Linux-Advocacy Digest #603, Volume #31 Sat, 20 Jan 01 06:13:06 EST
Contents:
Re: NTFS Limitations (Was: RE: Red hat becoming illegal?) (Edward Rosten)
Re: Linux is crude and inconsistant. (Peter =?ISO-8859-1?Q?K=F6hlmann?=)
Re: Windows 2000 Datacenter Server does support the "five nines" (.)
Re: 10,000 to 20,000 Linux/Alphas - CLUSTERED! (.)
Re: 10,000 to 20,000 Linux/Alphas - CLUSTERED! (.)
Re: A salutary lesson about open source (.)
Re: Linux 2.4 Major Advance (.)
Re: Red hat becoming illegal? (.)
Re: Red hat becoming illegal? (.)
Comparison of Linux/Apache versus Win2000 server uptime (Frank Kruchio)
Re: M$ *finally* admits it's OSs are failure prone ("Bobby D. Bryant")
Re: The Sins of William Gates ("Bobby D. Bryant")
Re: A salutary lesson about open source ("Bobby D. Bryant")
Re: A salutary lesson about open source ("Bobby D. Bryant")
Re: Comparison of Linux/Apache versus Win2000 server uptime ("Erik Funkenbusch")
Re: A salutary lesson about open source ("Bobby D. Bryant")
Re: M$ *finally* admits it's OSs are failure prone ("Erik Funkenbusch")
Re: 10,000 to 20,000 Linux/Alphas - CLUSTERED! ("Erik Funkenbusch")
Re: "The Linux Desktop", by T. Max Devlin ("David Brown")
Re: "The Linux Desktop", by T. Max Devlin ("David Brown")
----------------------------------------------------------------------------
From: Edward Rosten <[EMAIL PROTECTED]>
Crossposted-To:
alt.destroy.microsoft,comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: NTFS Limitations (Was: RE: Red hat becoming illegal?)
Date: Sat, 20 Jan 2001 08:48:15 +0000
Chad Myers wrote:
>
> <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
> > On Sat, 20 Jan 2001 04:25:52 GMT, T. Max Devlin <[EMAIL PROTECTED]> wrote:
> > >Said [EMAIL PROTECTED] () in alt.destroy.microsoft on Sat, 20 Jan
> > >>On Fri, 19 Jan 2001 15:25:22 GMT, T. Max Devlin <[EMAIL PROTECTED]> wrote:
> > >>>Said Ayende Rahien in alt.destroy.microsoft on Fri, 19 Jan 2001 06:58:01
> > > [...]
> > >>>>I'm not sure exactly *what* you can put into a file to get into that size.
> > >>>
> > >>>Precisely what they said about the 2 Gigabyte limit. ;-)
> > >>
> > >> Databases.
> > >
> > >A 'database' is not, by definition or even by convention, a single file.
> >
> > There's that magic word: "convention".
> >
> > That's all that separates a table spread across 10 files
> > and 5 physical disks from the video that for some
> > strange reason can't be similarly divided.
>
> Man, you must really have you head up your ass.
>
> I never said it couldn't, I merely said it would be incredibly
> wasteful and stupid to do such a thing.
>
> Like I said, (at least four times now) it would double or
> tripple the time / video for processing.
>
> All this, because of brain-dead Linux. No thank you.
>
> >
> > >
> > >> Then again, databases grew to that size long before there
> > >> were file systems to handle such file sizes. Good software
> > >> adapts to some degree to route around other 'faults' in the
> > >> system.
> > >
> > >Databases started out larger than a single file. The contrary idea
> > >didn't even occur to anyone, I would wager, until the advent of PC
> > >desktop applications.
> >
> > That's quite a long time actually.
> >
> > Oracle is barely older than that.
> >
> > >
> > >>>And they were really sure *they* were right, too. ;-)
> > >>[deletia]
> > >>
> > >> The real question is how much trouble is it to "route around"
> > >> such limitations. Considering the successes of databases in
> > >> this regard as well as mp3 players and DVD consoles, I don't
> > >> think this issue is such a tragedy.
> > >>
> > >> Compared to some of Microsoft's past mistakes, a 2G limitation
> > >> in an ext2 file is downright trivial.
