Namrok wrote on 2025-07-05, 17:41:

the3dfxdude wrote on 2025-07-05, 16:59:

Living wrote on 2025-07-05, 12:39:

that part comes when you want help online and find the toxic community that Linux is built around

Linux is not built around toxic online communities. The key to what you are saying is toxic online communities. It's toxic because it's online because there are cowards that wouldn't act this way otherwise because they get away with it online. Just find other people, preferably a friend you know that uses linux and it won't be toxic, and he'll show you things. Then you'll find that linux is just another OS when you remove yourself from toxic communities, getting away from the net a little. I know dozens of people that use linux, everyday, main OS, work, games, etc. They just use it, and just works. And nobody talks about toxic people online and then make the conclusion that linux is toxic.

You know, with the complete collapse of several communities I used to be a part of, I think I'm coming around to the idea that communities need a certain level of "toxicity" to maintain any sort of identity at all. And yeah, it's obnoxious when you ask a stupid question and some Linux nerd berates you for not knowing the shell script to fix it off the top of your head. But then again, that same community erupts in "toxicity" when Linux distros attempt to add telemetry, or drop 32-bit support, or other BS a meeker community might roll over and take in the name of not being "toxic".

Toxic != Having an opinion, explaining one's reasoning, citing facts, etc

Civil people can and do agree to disagree.

Having principles, ethics and a spine is not toxic.

IMHO, toxicity starts when humility and/or respect for others end.

Intel486dx33 wrote on 2025-07-04, 23:37:

Apple CEO Tim Cook said “what good is RAM if you are not using it”.

It's funny to hear that from the company that still puts 8GB of RAM in a lot of their computers (and claims that it's just as good as 16GB of RAM on Windows). I keep hearing about Apple Silicon Macs having high SSD failure rates from virtual memory usage.
[/quote]

It's funny that people seem to repeat the same message forever, even when facts have already changed. 😉

How do you expect Linux to thrive if most "programmers" can't even get their shit working properly on Windows in higher-level languages?

darry wrote on 2025-07-05, 23:29:

Toxic != Having an opinion, explaining one's reasoning, citing facts, etc […]
Show full quote

Namrok wrote on 2025-07-05, 17:41:

You know, with the complete collapse of several communities I used to be a part of, I think I'm coming around to the idea that communities need a certain level of "toxicity" to maintain any sort of identity at all. And yeah, it's obnoxious when you ask a stupid question and some Linux nerd berates you for not knowing the shell script to fix it off the top of your head. But then again, that same community erupts in "toxicity" when Linux distros attempt to add telemetry, or drop 32-bit support, or other BS a meeker community might roll over and take in the name of not being "toxic".

Toxic != Having an opinion, explaining one's reasoning, citing facts, etc

Civil people can and do agree to disagree.

Having principles, ethics and a spine is not toxic.

IMHO, toxicity starts when humility and/or respect for others end.

I basically agree with the both of you. People should be able to give feedback to their chosen vendor even if it sounds harsh. (one example is Microsoft, as with here) Because how else does a provider even know if know one speaks up? It's when the users, start taking sides, with one taking the side of the fanboys, that Microsoft can do nothing wrong, and best thing since sliced bread, and the users, threatening to leave, because they are dropping 32-bit support (or whatever TPM junky thing is required). But to the fanboys, it's always "get with the times". That's when I'm done with that conversation, because when it's users fighting users, it's just a waste of time. Linux does have both happening as well, and deserves feedback as well. But it isn't Linux's fault that a fight breaks out between the vendors and the users, but something just as necessary as when it happens with Microsoft, and the same will happen with Red Hat as that is likely a source people are getting Linux through. But when, online, the fanboys stand up in the way and disrupt the conversations, well... I'd just walk around them. Online forums are probably not really the place to provide feedback to the vendor. I do have to commend the people at vogons are generally pretty nice and don't wade into the weeds over controversies.

The most common linux failures I get involve /etc/fstab. It's still no windows volume management. Take a physical drive out before commenting the relevant volumes to mount? Have a power interruption and it has to fsck itself? Oh no you can't go to the desktop!!!!

leileilol wrote on 2025-07-06, 18:13:

Oh no you can't go to the desktop!!!!

Hint: Linux is perfectly usable without the desktop.

UCyborg wrote on 2025-07-06, 12:51:

How do you expect Linux to thrive if most "programmers" can't even get their shit working properly on Windows in higher-level languages?

