Linux-Advocacy Digest #618, Volume #26 Sat, 20 May 00 19:13:03 EDT
Contents:
No wanking or ploinking here please! (Charlie Ebert)
Re: Here is the solution (Leslie Mikesell)
Re: Need ideas for university funded project for linux (Leslie Mikesell)
Re: Need ideas for university funded project for linux (JEDIDIAH)
Re: Need ideas for university funded project for linux ([EMAIL PROTECTED])
Re: Things Linux can't do! (Leslie Mikesell)
Re: Things Linux can't do! (Leslie Mikesell)
----------------------------------------------------------------------------
From: Charlie Ebert <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: No wanking or ploinking here please!
Date: Sat, 20 May 2000 21:53:20 GMT
"Stephen S. Edwards II" wrote:
>
> abraxas wrote in message <8g6q6g$9im$[EMAIL PROTECTED]>...
>
> >In comp.os.linux.advocacy Charlie Ebert <[EMAIL PROTECTED]> wrote:
> >
> >> OH please! Stephen want's proof that NT or W2K blue screen.
>
> *sigh* He _STILL_ doesn't get it. No matter. I *PLOINK!*'ed the
> little wank minutes ago.
> .-----.
> |[_] :| Stephen S. Edwards II | NetBSD: Free of hype and license.
> | = :| "Artificial Intelligence -- The engineering of systems that
> | | yield results such as, 'The answer is 6.7E23... I think.'"
> |_..._| [EMAIL PROTECTED] | http://www.primenet.com/~rakmount
Your going to do or did do what?
Charlie
------------------------------
From: [EMAIL PROTECTED] (Leslie Mikesell)
Crossposted-To:
comp.sys.mac.advocacy,comp.os.ms-windows.nt.advocacy,comp.os.os2.advocacy
Subject: Re: Here is the solution
Date: 20 May 2000 17:29:28 -0500
In article <OmvV4.74662$[EMAIL PROTECTED]>,
Daniel Johnson <[EMAIL PROTECTED]> wrote:
>> We seem to have very different concepts of networking. Mine
>> is that you don't have to use any particular brand everywhere
>> just because you used a certain thing somewhere else. So
>> yes, I will use Linux some places, other things in other places.
>
>You can *insist* that networking is only networking if it uses
>Unix's protocols, but nobody else is obliged to care.
Please replace 'unix' above with 'platform independent'. In no
case have I ever suggested anything should be limited to unix.
It would make as little sense to say 'AS400' above, or Netware,
or Cisco, or the names of any of the hundreds of specific things
that interoperate with standard protocols. In fact most of
the interoperable protocols were originally developed on
platforms other than unix.
>If *you* care, what you can do about it is stick to Unix.
Unix vendors have been very up front about standards, but they
aren't the only ones. After all these years, Microsoft has
sort-of finally come around with an extra cost add-on pack
that covers a few things everyone else got right years ago.
>> I could easily have missed something, but I found where you
>> would manually syncronize files and then tried to set
>> up the scheduled runs and couldn't. According to the
>> on-line help I needed an active directory server to allow
>> scheduled file syncronization. The source was a remote
>> share, not on win2k, the destination would have been 2
>> remote win2k machines. I ended up using a cygwin-compiled
>> rsync in daemon mode on the target machines with a unix
>> cron job pushing the updates.
>
>I had no difficulty get my Windows 2000 Professional
>computer to to sync to files on a Windows 95 file share.
Yes, I could have done it once manually but that is not
what I wanted.
>LIke I said, perhaps I'm not thinking of the feature you are.
>Why would you wish to set up "scheduled runs"? It seems
>like you aren't doing what I'm doing.
I have a shared directory where web developers put their
changes locally. However, the images are really served
from a load-balanced set of servers at an off-site location
with better bandwidth. I don't want the developers to have
to deal with changes in the servers directly, so they just
get the images in place an hour before anything that references
them and the files are updated in the right places automatically.
Scheduled file syncronization sounded like the right choice.
I had hoped it would do something like rsync to send only
the necessary differences. I don't see the relationship
between AD and this, but that was what the help system said.
>The main problem DOS had was that it was really bad
>at handling printers. Arguably even *worse* than Unix,
>but frankly even *now* Unix isn't very much better than
>DOS, and is very much worse than MacOS.
>
>(NeXTStep excepted, of course.)
>From a technical perspective, postscript solves the
same problem as a common text/graphic metaformat
suitable for display and printing - better in
a lot of ways. Microsoft won this one on two fronts
price (compared to Adobe, but ghostscript demonstrates
that it is possible to use the documented standard without
duplicating the code), and development of fonts that
looked good at the low screen resolution of the time.