> > >
> > >The real issue is how trivially correctable it is. There are already
> >
> > Just buy an Alpha. '-)
> >
> > Besides, it has unmatched FPU performance.
>
> Sounds like the kind of answer you hate MS for.
>
> Oh yeah, and Alphas have unmatched prices.
>
> Why not buy a PC and buy Win2K, it's far cheaper than Linux + Alpha,
> plus you get so many more features with Win2K.
No you don't: you get an inferior computer with featureless operating
system, if you go with the latter option.
-Ed
--
Did you know that the reason that windows steam up in cold|Edward Rosten
weather is because of all the fish in the atmosphere? |u98ejr
- The Hackenthorpe Book of lies |@
|eng.ox.ac.uk
------------------------------
From: Peter =?ISO-8859-1?Q?K=F6hlmann?= <[EMAIL PROTECTED]>
Crossposted-To: alt.linux.sux
Subject: Re: Linux is crude and inconsistant.
Date: Sat, 20 Jan 2001 00:41:16 +0100
T. Max Devlin wrote:
>
> Well, since it isn't the fact that cards are hard to get which indicates
> that Microchannel was non-standard. It is the fact that only IBM ever
> produced PCs supporting it which most decisively supports the argument.
> And absolutely nothing, btw, which refutes it; Microchannel was
> non-standard. ISA, and possibly VLB(?), was.
>
Well, as much as I agree with you on the whole, here you are wrong.
I know for a fact (because I worked for that company 14 years) that
Honeywell / Bull produced Microchannel-machines AND boards.
And they were not the only ones. IBM was NOT alone with MC, although it
never was any good. The advantages were not good enough in the light of the
diasadvantages compared to ISA (VLB / EISA). PCI incorporated many of the
good things of MC.
------------------------------
From: [EMAIL PROTECTED] (.)
Subject: Re: Windows 2000 Datacenter Server does support the "five nines"
Date: 20 Jan 2001 09:22:13 GMT
Adam Warner <[EMAIL PROTECTED]> wrote:
> Hi Craig,
>> I'm sure it uses clustering to do this (no big deal).
> It does. Does this mean so long as at least one machine is running in the
> cluster it doesn't count again 99.999% availability if one machine has to be
> rebooted, etc?
> So as an extreme example, if you rebooted a different server in the cluster
> each day you could still quote 99.999% uptime?
You sure can. Which means all you'd need is 1,000,000 windows 2000 datacenter
machines in one cluster to meet the uptime of an asci white.
Or 600 of them to meet the uptime of a two-node beouwulf cluster.
=====.
------------------------------
From: [EMAIL PROTECTED] (.)
Subject: Re: 10,000 to 20,000 Linux/Alphas - CLUSTERED!
Date: 20 Jan 2001 09:24:50 GMT
sfcybear <[EMAIL PROTECTED]> wrote:
> WOW!
> http://computerworld.com/cwi/story/0%2C1199%2CNAV47_STO56666_NLTpm%2C00.html
Hmmm. Linux scales from my Tivo to a 100 trillion OPS supercomputer.
No matter how you cut it, windows 2000 does not have anywhere near a comparable
scale. This is a fact.
What exactly is it that windows does again?
=====.
------------------------------
From: [EMAIL PROTECTED] (.)
Subject: Re: 10,000 to 20,000 Linux/Alphas - CLUSTERED!
Date: 20 Jan 2001 09:26:52 GMT
Adam Warner <[EMAIL PROTECTED]> wrote:
> Hi "sfcybear",
>>
> http://computerworld.com/cwi/story/0%2C1199%2CNAV47_STO56666_NLTpm%2C00.html
Also, apparantly linux is able to scale to 20,000 processors.
Compared to windows 2000 datacenter's alleged 32.
Thats a pretty big difference.
=====.
------------------------------
From: [EMAIL PROTECTED] (.)
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: A salutary lesson about open source
Date: 20 Jan 2001 09:33:38 GMT
In comp.os.linux.advocacy Chad Myers <[EMAIL PROTECTED]> wrote:
> "Cliff Wagner" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
>> On Fri, 19 Jan 2001 13:46:37 GMT, Chad Myers typed something like:
>> >
>> >"J Sloan" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
>> >> Chad Myers wrote:
>> >>
>> >> > Oh you mean the heavily inflated web server thing? The grossly
> unscientific
>> >> > misrepresentative web server thing? Where every virtual host is counted
>> >> > as a sever thus doubling or trippling the server numbers?
