the other side—from my standpoint, human-generated community content has taken us to this level, we have a large impact, but there are also so many problems we can solve by leveraging this technology.
Reading these CEO's quotes filled with words like "realigning" and "leveraging" makes him sound like a typical MBA executive (yes, he's a Harvard MBA) who thinks that the same rules can be applied to a business in any industry. All while completely ignoring the feedback of a business's users/customers.
> I always felt this was their chicken that laid golden eggs.
But was it? Did they ever report on whether it brought them any profits? Potentially, its pricing was too low to make a good profit, but increasing it to reach profits or even just break even would have turned too many customers away.
Given they were bought by an investment firm for quite a frothy amount of money I'd assume it was at least partially debt leveraged and they're suffering from interest payments.
