This is part of The Pile, a partial archive of some open source mailing lists and newsgroups.
Subject: Re: Opinions From: Bernhard Rosenkraenzer <bero@redhat.de> Date: Tue, 11 Jan 2000 21:29:25 +0100 (CET) On Tue, 11 Jan 2000, Jeremy Bradley wrote: > I would like to hear opinions of why users of this list are so happy > with Redhat compared to Microsoft. - less bugs - more stable, FAR fewer crashes (my Red Hat Linux boxes haven't crashed for several years - the only reasons I've ever had to reboot them were upgrades) - speed - security - source code availability - free (source and $0) - customizability - compliance with standards (RFCs, for example) - usability (I know it probably needs getting used to, but once you're familiar with the OS, its interfaces are FAR more productive than "all-you-can-do-is-click" OS) - hardware requirements - Linux still works well on a 486 - cleaner design - the chance to change something about the OS - send a bug report or a feature request to Microsoft and it goes straight to the next waste basket. - better support (if there's a real problem, you can usually get a response in a matter of hours, not years - and who at Microsoft would bother to reply to a message like this one?) - not part of some evil wannabe-emperor to control the world - probably 1000 more things I've forgotten to mention. > I found it to be a little tricky configuring things such as the > internet connection and such. Where exactly were you having problems? We're always glad to get feedback from new users so we can improve things. For someone used to the Unix ways of doing things, it's sometimes hard to imagine what a beginner doesn't know about. ;) > after being on the lists I see that users are constantly adding patches and > recompiling kernels and such. Is this type the type of stuff that has to be > done, or is this mostly just the little hacker and developer type things? Both actually. A couple of security problems have been found after the release, so if you're on a server you'll actually want to update these packages. The reason you're seeing so many packages is that we're not releasing one huge Microsoftish fix-them-all-and-introduce-1000-new-bugs-and-don't-let-the- user-see-what's-going-on fixpack, but giving the user control of what he wants to update. Many other things, like recompiling kernels, or moving to gcc 2.95.2 to get better optimizations, are hacker and development type things - stuff to get the most out of your system, but that isn't really necessary. > and have found that it has bugs too... Aside from maybe some "Hello World" program, you won't find any program that doesn't have any bugs. There are definitely some bugs left in Linux, they're just not as bad as those found in some other systems that tend to crash as soon as one application misbehaves, or as soon as it's been running for a week or so. > I spent all weekend trying to get Linux to find my printer port, Did you read the "known bugs" thing? http://www.redhat.com/support/docs/gotchas/6.1/gotchas-6.1.html > One other thing I don't like, > is that when I try to browse help files and stuff, most of the data that I > find is for earlier versions, it seems like the help files aren't > maintained. If something hasn't changed, there's no need to update the help files. If version 2.0 of package xyz just fixes bugs and optimizes some stuff, without changing its interface at all, there's no real need to update help files. > You guys start talking about > configuring modules and compiling kernels, it makes my head hurt and makes > me want to run back to my NT server! It's stuff you don't really need to do. One of the nice things about Linux is that you get the kernel source and can adapt everything to whatever you need. Microsoft would just laugh at you if you told them "hey, I've noticed you're including drivers for 250 network cards I don't want to use, and they're wasting my disk space! Give me your kernel so I can throw them out." or "I've upgraded my machine with the new AMD Pentium 6x86MX processor, and I'd like to tune the kernel for optimal performance on that CPU, please let me recompile it!". Many people make the mistake of assuming that Linux is very difficult to handle just because you *can* do these things. Similarily, don't be scared when someone shows you a shell script that makes heavy use of regular expressions and such - this is all stuff you don't need to know about (but it can make your life easier if you do). The command line in Linux is a very powerful tool - stuff like find . -name "*.c" |xargs perl -pi -e "s/^C:\\/\//g" looks very complicated - but how would you tell your Windows server to "find all c files in the current subdirectory, and automatically replace all occurences of 'C:\' at the beginning of a line with '/'"? > I don't think I could compile a kernel to save my life! Compiling a kernel is actually quite easy if you know your hardware, because there's a frontend that does the work for you. Just try it: cd /usr/src/linux make xconfig dep clean zImage modules modules_install LLaP bero ===