>> >>
>> >> Websites are websites, and should be counted as such.
>> >
>> >Right. 500 "My Cat Fluffy" websites vs 500 e-Commerce Fortune 500
>> >company web sites means the same thing.
>>
>> Please provide proof of this statement.
>> From my experience, most "My Cat Fluffy" sites are hosted
>> on places like geocities and homestead and places
>> like that because people generally don't want to
>> pay money to host something so inane.
> If you compare surveys from other parties (besides Netcraft), they
> mostly survey Fortune500, Global500, etc. Those numbers, IIS is
> in the lead or closely follows iPlanet and Apache is far behind.
Ah. So now its not most of the websites in the WORLD, its just the
bloated commercial ones; not the "my cat fluffy ones". Wait, and
its also not just all the bloated commercial ones in the world, its
only the ones listed in Fortune 500 and Global 500. (that would
be 1000 sites, chad).
But wait, its not even all of those, its simply the MAJORITY of them.
So less than 1000 sites fit your bill. Out of all the sites in the
world.
Thats very interesting.
=====.
------------------------------
From: [EMAIL PROTECTED] (.)
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Linux 2.4 Major Advance
Date: 20 Jan 2001 09:35:28 GMT
In comp.os.linux.advocacy Chad Myers <[EMAIL PROTECTED]> wrote:
> "T. Max Devlin" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
>> Said Chad Myers in comp.os.linux.advocacy on Tue, 09 Jan 2001 13:59:29
>> [...]
>> >The filesystem doesn't "get in the way" and it's never been an issue. Even
>> >NT 4 still kicks Linux's ass in all things performance.
>>
>> BIG lie.
> Not really. The only benchmark I've seen Linux win was with a web server
> that no one uses. One benchmark. Please show me ones where Linux wins
> (oh yeah, and the FUD ones from c't don't count, only major reputible
> companies with standardized benchmarks, not grudges against Microsoft).
Thats right. Windows will run on anything its ported to, and only benchmarks
that chad likes count.
=====.
------------------------------
From: [EMAIL PROTECTED] (.)
Crossposted-To:
alt.destroy.microsoft,comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: Red hat becoming illegal?
Date: 20 Jan 2001 09:37:03 GMT
In comp.os.linux.advocacy Chad Myers <[EMAIL PROTECTED]> wrote:
> "Steve Mading" <[EMAIL PROTECTED]> wrote in message
> news:94a0ud$lqs$[EMAIL PROTECTED]...
>> In comp.os.linux.advocacy Chad Myers <[EMAIL PROTECTED]> wrote:
>>
>> : - Import the video from firewire (usually 3:1 or 5:1 with good capture
>> : cards)
>> : - Load the video into Premiere or whatever app they're using for editing
>> : - Save raw video file for posterity.
>> : - Perform edits, insert audio, stills, etc
>> : - Save edits to video file
>> : - Resize video to internet video size (192x144)
>>
>> Alarm bells went off when I read this. How long is this video that
>> it takes 2 GB at 192x144 size?? Does the video last all day?
> Well, thank you for clipping the part that answered your own question.
> By the time its resized, it's rarely 2GB. However, we did have some
> videos that, when resized to 192x144, were still over 2GB (they were
> 30-45 minutes in length). This was before heavy amounts of
> compression.
AAAAAAAAAAAAHAHAHAHAHHAA.
Haha.
Either you are lying (very likely), or you utterly and quite completely
have no idea what youre doing at all. (very likely)
=====.
------------------------------
From: [EMAIL PROTECTED] (.)
Crossposted-To:
alt.destroy.microsoft,comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: Red hat becoming illegal?
Date: 20 Jan 2001 09:37:51 GMT
In comp.os.linux.advocacy Chad Myers <[EMAIL PROTECTED]> wrote:
> "T. Max Devlin" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
>> Said Steve Mading in alt.destroy.microsoft on 19 Jan 2001 18:27:25 GMT;
>> >In comp.os.linux.advocacy Chad Myers <[EMAIL PROTECTED]> wrote:
>> >
>> >: - Import the video from firewire (usually 3:1 or 5:1 with good capture
>> >: cards)
>> >: - Load the video into Premiere or whatever app they're using for editing
>> >: - Save raw video file for posterity.
