Jump to content

iphitus

Global Moderator
  • Posts

    3831
  • Joined

  • Last visited

Everything posted by iphitus

  1. which could just be a misconfiguration on your end. Access points dont do "os detection" at an association level, and 802.11 association is a standard well implemented under linux. James
  2. Welcome to the board draker! Mandriva ought to include madwifi, but if it doesnt, you can install it with dkms. I don't know the exact command, but im sure someone will come here soon and show you. There's no need for ndiswrapper here, as madwifi is an excellent and stable driver, so hopefully nobody suggests that. James
  3. The issue here is that mandriva comes with newer versions of a range of libraries, including GLIBC. Maple is linked against older versions, intended to be installed on distros with those versions. It's probably got a list of "supported" distros, like RHEL, SuSE, and others. Maple's primary audience is more likely to use these stable "production" distros. Symlinking will not work (except in the most simple of cases, this does not include glibc). Do not install an old glibc over your current one, it'll effectively break your system. If you want it working on mandriva, either: - find a statically compiled version of maple or one that includes the libs - get a newer version, if one exists. - install compatibility libs. thats what the novell solution details to an extent. Before you do any of that, remove the symlinks, they'll just get in the way. James
  4. Security isn't that simple, otherwise all the major projects would never be releasing patches and security updates. As for "lightweight" -- a web browser (firefox anyone?), web server, processes for dynamic pages (php/python/ruby/perl/etc)... isnt lightweight. It's overkill. Feature Creep. Nor can i think of any usage situation where this thing would actually be useful. It'd be slower, have a much weaker UI, and be far less capable. It'd be far more complex, heavier on the system, and putting something in a browser that doesnt need to be, gaining nothing extra by being there. James
  5. I have this one myself, a great introduction to the kernel and how it operates. From this you could start messing around, and then delve into something a bit more complicated. http://www.amazon.com/Linux-Kernel-Develop...s/dp/0672327201 After that, there's the in kernel Documentation/, linux device drivers (3rd Ed) - available free online too, and probably other books I don't know. Careful buying books, that you get the latest edition, otherwise you'll end up with a useless book about Kernel 2.4. Things have changed a *lot* since then, so don't even consider old editions. James
  6. and adds a pile of security issues. James
  7. blasphemer! the man was aru! we await a second coming! James
  8. I was being a tad sarcastic :)
  9. thecompletecomputerresource or thelongesturliveseenforalongtime.com? :P Follow the gnome hig. Don't follow everything exactly if you don't want, but it's got plenty of smart suggestions on how to construct UI's and keep them consistent. http://developer.gnome.org/projects/gup/hig/ what sticks out to me in the screenies: * no spacing whatsoever. cramped! * Icons go on buttons, not menu items. What's that find thing on the menu? Make it a button and move it to the toolbar. * Usually toolbar buttons have larger icons * Use GTK stock icons only. * Buttons on tabs?? And the other thing. it's fun to plan the UI, but it's even better to write the code. Before you've even designed the UI you know how the program's going to work and what many of the routines will be. I've seen lots of projects dead in the water, and they've only ever designed the UI. James
  10. There's probably a libstdc++ compat rpm in urpmi, there shouldnt be a need to download third party and other distro rpms. James
  11. The post triggered deja'vu, so i searched up to find some articles, so while the inquirer may not be the best, i do remember it being reported elsewhere. smh is a reliable resource, the better newspaper here :) James
  12. deja (2003) http://www.smh.com.au/articles/2003/07/08/1057430181568.html vu (2005) http://www.theinquirer.net/default.aspx?article=22967 and yes, they did actually sell machines: http://linuxlala.net/thoughts/?p=21 James
  13. linus' argument lacks logic and is ridiculously short sighted. Con always made an effort to chase up bugs, even the most obscure bugs. I admired his patience. The one "bug" that Linus saw, Con did chase up, and the scheduler was not at fault. The user who was reporting it refused to accept this, and it caused a bit of a stink on the mailing list. Linus' argument is based on this one single report. If he'd bothered to look on the CK mailing list, or at any other work of Con's, he'd have seen diligent and careful bug handling. Additionally, Con stopped developing *after* it was clear that SD had no future in mainstream. So saying that he is not capable of being a long term maintainer is based on something that developed *after* Con's SD was rejected, and could not have been part of Linus' original decision. Regardless, would Con bother maintaining SD if it was going to be seen by most as "second place". Con is more than capable of long term maintenance, as shown by his continual improvement and maintenance of his staircase scheduler, and swap prefetch for many years. This saga and many others, have shown some pretty nasty flaws in how the kernel is developed. Ugh. James
  14. Give him credit. His decision isn't just directed at Ingo, it's been something long in the making, before this whole spat. The current development style has some big problems, and they get in the way of things getting done. Con does exaggerate slightly, but mostly, he's correct. James
  15. Use "mv" to move it to OMGAUDIO. What filesystem is it on?
  16. Not a good idea to put CVS variables in bash_profile, as if you ever need to use another CVS directory, it could cause some problems.
  17. It's a firefox extention, thus it's cross platform. Or at least, I know this one is. There's also a tonne of metalink command line clients. No need to be so cold ianw :)
  18. by 3D desktop do you mean XGL? XGL is inherently broken. Don't use it. James
  19. for reference, googling found me this: http://mambo.kuhp.kyoto-u.ac.jp/~takushi/#canon an apt repo which has the drivers packaged up and configured correctly for debian. James
  20. slow down. don't try to do everything at once. first you're better off learning how an SCM works, before you go on complicating things and trying to share it. Most people are gradually moving away from CVS to other, and simpler alternatives. Subversion (svn) might well be better suited for you as you're looking at a centralised setup. A quick google finds these, which will give you a good introduction to scms and svn in particular: http://www.phptr.com/articles/article.asp?p=408888&rl=1 http://abbeyworkshop.com/howto/misc/svn01/ James
  21. What's the kernel sound module? "lsmod|grep snd" should show this. Getting apps that interact with the sound card like that can be black magic under alsa. If it doesnt work with aoss, like some games, things get difficult. I find the easiest approach is to go the opposite way. Use OSS. http://www.opensound.com/ That *was* a closed source implementation of OSS, but it was recently open sourced. The open source implementation OSS that exists in the kernel currently is deprecated, but OSS as a general sound system is not. It works great for me, and does a much better job of implementing mixing than alsa. It has an alsa compatibility, but most applications still have OSS support anyway. Install that (it has an rpm), enable it, set apps to OSS, and things should work a bit easier. James
  22. It's been going that way for years, but it's long been clear that it doesnt work. Macrovision and CSS (DVD), HDCP and AACS (bluray/hddvd). They've each been broken and proven flawed. DRM is a dead end. Each iteration will be broken, with the downside that it will somewhat inconvenience consumers for a short while. James
  23. No, vista should not affect anything. You could have AmigaOS for all Linux cares, it wouldnt change a thing.
  24. but this isnt a good idea. as it is, it seems that people deal with dane elec and you can only hope that dane elec fix it. You've knocked people off your own support processes and onto another company's. In doing so, you should at least have some way of guaranteeing a reasonable level of support. Either route their queries through you, or try to hold dane-elec accountible in some way. Otherwise, why should dane-elec care about your customers? Mandriva will end up with the bad reputation, and dane-elec lose nothing. Mandriva really ought to start thinking about their customers once in a while... James
×
×
  • Create New...