So this one is grunge work, not innovation.
>But it is worth noting that Microsoft copied the Mac *first*;
>they didn't bother with pre-emptive multitasking, memory
>protection, security, or multi-user support until *after*
>they had the indispensible basics- like printer and
>display support- working.
Then they should have partnered with someone who could deliver
the server side. And in fact they did - I used AT&T's starlan,
then StarGroup servers which put DOS, then Windows-for-workgroups,
then Win95 networking on top of a reliable OS with standard
networking services as well. However, like all MS partnerships
it was rocky and eventually fell apart. I understand HP also
gave up on their equivalent product. IBM may have one too but
I think it has gone the lanman route instead of trying to deal
with the weirdness of NT. Anyway, under unix-as-server, dealing
with DOS printers was not particularly difficult because you
could send postscript and have it processed by ghostscript
if necessary.
>> IBM had deep pockets and no shortage of lawyers. There has
>> to be more to the story. Perhaps a movie plot someday.
>
>Well, you are probably right. But I caution you against
>assuming that the "more to the story" is necessarily
>damning to Microsoft. Perhaps it was more that
>IBM was shockingly stupid, than that Microsoft was
>shockingly clever.
Maybe - I was thinking more along the lines of someone high
up at IBM having done something that exposed them to blackmail
from someone at Microsoft. Maybe I've been watching too
much TV.
>[snip]
>> >Well, I don't take taht sort of blanket claim seriously-
>>
>> It's true - and the reason I was using it. Dig out some trade
>> magazines from the time.
>
>Advocates *always* say this about *every* tool. It is boring. :D
OK, but I'm the one who used both Foxpro and Access, and the
Foxpro advocates were right.
>No competitor has ever beaten MIcrosoft because
>MS *let* them; they did so because they out performed
>Microsoft, when it happened.
But MS did not outperform many of the competitors they put
out of business. That is my real complaint, along with
the issue of putting all of your eggs in one basket.
>> It is the things I've seen recently
>> about the various threats that MS used to force the issue
>> that makes me think in retrospect that Dell didn't have
>> a choice.
>
>I tend to think they didn't either, but for a different reason: Unix
>was and is a very poor choice for a desktop machine, and had
>Dell stuck to it, they would have gone out of business when
>their competitors got Windows.
But the ISP business was just getting hot and with a tiny
amount of tweaking this could have been the perfect platform.
(In fact a local guy started with it and did very well).
This was just before BSDI brought out their relatively cheap
unix for generic PC's.
>> >Aren't we presuming that MS has a monolithic opinion
>> >on this?
>>
>> How many stockholders did it take for a majority vote
>> at the time?
>
>No idea. Are stockholders opinions the only ones you are
>interested in? I was under the impression that the decision
>to abandon OS/2 was made at the top.
Errr, the stockholders are the top.
>> >I bet the feelings of the Windows development team
>> >turned anti-OS/2 way earlier than those of, say, the
>> >Word team...
>>
>> Same stockholder(s) in control of decisions.
>
>Are they really? Or do they let MS's management make
>the decisions? If I were an MS stockholder, I wouldn't
>be interfering much.
In Microsoft's case, the top execs have always been the
majority stockholders. For a long time Bill G. himself
held the controlling interest - I'm not sure when that
changed, but regardless, this decision was made by a
small circle of friends with tight control over both the
OS and the apps. The thing that made windows click from
the start was the availability of apps, so you might say
that the OS had an unfair advantage starting out, just as
you might say the apps have it now. However, since they
did not have monopoly control at that point, all we can
complain about is the deception of everyone else.
>> No, standards can be modified/updated/replaced. But
>> a new standard doesn't happen because a single company
>> makes secret modifications.
>
>Admittedly. In *this* case MS didn't do that- they talked to
>the MIT people and got their modification into standard.
Does that mean it is documented well enough for others to
use?
>It does not seem to have done them any good on the PR
>front. :(
I don't understand. If they put the data you need to act
as a win2k controller into the standard, there should be
competing server versions by now and everyone should be happy.
>But it does make it relatively easy for non-MS clients
>to use MS Kerberos servers. They just ignore the
>(irrelevant to them) domain info.
Standards go both directions.
>I don't know what clients you consider "standard"; I suspect
>your opinion differs from that of the business users, most
>of whom see Windows as the standard.