>> >: - Perform edits, insert audio, stills, etc
>> >: - Save edits to video file
>> >: - Resize video to internet video size (192x144)
>> >
>> >Alarm bells went off when I read this. How long is this video that
>> >it takes 2 GB at 192x144 size?? Does the video last all day?
>>
>> Its most probably like this:
>>
>> The production company gets a post-production video tape. Their task is
>> to produce a number of short, small 'clips' that will be presented as
>> 'web video' on some web site. (Probably ASF format. Guffaw.) Reduced
>> size and resolution, etc, would result, but the input data is still in a
>> >2Gig file.
>>
> That's wrong.
> We shot the videos ourselves. Hours and hours of video. They broke the
> videos into shorter segments (they were college educational lectures)
> which were typically 15minutes to 60 minutes in length depending on
> the subject matter. We had to digitize the videos to the computer
> and archive the raw video for prosterity in case we had to re-edit
> the video due to errors or style change.
Chad edits for a streaming porn site, apparantly.
Way to go chad. That makes you an expert.
=====.
------------------------------
From: Frank Kruchio <[EMAIL PROTECTED]>
Subject: Comparison of Linux/Apache versus Win2000 server uptime
Reply-To: [EMAIL PROTECTED]
Date: Sat, 20 Jan 2001 22:52:59 +1300
http://uptime.netcraft.com
www.suse.com Suse Linux with Apache server, latest 90 day moving average is
191.45 days
www.microsoft.com Win2000 server, latest 90 day moving average
is 15.93 days
Does this count in favour of Linux to be reliable as a server ?
What do YOU think ?
--
Digitally signed message(2048 bit)
GPG key/Home Page: http://homepages.paradise.net.nz/frankkru Wellington, New Zealand
------------------------------
From: "Bobby D. Bryant" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.nt.advocacy
Subject: Re: M$ *finally* admits it's OSs are failure prone
Date: Sat, 20 Jan 2001 03:52:54 -0600
Erik Funkenbusch wrote:
> semi-random? As in "I posted the sites that met my criteria". You can have
> 100 sites with 10 year uptimes, but if you have 10,000 sites with 1 day
> utimes you're average is going to be quite low. Since you chose to ONLY
> include high uptimes in what you posted, that is not a valid statistic.
You obviously didn't read the part that explained my methodology. You may be
looking at the wrong link; look at the one where I actually calculated the
averages.
Bobby Bryant
Austin, Texas
------------------------------
From: "Bobby D. Bryant" <[EMAIL PROTECTED]>
Subject: Re: The Sins of William Gates
Date: Sat, 20 Jan 2001 03:49:49 -0600
Houston Review wrote:
> The Sins of William Gates:
>
> A Society Avenges its Embarassment.
>
> By Derek Copold
I skimmed it, but couldn't be bothered to actually read it, because the
skim was enough to reveal that Derek is telling people what *my* motives
for despising BG are, and that he's entirely incorrect.
The article is very biased, and could hardly be more apropos for a
MS-sponsored hack job. Who's footing the "bill", Derek?
Bobby Bryant
Austin, Texas
------------------------------
From: "Bobby D. Bryant" <[EMAIL PROTECTED]>
Subject: Re: A salutary lesson about open source
Date: Sat, 20 Jan 2001 03:58:53 -0600
Mig wrote:
> More intersting would be to look at volume served on ordinary websites (no
> downloads) with both static and dynamic content. There must be a survey
> somewhere.
I'm not sure, but I think the Hot 100 are rated "hot" because of the amount of
traffic they carry.
Bobby Bryant
Austin, Texas
------------------------------
From: "Bobby D. Bryant" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: A salutary lesson about open source
Date: Sat, 20 Jan 2001 03:57:58 -0600
Ayende Rahien wrote:
> I did just that some time ago.
> The result was Apache first, IIS second, and various others third.
> Unix first, NT/2K second, linux third.
Could you repost them, or direct us to the link in deja?
