Quote: Original post by Arild Fines
Unfortunately, there seems to be some kind of emperor's new clothes effect(combined with a healthy dose of groupthink) at play in the unix community. People generally don't point out the glaring inadequacies in the tools; instead concluding that they themselves are at fault for not mastering them(the smug feeling of superiority that comes after memorizing a particularly cryptic command set for a tool probably helps too).

This is very true.

However, google doesn't have many inadequacies. ;)

Well some people feel that info is a good replacement for man. Subversion is a replacement for cvs. Not everyone thinks that good enough is good enough, but the same can be said for Windows users. Rebooting every time an application is installed or when a program crashes is good enough for a lot of people, but not everyone agrees.


But still, if there is a perceived lack, then perhaps this is something that needs to be addressed. It is like using Ctrl-C to copy and paste in Windows, but not having that same functionality consistent across applications on Linux. Copy and paste is different (highlight and paste with middle click), but people still want Ctrl-C. I know that has sparked a big debate.

So perhaps it is the same with help files/man pages. Some people would prefer them to change and be more integrated, others want them to stay they same.

I don't see a problem with a mixture, but is there that much of a need for development of an integrated documentation system over and beyond what info provides?

Oh god.

First off, forget BSD. We could go on and on about it, but it's simply not relevant.

Now, Linux and gaming. Well, for one thing, there's just the flat out gap in market share, particularly home desktop market share. Linux barely has any. Any professional game dev that wants to achieve anything needs to develop for Windows in order to survive. I don't believe that anything other than the sheer quantity of potential buyers is directly relevant. So the question becomes two parts:
1) Why does Linux have such a small desktop market share?
2) Are games critical to increasing Linux's desktop viability?

I won't go into detail about #1 at the moment, as it's a long discussion and you guys are free to take it up without me. But for #2, I'm inclined to say no. If more PCs transistion to a dual boot mentality, where Linux is good for everything except gaming, then slowly the market share will become such that games will appear for Linux. So my conclusion is that the lack of games for Linux isn't really related to gaming at all.

The only reason I still have Windows is games.. and the fact that I have an ATI card which has %¤/&# Linux drivers (can't even play native Linux games, thanks ATI!)

depends. what does not work? with a bit of tweaking you should get it working with not much ado.

Quote: Original post by RPTD
depends. what does not work? with a bit of tweaking you should get it working with not much ado.

Did you direct that post at me?

Well, the problem isn't that the ATI drivers aren't working (they work). The problem is that they are horribly slow and unstable. I can't even run UT2004 on Linux at 800x600 resolution without stuttering/lag. On Windows the game runs smoothly at 1024x768. I don't think I should even think about trying to run Doom 3 with an ATI card in Linux..

I should've bought an nVIDIA :(

Quote: Original post by tentoid
Did you direct that post at me?

Well, the problem isn't that the ATI drivers aren't working (they work). The problem is that they are horribly slow and unstable. I can't even run UT2004 on Linux at 800x600 resolution without stuttering/lag. On Windows the game runs smoothly at 1024x768. I don't think I should even think about trying to run Doom 3 with an ATI card in Linux..

I should've bought an nVIDIA :(

this is rather strange as i am running ATI on a 1024x800. UT2004 worked well and even Doom3 worked nearly as quick as under windows.

what kernel have you in use? have you compiled the ATI kernel module or used a predefined one? what does 'glxinfo | grep direct' say?

Quote: Original post by C-Junkie
I call BS. I have never one had to go to the source to figure something out. Documentation abounds.

Call BS all you want. Not all open source projects are well-documented (I'm not referring to just the mainstream ones), and you can't claim they are without being an expert on all of them. Even if you're working with a mature tool, you often need a plugin or a complementary product that is not mature.

Quote: Closed source drivers are buggy drivers. They are pissed upon by the majority of the kernel people for technical reasons, not religious ones, though there are some that do for both.

I call BS to this. We found and had to fix a few bugs in the open source Linux drivers for the 3Com 3C905x cards. While it was very nice that we were able to fix them in our shipping product without waiting for someone to provide a fix, the same bugs did not show up in 3Com's Windows driver. Not all closed source drivers are crappy or buggy, and not all open source drivers are good.

As a side note, ATI and nVidia will not open-source their drivers for a long time, if ever. Much of the value in their product comes in the driver optimizations they make. Driver optimizations alone can give them an edge over a competitor, and they're not going to release something like that to the public. The Linux license should not force them to create a custom build of the kernel to get around the GPL to protect their trade secrets.

Quote: Linux CREATED the open source community. The BSDs just sat there. They were cast by the wayside for very good reasons. The biggest being: they weren't doing anything.

Once again I call BS. The open source community was alive and well long before Linux was started. Linus started working on Linux in 1991. The GNU project started in 1984. Without the GNU suite of products already in place, Linus's fledgeling OS would have taken a lot longer to get off the ground. So it was really the other way around. The open source community created Linux.

Quote:
Let your friend know that BSD IS dead. Except for Apple.

How many times can I call BS in one post? I believe that BSD is being installed and used in more systems than Linux today. Most of that is in embedded systems like mobile phones, but that doesn't make it any less true. In the business world, BSD is generally considered to be faster, more stable, more secure, and more free (i.e. less restrictive licensing).

All Linux has really has over it is hype, which gets it more driver support. More driver support is definitely an advantage, but it's not a trend that can't be reversed. I think we'd be better off if it did reverse itself because then we wouldn't have to recompile the kernel to install the nVidia drivers.

you guys are missing the whole bit on sound support.. I'd prolly use linux alot more if it wouldn't choke on the sound cards I have. Until the people at ALSA(or whatever sound drivers linux uses in the future) get it together I'll be sticking with my FreeBSD.

Quote: Original post by GBGames
Well some people feel that info is a good replacement for man.

It isn't.

If one tool is consistently harder to learn to use across all users (and I dare you to suggest that learning to use man or info is easier than using a web browser), then the tool is not well-suited to general-purpose intermittent use. What that says is that a more difficult-to-learn interface may be appropriate because of the efficiency benefits it confers on frequent users, but for systems/applications that are used only once in a while or do not have a "mastery level" it is clearly an inferior choice.

Quote: It is like using Ctrl-C to copy and paste in Windows, but not having that same functionality consistent across applications on Linux. Copy and paste is different (highlight and paste with middle click), but people still want Ctrl-C. I know that has sparked a big debate.

There shouldn't be any debate. If the tools/applications/systems are targeted at users who have existing modes of use, then the tools/applications/systems need to adopt those modes, at least as one option.

Quote: I don't see a problem with a mixture, but is there that much of a need for development of an integrated documentation system over and beyond what info provides?

An "integrated documentation system" in the windows world is basically HTML files and a contents index. Displaying that same data in the man or info shells should be trivial. Doing the inverse - displaying infopages in a web browser - isn't quite as easy.

Pragmatism suggests that the solution is a no-brainer.