Explain how anyone can consider one vendor's offering as
a standard, unless they don't know the meaning of the word.
>They have done so with this Kerberos thing, but they get
>no credit for it. Sometimes, it sucks to be Microsoft.
I'm sure everyone will be happy again when MIT releases the
reference code that follows this new standard and operates
as a domain controller as well as the kerberos server.
>> How about Netscape? Do you think MS had some effect on
>> their fate? Stac?
>
>These are both cases where they did. In Netscapes case,
>Microsoft built a better browser and gave it away for
>free. Netscape could give their browser away, and they did,
>but they were apparently not able to match Microsoft's
>quality.
If Microsoft could have won on quality, why did they have
to integrate the browser into the OS and force hardware
vendors to bundle it and not include Netscape?
>They still held out for a remarkably long time- which
>shows how useful it is to be an entrenched near-monopoly
>in your field.
No, it shows how dreadful the versions of IE before 5.0
were. Even with the dirty tricks to get them on everyone's
machine they still replaced them with the better alternative.
And now the company producing it has been effectively
stoped, so we won't see more of these innovations.
>I've heard conflicting stories about the Stac case; some
>make MS out to be quite the villain, outright pilfering Stac
>code. Others just have MS enhancing their OS with
>a similar capability. I bet I can guess which one you
>believe. :D
As far as I can tell it is par for the course for anyone
who tried to do business with MS. Do you have some examples
of companies with a long and profitable relationship in
a joint venture? Maybe IBM is the longest running case.
How does Lotus feel about them (remember - the company whose
product drew everyone to DOS in the first place)?
>> Historically they have been responsive only to competition.
>
>Oh, I don't know. DirectX met their users demands; as
>far as I know it had no competition then. (Now it does,
>in the form of Apple's GameSprockets, but I think
>DirectX came first)
When did OpenGL come along?
>I don't know if Standard Oil would *really* have raised
>prices eventually. They were broken up to keep them
>from doing that, on the theory that if they were not they
>would have been able to gouge customers once
>their monopoly was totally secure.
>
>But none of that happened, and it is speculation to
>say it would have.
Just that historically it has happened every time
the opportunity was there.
>> I don't, and I don't think we need a single company
>> in control of software either.
>
>I don't see that as a real prospect.
It is a very real (and scary) prospect that many enterprises will
be arranged so that their logins, passwords, and access permissions
can only be controlled by a product from a single company - a
company that does not allow open inspection of its code and has a
rather embarassing history of security issues. It is impossible
to prove that there are no built-in back doors in this scheme
and there are no alternative implementations from a different,
less secret code base.
>But it is illustrative. You've spoken of "standards";
>there are presently two known ways to get standards.
>One way is the way Microsoft promulgates standards-
>they have a dominant product, and whatever they do is
>"standard" because of it.
>
>The other way is by negotiated mutual agreement between
>vendors; lots of vendors get together and hammer out something
>they can agree on.
No, only one of those things is a standard.
>So far it seems to me that the first usually works better;
>it permits you to go beyond the mere codification of
>existing practices and to 'push the envelope' with new
>features and ideas.
This may appear to be true for a short time. Then when someone
else does it even better you are locked into the previous
technique because you would have to change every component
at once to accomodate the new one.
>The 'standards body' approach seems
>to be largely unable to do this.
Yes, they correctly understand the problems of using protocols
that do not allow choices among platforms or vendors.
>The problem seems to be that the 'standards body' approach
>is *fair*; it can't favor one vendor over another very easily. It
>can't do this because the other vendor- the one being placed
>at a disadvantage- can scuttle the standard is sufficiently
>irritated. It won't work unless this is avoided.
And that is a problem? Why should one vendor be able to lock
you into their product?
>This is just one issue; it does not by itself prove that
>having a dominate company dictating standards
>is the One True Way. But it shows that this is not
>a *uniformly* bad thing.
No, but other things show that it is uniformly a bad thing.
Like paying attention to history. It is a quirk
that popular processors have maintained backwards compatibility
for such a long time in recent history. Looking at it
objectively, most people would probably think it is actually
bad in the long run and it is time for a change. If you
have locked yourself into something that requires a
particular CPU type or backwards compatibility to it, what
do you do when a real improvement comes around - or the
company that produces the one-and-only choice you have
goes out of business or just decides to stop making that
item?
>In light of that, I think it must be admitted that the
>presumption that "monopoly" = "bad" should
>be questioned; it may be true but it needs to be
>jusitified- perhaps on a case-by-case basis.