Thanks,
Bobby Bryant
Austin, Texas
------------------------------
From: "Erik Funkenbusch" <[EMAIL PROTECTED]>
Subject: Re: Comparison of Linux/Apache versus Win2000 server uptime
Date: Sat, 20 Jan 2001 04:20:03 -0600
"Frank Kruchio" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> http://uptime.netcraft.com
>
> www.suse.com Suse Linux with Apache server, latest 90 day moving
average is 191.45 days
>
> www.microsoft.com Win2000 server, latest 90 day
moving average is 15.93 days
http://uptime.netcraft.com/up/graph?site=redir.windowsmedia.com
latest avage 201 days.
http://uptime.netcraft.com/up/graph?site=insider.microsoft.com
Latest average 196 days
http://uptime.netcraft.com/up/graph?site=beta.visualstudio.net
Latest average 129 days
BTW, this last entry shows a problem with the netcraft statistics. They
only list 41 samples for the platform, and 57 samples for the average, which
seems to indicate that 16 samples got lost somewhere and were recorded as
0's, bringing the average down by almost 30 days than it should be.
> Does this count in favour of Linux to be reliable as a server ?
No.
------------------------------
From: "Bobby D. Bryant" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: A salutary lesson about open source
Date: Sat, 20 Jan 2001 04:05:02 -0600
Cliff Wagner wrote:
> Look at the "Most Requested"
> sites on netcraft. Microsoft has 9 spots on the top...50.
> And that's with hotmail on there twice.
FYI, those "Most Requested" are just the sites that are looked up on Netcraft
the most, not the sites with the most traffic. (Most of those Hotmail hits
are probably Slashdotters checking for down time.)
Bobby Bryant
Austin, Texas
------------------------------
From: "Erik Funkenbusch" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.nt.advocacy
Subject: Re: M$ *finally* admits it's OSs are failure prone
Date: Sat, 20 Jan 2001 04:24:09 -0600
"Bobby D. Bryant" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Erik Funkenbusch wrote:
>
> > semi-random? As in "I posted the sites that met my criteria". You can
have
> > 100 sites with 10 year uptimes, but if you have 10,000 sites with 1 day
> > utimes you're average is going to be quite low. Since you chose to ONLY
> > include high uptimes in what you posted, that is not a valid statistic.
>
> You obviously didn't read the part that explained my methodology. You may
be
> looking at the wrong link; look at the one where I actually calculated the
> averages.
Which message is that? According to Deja, you've only posted 5 messages
since July with the word "average" in it, and none of them are it.
------------------------------
From: "Erik Funkenbusch" <[EMAIL PROTECTED]>
Subject: Re: 10,000 to 20,000 Linux/Alphas - CLUSTERED!
Date: Sat, 20 Jan 2001 04:29:26 -0600
"." <[EMAIL PROTECTED]> wrote in message
news:94blks$5ov$[EMAIL PROTECTED]...
> Adam Warner <[EMAIL PROTECTED]> wrote:
> > Hi "sfcybear",
>
> >>
> >
http://computerworld.com/cwi/story/0%2C1199%2CNAV47_STO56666_NLTpm%2C00.html
>
> Also, apparantly linux is able to scale to 20,000 processors.
>
> Compared to windows 2000 datacenter's alleged 32.
>
> Thats a pretty big difference.
Man, this isn't even vapor.. it hasn't even been *STARTED*. They claim to
not have the machine ready till 2004. Lots of things will change in both
the Linux and Windows side before then.
------------------------------
From: "David Brown" <[EMAIL PROTECTED]>
Crossposted-To: alt.destroy.microsoft
Subject: Re: "The Linux Desktop", by T. Max Devlin
Date: Sat, 20 Jan 2001 11:39:48 +0100
Damien wrote in message <3a693a3d$0$[EMAIL PROTECTED]>...
>On Sat, 20 Jan 2001 00:32:26 GMT, T. Max Devlin <[EMAIL PROTECTED]> wrote:
>> Said Damien in alt.destroy.microsoft on 19 Jan 2001 23:41:17 GMT;
>> >On Fri, 19 Jan 2001 21:21:30 GMT, T. Max Devlin <[EMAIL PROTECTED]>
>> >wrote:
>> >
>> >[*nice specs*]
>> >
>> >> It should be here next week. I didn't get the dual-boot option, but I
>> >> plan to install 95, and maybe NT, once its up and running. So here we
>> >
>> >You might run into some problems getting this machine to dual boot.
