James, I usually leave the political discussions alone, but this is near and dea= r to=20 me currently since I recently made the switch to Linux. I'll admit up front that I've not read that full article, since I feel ce= rtain=20 that it claims Linux has many bugs, and probably more than Win___. Even=20 though I've mostly moved to Linux (2 of 3 machines), I don't doubt that i= t=20 has its share of issues, and I've seen some already. But I'm sure there = are=20 many surveys that will claim compelling results in either direction. FWI= W, I=20 recently saw a MS ad f/Windows Server touting "... Linux reliability...",= so=20 even they seem convinced. I have personally had much more bug-free computing with Linux. 3 HD cras= hes=20 in the past 2 years -- 2 of which were SW related (Win2k and WinNT), and = one=20 which was a HD-controller failure. I've abused Linux (accidentally pulli= ng=20 power off), and it recovers nicely. Over months of my Win installation, = it=20 would slow down and some of my best Windows-consultant friends could not = fix=20 it. The answer is always to re-install from scratch. Of all the Win=20 versions I've tried, Win2k has been the most stable, and I could get abou= t 5=20 days off it before having to re-boot. Linux ... many weeks so far. And = I'm=20 on dual-boot here, so haven't really given it a chance to run un-rebooted= for=20 long. The problem with Windows bugs is that there is little or no motivation or= =20 attempt to fix it. There is no incentive for MS to do so, as the revenue= =20 comes from releasing a new product, not investing time/money into a produ= ct=20 which has already been sold. And it *is* all about the $$$. =20 Seventy-something percent bug-pass rate was considered user-deliverable=20 quality (on one of their compiler products). And I know this up-front as= I=20 used to work for MS. Of course that exists in other companies, but the=20 attitude at IBM (also worked there) for instance has been one of leaving = off=20 feature ideas so end-users have a reason to upgrade. With Linux, when a=20 concern arises, I know I have control over fixing/solving it. The virus situation is another issue. That's self-explanatory. Now, even though the topic here is bugs, the most important problem to me= is=20 the lack of security, and MS's own spy attempts. Other than reliability,= =20 trust has to be one of the biggest features of a piece of SW that forms t= he=20 base of all of a user's computing. If my OS mfgr is going to spy on me, = then=20 I walk. I never went past SP2 on Win2k due to spying concerns on SP3, an= d=20 I've heard its even worse on XP. So I've never touched it. But I get=20 regular calls from friends asking for help with fixing Win XP crashes/bug= s. The 27 posted fixes you mention -- does that mean that there were only 27= =20 reported bugs that were in MS'es database? Or that only 27 of the thousa= nds=20 were actually fixed. That metric needs to be clarified. But everything has it's place, and I agree with skill-set requirement you= =20 mention. For my non-techie friends (who have been passing on emails abou= t=20 the jdbmgr hoax this week, as if it were a major issue :-), I recommend W= in=20 products still. Just cause I would hate to have to explain which librari= es=20 they need during installation for example. Redhat helps, but is still no= t as=20 brain-dead as Win2k installation. For myself, I've learned enough about=20 Linux in the couple months I've been using it, to have better success tha= n=20 Win2k so far. I really only keep Win2k around for 2 apps ... MPLAB SIM, = and=20 my electonic-copy Audi service manual. And also any other SW that comes = my=20 way that is Win only. To each their own. Cheers, -Neil. On Wednesday 28 May 2003 22:37, James Newton, webmaster scribbled: > Please note, before you cry "...off with his head!", that I fully > support and hope for the best from all open source software. > Richard Stallman is probably my #1 hero in the computer world. > All hail GNU/Linix and so on... BUT, My experience has shown that > there is another side to it. Just about every Linux box I have > worked on has been hacked and I've never lost an NT server. I > have long wondered if the general opinions about Linux being more > stable / secure were wishfull thinking or if they are born out by > numbers. > > http://story.news.yahoo.com/news?tmpl=3Dstory&u=3D/nf/20030523/bs_nf/21= 583 > > A count of the problems reported for XP Professional is available > on the Microsoft Web page that lists all of its security > bulletins. Use the pull-down menu to find the bulletins for > Windows XP Professional. The list starts in November 2001. In the > 18 months since then, 27 bulletins about security flaws or other > bugs have been posted for Professional XP. > > To count the fixes and bugs for Red Hat Linux 7.2, go to the > company's errata page and begin counting from November 2001. From > November 2001 until now, the company has issued 158 security > bulletins or bug fixes (not counting the enhancements listed on > that page). > > Also, the Linux-Unix OS is largely in the server environment, > where the vast majority of Windows installations are in the > client environment, The difference in technical skills in those > two user bases could greatly influence perceptions of OS > stability. {i.e. Linux has smart people running it and Windows > has to do its best with idiots like me! } > > ...with Windows, there's a rather aggressive community trying to > find bugs to denigrate Microsoft and Windows. > > http://story.news.yahoo.com/news?tmpl=3Dstory&cid=3D1093&ncid=3D1209&e=3D= 1&u=3D/pcwor >ld/20030528/tc_pcworld/110906 > > For the second time in as many months, the Apache Software > Foundation released an updated version of the popular open-source > Web server software, only to warn users of a critical security > hole in previous versions of the software that the update > patches. > > Among those fixes is a patch for a security hole in the mod_dav > module that could be exploited remotely, causing an Apache Web > server process to crash, according to the bulletin. > > A second fix is for a denial-of-service vulnerability affecting > Apache's authentication module. By exploiting a bug in > configuration scripts used for password validation, attackers > could launch remote denial-of-service attacks that would cause > valid user names and passwords to be rejected, the bulletin said. > > James Newton (webmaster, former admin #3) > http://www.piclist.com > jamesnewton@piclist.com > 1-619-652-0593 VM > 1-208-279-8767 FAX -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.