1,739 karma · joined August 21, 2009
A great idea in theory, but in practice just too hard.
There is a great deal of genius in coming up with these ideas.
There is just as much genius in finding the 80-90% solution that incorporates most of the original, but in a workable way.
Looking forward to seeing where stuff like Nix ends up.
In fact, at the risk levels that we are accustomed to dealing with on a day-to-day basis, there are plenty of adequate solutions to disposal and storage of nuclear waste.
Disclaimer: H2 dev
We build several products in Java using H2 as our database. What a pleasure. Full SQL support, zero configuration/installation/etc.
Just copy data-files/dbs around using normal file tools.
Just start up one process and your app is ready.
Which is not surprising, because pipewire, too, decided to start from from scratch instead of working inside gstreamer.
(a) create a sandboxed child process with some code inside it (b) talk to that sandboxed child process (e.g. Firefox and Chrome both have IPC libraries) (c) handle the resulting crashes nicely, restarting child process when necessary
Now if only there was a nice open-source, cross-platform way to do this..... hint..... hint......
:-)
For a number of useful applications, this is exactly the right, correct, and most useful approach.
I currently maintain several successful (within our commercial niche) 100kLOC+ programs that largely use such an architecture.
It puts the error-handling code in one place, and enables common logging, recovery, filtering and display.
It means that the vast majority of the code can happily just assume that the world is full of unicorns and light.
And given that it is written in Java, the program just largely keeps on running, even in the presence of bugs and weird edge cases, and suchlike, a feature our users really like.
Human are pretty good at going "OK, so that part of the program is having a bad day, I'll report the bug and keep on using the rest of the program".
You're talking about people who wear tactical gear to go to the corner store. Who carry guns and flex-cuffs because they are desperate to be some kind of "big man".
Probably a handful of them intended to actually "do" something.
But mostly they just milled about looking like penguins escaping the zoo.
For example, someone “donated” a gradle build script to an open source project I help maintain. It lasted less than 6 months before it began breaking, not because we changed stuff, but because gradle updates had changed how some things worked and a plug-in stopped working. Now “obviously” the answer is that we should have pegged the version of gradle required to use the script, but (a) I don’t even know if that’s possible (b) the donator definitely didn’t think to do that (c) who changes such things in such a production level tool during a stable release cycle?
Now the gradle people certainly seem decent, in my interactions with them, but I think there is unfortunately a strong case of Stockholm Syndrome going on there.
Personally, in this situation, I would find some multithreaded RPC proxy thingy where a single instance would sit between the worker processes and the rest of the services.
NFS is just about impossible (except in narrowly restricted situations)
There are however, a couple of small companies, and some individuals using it, and providing contributions.
I was writing UIs in Swing 10 years ago, and we managed fine even back then, but our target customer is technical people so maybe our definition of "looks good" is somewhat easier.
But we are trying it, and opinions (internal to my company) are varied. Some people like it, but I don't. Because (a) it has missing functionality here and there (b) it still has a long road to go to become battle-hardened (c) I don't see any of the current supporters having deep enough pockets to fix (a) and (b)
So we're taking a mostly wait-and-see attitude.
But we have an existing deep and wide investment in Swing in our products, so our risk profile is probably somewhat different to yours.
If you need 3d, we drop into OpenGL and use that (e.g. when integrating WorldWind).
Sometimes I feel bad about how easy it is to write desktop UIs in Java - the API is stable, nobody keeps changing it under me, I don't have to wonder what Microsoft is pushing this year, it is (mostly) easy to debug, has open source forms designers (Netbeans), etc, etc.
Even if you need to implement some kind of "Enterprise Java" app, you can do so with much better libraries and tools than back then, that do not suffer from the excessive abstraction problem.
The driving force behind Wayland appears to be commercial interests that have no interest in the desktop use-case per-se.
The linux-only outcome is purely an unintended side-effect.
The wayland people replaced that with a half-baked solution because they insisted on boiling the ocean - replacing the entire thing in one go, instead of working piecemeal (which the X protocol was explicitly designed to allow).
Which is a great pity, because now the day of the Linux Desktop is even further off.