>> >The default Redhat Workstation install (which the OEM probably used)
>> >will likely have only two partitions (root and swap) which doesn't
>> >leave you anywhere to put Windows. Partition magic will get you over
>> >that hurdle, for a price.
>>
>> I'm fine with fdisk, actually. Is there a how-to?
>
>I'm sure you are. But fdisk will not help you. It cannot resize
>partitions, and since you'll only have two (a root, and a swap, both
>of which you need), you are not going to have anywhere to put windows.
>
Try the Ranish (sp?) partion manager. It is not quite as pretty as Partion
magic, but it does just about the same job, it's free, and it comes with its
own boot disk image (FreeDOS, I believe), so that it is very easy to run.
>As for a how-to, I'm pretty sure there isn't one. Why would someone
>want to put Windows on a perfectly functioning computer?
>
There are dozens of how-tos, and mini-how-tos, and other faqs and
information sites, about installing both Linux and every combination of the
dozen windoze varieties. It is extremly common to want both Linux and one
or more windoze flavours on the same machine (although most of the time,
people want to install Linux on a windoze machine rather than the other way
round).
>Something I didn't mention earlier, but someone else did I feel it's
>worth restating. Windows installation with overwrite all your
>partitions, destroying everything in it's path. You experiment is
>doomed.
------------------------------
From: "David Brown" <[EMAIL PROTECTED]>
Crossposted-To: alt.destroy.microsoft
Subject: Re: "The Linux Desktop", by T. Max Devlin
Date: Sat, 20 Jan 2001 11:34:43 +0100
T. Max Devlin wrote in message ...
>Said Ayende Rahien in alt.destroy.microsoft on Fri, 19 Jan 2001 23:59:11
>>"T. Max Devlin" <[EMAIL PROTECTED]> wrote in message
>>news:[EMAIL PROTECTED]...
>>> Well, here we go.
>>>
>>> I've got the "Linux Desktop" on order, from a company listed on
>>> linux.org. Its an 850MHz Athlon with 128 Meg of ram and a 40G ATA 100
>>> drive. CD-writer, printer, Logitech wheel mouse, PCI modem and a cheap
>>> Ethernet card; 19 inch monitor. RedHat 7.0, and I paid the extra bucks
>>> for the Deluxe box.
>>
>>RH 7.0 ?
>>On general, you should stay away from RH, and especially from .0 releases.
>>RH tend to put all sorts of bleeding edge stuff in those things, stuff
that
>>will make you bleed.
>
>Don't worry, I can handle it. ;-)
Have you considered Mandrake 7.2? Mandrake is often more "bleeding edge"
than Red Hat, and many find it easier to install and use. As far as I knew,
Mandrake is in fact the most popular distribution, not Red Hat (I may be
wrong, and it may depend on the market - I think Mandrake is especially
popular in the home, while Red Hat is preferred by businesses).
>
>>Most notable example is gcc in RH 7, I remember that there was some
problem
>>with 5.0, can't recall if there was something of the like in 6.0
>
>I have no plans whatsoever to ever need gcc, or care about its edges
>even if I do.
*Do* plan to use gcc - you are imposing silly restrictions on your system if
you do not install it. After all, it costs nothing (except a large chunk of
disk space), and it means you can make your own optomised kernel, or other
packages - a lot of programs benifit strongly from being locally compiled
with a configuration to suit your own setup. The binaries that come with
most distributions are fine for simple use, but often not for serious use
(for example, pre-compiled Apache is fine for a local net, but you would
want to configure and re-compile it before openning it on the internet).
Additionally, with gcc set up, you can update packages by downloading small
patches and re-compiling rather than waiting for the updates to trickle
through to Red Hat and official rpms. If you like the "bleeding edge", then
this is necessary. For example, you might want to keep closely up-to-date
with wine, so that you can run your windows programs without actually having
to reboot.
Skipping over gcc, and other language tools, is windows thinking. You don't
have to learn how to use it, or any related tools, nor do you have to
actually look at the source code for packages - just have them installed so
that they are available as needed.
Regarding the following explanation, I may well be wrong - if other people
know better, please correct me.