I have to wonder how much stuff even works by pure accident. (don't mind me I'm still amused by your notes when you were working on the I76 AiO patch)

Grzyb wrote on 2025-07-06, 19:59:

Hint: Linux is perfectly usable without the desktop.

Hint: Look at the bigger picture; no one is willingly dumping win10 to use GNU/emacs in a fucking prompt.

There's many usability issues with Linux currently and I'm not brushing those away. xfce4's my favorite DE and that has a lot of things it can't do that Windows 95 can.

Also good luck doing art and playing games without a desktop 🤣.

leileilol wrote on 2025-07-06, 18:13:

The most common linux failures I get involve /etc/fstab. It's still no windows volume management. Take a physical drive out before commenting the relevant volumes to mount? Have a power interruption and it has to fsck itself? Oh no you can't go to the desktop!!!!

If on windows, you can't boot from your system drive due to a corrupted environment, you'll get a black screen telling you to boot a setup disk to try to repair. You won't get a desktop either. How do you expect to get a desktop from a broken system? Either use the provided shell if possible or boot from a portable linux system or rescue disk and get a "desktop". There are many choices here if you like using a graphical desktop.

In 25 years of using linux, fsck was automatic, fixed and done, and continued booting the system, except once. That one time actually was due to faulty hardware.

Did you pull out a drive needed to boot the system?

the3dfxdude wrote on 2025-07-07, 03:26:

If on windows, you can't boot from your system drive due to a corrupted environment, you'll get a black screen telling you to boot a setup disk to try to repair.

In linux's case, it didn't even have to be a system volume for it to panic- just one id it can't find and you're sent to shell. Nice try though

leileilol wrote on 2025-07-07, 03:39:

the3dfxdude wrote on 2025-07-07, 03:26:

If on windows, you can't boot from your system drive due to a corrupted environment, you'll get a black screen telling you to boot a setup disk to try to repair.

In linux's case, it didn't even have to be a system volume for it to panic- just one id it can't find and you're sent to shell. Nice try though

Nice try, what? You configured a filesystem to be required at boot, and to be checked. Then you either pulled the drive or corrupted the system to not boot. If the disk is not required to boot, then don't configure it to be required to be checked and mounted at boot. Besides, there are such things as rescue disks to save you from not having a gui desktop. This is not linux's fault. You'll get the same thing if you use windows that way.

DracoNihil wrote on 2025-07-06, 21:35:

I have to wonder how much stuff even works by pure accident. (don't mind me I'm still amused by your notes when you were working on the I76 AiO patch)

Oh that's one fun example. Don't get me wrong, programming is hard, hence why I'll never be one, I'm too dumb, admittedly. And I get overwhelmed quickly. I just got the hunch that most of us, including developers, don't even really know what we're doing.

I found a solution to further improve stability in Interstate '76, now nGlide wrapper and audio survive for longer periods, just have to tidy up a bit and transplant patches to the expansion, which I haven't got around to yet.

leileilol wrote on 2025-07-06, 18:13:

The most common linux failures I get involve /etc/fstab. It's still no windows volume management. Take a physical drive out before commenting the relevant volumes to mount? Have a power interruption and it has to fsck itself? Oh no you can't go to the desktop!!!!

What happened to me not too long ago on older KUbuntu, when Cloudflare still locked up Pale Moon web browser forever, the whole graphical session froze. Didn't respond to CTRL + Alt + Backspace either. I had to push reset button and when it was supposed to start booting, all I got was a blinking cursor. I ended up fixing it by running fsck on its partition from live distro on USB.

So if fsck was supposed to be kick in automatically, it didn't.

UCyborg wrote on 2025-07-07, 11:03:

now nGlide wrapper and audio survive for longer periods, just have to tidy up a bit and transplant patches to the expansion, which I haven't got around to yet.

Unfortunately recent regressions in Wine make the nGlide wrapper impossible to use, it just causes Wine to segfault from some stack/buffer overflow.

I primarily play the game with the -gdi switch because the software renderer both looks nicer to me and does things the hardware renderers don't. (the "motion blur" when you get shellshocked from collisions)

UCyborg wrote on 2025-07-07, 11:03:

What happened to me not too long ago on older KUbuntu, when Cloudflare still locked up Pale Moon web browser forever, the whole graphical session froze. Didn't respond to CTRL + Alt + Backspace either. I had to push reset button and when it was supposed to start booting, all I got was a blinking cursor. I ended up fixing it by running fsck on its partition from live distro on USB.

