I thoroughly investigated every open-source (mostly open-core) self-hostable Slack alternative, and the conclusion was clear: only a self-hosted Matrix is a viable option.
There are no arbitrary limits for 'community editions', no risk of relicensing, no risk of being held hostage for features (like Gitlab did at some point).
You can work around all the missing features easily with self-built webhooks and other tools.
Starting with Mattermost or Zulip or similar is just way too risky.
What about their mobile push notifications? I don’t understand whether that’s something one can self-host. I have just read Zulip’s https://zulip.readthedocs.io/en/latest/production/mobile-pus... and they give me the impression only they can do this, so ‘pay us money or else’
I wonder whether this is something any other provider can rug pull in the figure and to be worried about.
There's a very common line of thinking that goes like this:
From the end of WWII until the fall of communism, the public in the West (as opposed to the elites) enjoyed much better treatment, and prospered more than ever before or since. This would include both fiscal gains, and the public's opinion being truly taken into consideration. This is mainly because the elites were afraid of people turning socialist / communist, so they gave them a reason to actually be invested in the system. Once that threat of communism evaporated, the elites could proceed to gut the majority as in the previous centuries with no fear whatsoever.
My comments:
I'm not sure I agree with that, though, too simplistic. On the other hand, I also think that people have a rose-tinted view of what "democracy" always was - with enough money / media control and a bit of time, you can convince the majority of anything, anywhere. Letting people prosper does make it easier. Maybe it did play a bit of a role. A counter argument is that (independent) media coverage made the Vietnam war unpopular, and then the US pulled out because of that, a miracle of democracy which never really came close to happening again ever after.
But I think the USSR itself murdered any real chances of communism's further spread in 1968, when they invaded Czechoslovakia. (The Hungarian thing in 1956 isn't nearly as important because of country's undeniable previous Axis affiliation; few had sympathy for that back then). The US and west in general couldn't get rid of their Woody Guthries, and their Klaus Fuchses, until USSR did it for them through sheer idiocy. But after that, was communism really a threat?
But I do think that the 1950s policies were affected by the war (+ Korean war) even more than communism itself. All these traumatized vets, desensitized to violence, were now back home, and the elites were truly afraid. But that doesn't seem like it brought democracy in today's sense of the word? There's a reason why feminism regressed in the 50s - letting men be little despots in their own (cheaply bought) homes was the least the government could do. But that seems to have lasted only until the mid 60s, then the Vietnam thing happened, ... Let's not go further.
Is there a device you can recommend for printing (sticky) labels occasionally? I have a little Brother printer for those narrow little labels, one with a rubber keyboard, but would love something with sticky labels AND Linux connectivity. Something I could script when organizing my workspace, parts, ... to print the appropriate label.
If your printer speaks ZPL, I might have a solution in the near future. I'm working on a ZPL server that handles printers (USB and network), label templates, CSV uploads (for batch printing labels), and the like.
But on the other hand, let's not kid ourselves, array out of bounds, use after free, resource leaks and bad type system, all of this isn't even close to an exhaustive list of C downsides. Beyond its direct limitations, C inspires an approach that is vastly inferior even if you follow all the best practices. Even compared to (modern) C++ it's much worse. I say this and I kind of like C.
If the approaches described in the article save us 30% of the effort of translating C codebases to Rust, it's still worth trying; we're unfortunately not very close to complete automation, but that's something worthy of pursuit.
When you hear this reiterated by employees, who actually believe it, then it's sad. Obviously not in this situation, but I've actually heard this from people. Some of them were even pros. "There is no fool like an educated fool."
This is one of the most fascinating things I've found in the past ten or so years. When did people in general begin to buy into the bullshit spewed by big shots in corporate or really any commercial venture. People at least implicitly understood that the boss just wanted money and would fuck you, nature or his own firmly held beliefs to get it.
People are now shocked when a company cuts a loved product or their boss fires them when someone cheaper comes along.
Anyone who has worked at OpenAI or is currently working there has lost all credibility in my eyes. When their dear leader, Sam was "fired", they staged a coup to save their paychecks.
These people are just out there to may a buck and scam people with "AGI" and now that there is plenty of competition and superior models, I'm hearing crickets from them.
All they had going for them was first to market and they managed to damage the brand, lose their top talent, deliver a subpar product and convert a nonprofit into for profit.
Woodworers make 70% - 200% of dev wages in Slovenia, especially if you're decent at it.
It's of course quite dull, kitchens, cabinets, ... Mostly tailored to living spaces of upper middle class and rich people of varying taste, so you'll be doing a bunch of tacky stuff no doubt.
There was always an underlying Randian impulse to the EA crowd - as if we could solve any issue if we just get the right minds onto tackling the problem. The black-and-white thinking, group think, hero worship and charicaturist literature are all there.
When I was a kid in the 90s and early 00s in Serbia, later Montenegro, I knew 2-3 kids that had leukemia at some point, and if my memory serves me well, all of them survived. One neighbor kid had it really rough (was in therapy for years, looked like hell), but still survived. Even then, the treatments were so available that surviving it was the expected outcome.
Just years before it was quite common for children to die from it - I know 2 couples who lost kids to it in the 80s and early 90s.
Another thing was that a couple of kids had congenital heart conditions. Those didn't fare that well. My classmate from elementary survived that with a pacemaker, but a neighbor suddenly died in her mid-20s, that was really sad.
There are no arbitrary limits for 'community editions', no risk of relicensing, no risk of being held hostage for features (like Gitlab did at some point).
You can work around all the missing features easily with self-built webhooks and other tools.
Starting with Mattermost or Zulip or similar is just way too risky.
reply