Regarding Red Hat 7.0's gcc (assuming you stick with RH7), you will want to
change to a more standard gcc (Red Hat will undoubtably have it ready as a
download). As far as I understand the problem, the gcc they installed has
incompatibilities in the object format (someone else mentioned name
mangling, suggesting the problem is mainly with c++). In itself, as a
stand-alone compiler, there is no problem - the hassle comes when trying to
link gcc-2.96 compiled objects and libraries with pre-2.96 compiled objects
and libraries. This means that when you re-compile the latest wine source,
it will not work unless you first re-compile the libraries. Then other
programs that use the same libraries will need re-compiling, and so on - a
serious mess.
The reason kernel compilation works fine is quite simple - the kernel is
compiled with a different version of gcc. It is very common for linux (or
other unix) systems to have more than one compiler installed, normally with
slightly different names. For example, the pentium-optomised version of gcc
is called pgcc, and a system might have both standard gcc and pgcc
installed. Normally, programs are compiled by default using gcc - this may
be a standard version, or it may be linked to pgcc, or whatever. The
kernel, however, is compiled by kgcc, so that it is easy to have a seperate
compiler for the kernel. On most systems, this is simply linked to the
system's main gcc, but it can be useful to have a known-good compiler for
the critical kernel compilation, while using a cutting-edge compiler for
everything else. Perhaps your system has standard (pre-2.96) gcc for its
kgcc, and the pentium-optimised pgcc for the main compiler. It is known
that pgcc can occasionally generate incorrect code - you can fix this in
most programs, but you don't want to risk compiler bugs getting into your
kernel. As far as I understand it, this is the sort of setup in RH7 - there
is a pre-2.96 compiler for kgcc, and the incompatible 2.96 for the main gcc.
Thus those who have successfully compiled their kernels with RH7 did so with
an earlier compiler, not 2.96.
I hope that makes sense (and I hope also that it is correct - I don't have
RH7, but I have followed some of the fuss about it in various Linux news
sites).
>
>>> It should be here next week. I didn't get the dual-boot option, but I
>>> plan to install 95, and maybe NT, once its up and running. So here we
>>> have a real-world comparison, taking into account and reflecting on the
>>> monopoly, pre-load, and ease of installation. The Win-whiners aren't
>>> going to agree, of course, but I think seeing just how easy it is to
>>> install 95 or NT on a box that has Linux preloaded is going to be very
>>> instructive. I've said I'd never build a PC from scratch again, and
>>> would prefer an OEM earned their profit by integrating the system for
>>> me. But in this case, the exact same hardware is supported by the same
>>> vendor as a dual-boot option, (can you believe it? an OEM selling
>>> dual-boot), so I don't think I'm going out on a limb. Plus which, if
>>> Windows for some reason is too much of a hassle to get up, I'll still
>>> have a functional system, so that might help eliminate the 'frustration
>>> and desperation factor' which so badly reflects on the monopoly in the
>>> typical scenario.
>>
>>Be sure to have a LILO boot disk around, you'll need it to reinstall LILO
>>(or your boot manager of choice) on the MBR after you install Windows.
>
>Yes, I know that getting Windows to 'play nice' will still be tricky.
>I'm wondering if anyone knows where there is a good step-by-step, or at
>least a how-to, on this? I've been looking, but there's SO much about
>Linux available that its really tough to know where to look.
>
It can often be safer to install windows first, then Linux afterwards. You
might also want to try alternative boot loaders, such as grub (modern linux
distributions, such as SUSE and Mandrake, use nicer and more convenient boot
loaders than the conservative Red Hat). Another fun tool is XOSL - extended
operating system loader, along with the Ranish (sp?) parition manager. This
will give you a neat graphical boot menu, with a lot of configuration
possibilities and quick access to a powerful partion manager.
If you have the time, and want to learn about the system, expect to install
and re-install both Windows and Linux a few times.
>--
>T. Max Devlin
> *** The best way to convince another is
> to state your case moderately and
> accurately. - Benjamin Franklin ***
------------------------------
** FOR YOUR REFERENCE **
The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:
Internet: [EMAIL PROTECTED]
You can send mail to the entire list by posting to comp.os.linux.advocacy.
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Advocacy Digest
******************************