I've had Linux freeze on me (with nothing listed in journalctl), and when I restarted, it was able to check and fix everything except NTFS partitions. There are separate NTFS tools (I don't remember the name now), but they didn't make the drive accessible again. The only to make NTFS partitions accessible again was to plug the drive into a Windows machine and run chkdsk. Chkdsk always said that there were no errors (and didn't even say that the partition was marked as dirty), but the drive was accessible in Linux again.

While I'm on that subject, I don't think there is a good cross-platform-compatible file system. I have one drive that's formatted as exFAT, and it takes multiple tries to mount. It will give errors the first couple tries. I don't know if it's just that one drive.

@DracoNihil
It may just glitch somewhere else without nGlide, the issue being with using memory from default process heap (where other libraries may allocate from), which it tends to corrupt. Yeah, there always had to be some compromises with hardware renderers back then. But oh my, 640x480 window isn't very immersive.

I once brought about up a bug in WINE with inet_addr() on their forum. Windows XP and earlier return INADDR_ANY if you pass an empty string ("" in C), Server 2003 and later return INADDR_NONE. The former is not emulated if you choose compatibility with older Windows, the latter seems to be standard behavior among different implementations. They didn't want to hear about it and locked the thread just because I mentioned poking around in an old game, I think I said "reverse engineering".

@lti
Last time I checked, NTFS checking in Linux was very basic and it marks partition explicitly as dirty to be later checked by Windows' chkdsk. Windows has the message at boot saying "one of your disks needs to be checked for consistency", if I'm not mistaken, that's the indicator of partition flagged as dirty. Not sure if term "dirty" is used anywhere else in Windows except fsutil.exe.

Strange, exFAT is no longer proprietary and doesn't have permissions to worry about. Are you sure the drive is in good working order?

UCyborg wrote on 2025-07-08, 08:14:

I once brought about up a bug in WINE with inet_addr() on their forum. Windows XP and earlier return INADDR_ANY if you pass an empty string ("" in C), Server 2003 and later return INADDR_NONE. The former is not emulated if you choose compatibility with older Windows, the latter seems to be standard behavior among different implementations. They didn't want to hear about it and locked the thread just because I mentioned poking around in an old game, I think I said "reverse engineering".

That's strange you had someone lock your thread over information that can be obtained by a test case without reverse engineering. It's probably someone who just didn't want to deal with whatever you were saying. If you send in the test case for inet_addr for their test case suite, they should be running the Wine tests across different versions of actual Windows systems, and they'll see the problem and they will know they have to fix it. So if it is not fixed now, then they are not testing for that and don't see the problem. If what you found is causing a problem in a program, we probably should try to send in the fix. If you have documented cases like this you want to get fixes into Wine, message me and I'll see if I can help.

You can read it right there in Microsoft's own documentation!

https://learn.microsoft.com/en-us/windows/win … sock2-inet_addr

On Windows Server 2003 and later if the string in the cp parameter is an empty string, then inet_addr returns the value INADDR_NONE. If NULL is passed in the cp parameter, then inet_addr returns the value INADDR_NONE.

On Windows XP and earlier if the string in the cp parameter is an empty string, then inet_addr returns the value INADDR_ANY. If NULL is passed in the cp parameter, then inet_addr returns the value INADDR_NONE.

Official versions of Drakan: Order of the Flame, the ones using Windows Sockets for networking rather than DirectPlay, you normally can't start multiplayer server, gives you session creation failed error, unless you know to pass -ip 0.0.0.0 cmd-line parameter to the game, which I guess almost no one knew at the time. I guess -ip " " should work as well and of course, specific IP of the adapter you want the server bound to works as well.

This game is probably rare example where this quirk shows in, Microsoft doesn't emulate it in their compatibility modes either AFAIK. But you'd think project like WINE would want to emulate it since they have the option to pick Windows version and this behavior is documented in the official API documentation.

---

In other news, Google Chrome will check Windows 11 eligibility on your PC for Windows 10 EOL.

UCyborg wrote on 2025-07-08, 08:14:

It may just glitch somewhere else without nGlide, the issue being with using memory from default process heap (where other libraries may allocate from), which it tends to corrupt. Yeah, there always had to be some compromises with hardware renderers back then. But oh my, 640x480 window isn't very immersive.

It's not specifically I76, it's literally anything I try to run the nGlide wrapper on just ends up segfaulting with some virtual memory overflow nonsense that never happened several Wine releases ago.

They've been overhauling a lot of stuff and generating significant regressions as a result, but I'm unable to report on any of it because their bug tracker is cumbersome and problematic (oh the irony!) and I do not have a spare system on hand to do their mandatory grunt work they expect everyone reporting bugs to do.