Monopolies may be acceptable, but then they need to
be closely regulated. They have the power to cause
near-universal damage on a whim. Even if (in spite
of all evidence...) you think they won't exercise
that whim, regulation must exist to make it impossible.
Les Mikesell
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Leslie Mikesell)
Crossposted-To:
comp.os.linux,comp.os.linux.development,comp.os.linux.development.apps,comp.os.linux.development.system,comp.os.linux.misc,comp.os.linux.setup
Subject: Re: Need ideas for university funded project for linux
Date: 20 May 2000 17:37:50 -0500
In article <[EMAIL PROTECTED]>,
JEDIDIAH <[EMAIL PROTECTED]> wrote:
>>>It is the RPM BS that has caused me to abandon that format whenever
>>>possible. Instead, I prefere to install software from source.
>>>Packages that conform to the ./configure, make, make install mantra
>>>are easy to build and put where you want them.
>>
>>You left out the dozen obligatory arguments to ./configure that
>>are different for every package to make it interoperate with
>
> I dunno about you, but I rarely if ever actually need to
> use any of those options...
You need them IF you have installed a modern distribution that
already includes the thing you are updating, and you want
it all to work the same after you apply your fix.
>[deletia]
>
> The point of automation is to avoid such manual futzing.
You do avoid it if you wait till someone else does it and
then just install the packaged version. If you have some
reason to need a fix the day a patch is out or need some
local changes until the next release, you need the futzing
but with the rpm scheme even most of the futzing is automated.
Les Mikesell
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (JEDIDIAH)
Crossposted-To:
comp.os.linux,comp.os.linux.development,comp.os.linux.development.apps,comp.os.linux.development.system,comp.os.linux.misc,comp.os.linux.setup
Subject: Re: Need ideas for university funded project for linux
Date: Sat, 20 May 2000 22:46:43 GMT
On 20 May 2000 17:37:50 -0500, Leslie Mikesell <[EMAIL PROTECTED]> wrote:
>In article <[EMAIL PROTECTED]>,
>JEDIDIAH <[EMAIL PROTECTED]> wrote:
>
>>>>It is the RPM BS that has caused me to abandon that format whenever
>>>>possible. Instead, I prefere to install software from source.
>>>>Packages that conform to the ./configure, make, make install mantra
>>>>are easy to build and put where you want them.
>>>
>>>You left out the dozen obligatory arguments to ./configure that
>>>are different for every package to make it interoperate with
>>
>> I dunno about you, but I rarely if ever actually need to
>> use any of those options...
>
>You need them IF you have installed a modern distribution that
>already includes the thing you are updating, and you want
Who said anything about 'updating'. I'm talking about the
'new' stuff I compile. I typically don't bother with source
for more stable projects.
>it all to work the same after you apply your fix.
>
>>[deletia]
>>
>> The point of automation is to avoid such manual futzing.
>
>You do avoid it if you wait till someone else does it and
>then just install the packaged version. If you have some
Is this supposed to be describing binary packages or
makefiles, as I've always thought of reasonably
complete source packages as serving this purpose.
>reason to need a fix the day a patch is out or need some
>local changes until the next release, you need the futzing
>but with the rpm scheme even most of the futzing is automated.
Where I've found rpm most useful are those projects that
seem to be made of a million or so parts and doing a
'build World' is a manual process.
--
In what language does 'open' mean 'execute the evil contents of' |||
a document? --Les Mikesell / | \
Need sane PPP docs? Try penguin.lvcm.com.
------------------------------
Crossposted-To:
comp.os.linux,comp.os.linux.development,comp.os.linux.development.apps,comp.os.linux.development.system,comp.os.linux.misc,comp.os.linux.setup
Subject: Re: Need ideas for university funded project for linux
From: [EMAIL PROTECTED]
Date: Sat, 20 May 2000 22:59:29 GMT
[EMAIL PROTECTED] (JEDIDIAH) writes:
[is OpenGL hardware-accelerated?]
> >No. It isn't.
> >It may have the potential to be accelerated at some point in the
> >future, but, as of this writing, it is not. NVIDIA has flatly stated
> >that they will not be doing hardware-accelerated OpenGL until XF86
> >4.0. As XF86 4.0 is not the official XF86 at this point, there is
> Says who? There's already at least one distro that's shipping it.
Sorry. I was operating under outdated information.
You're right; it is the official XF86. (I must've checked its status
the day before it was released...)
> >> > A killer app is something that most computer users will find
> >> > useful.
> >> Of course Apache is a killer app.
> >Of course it is not.
> Netcraft and the hype in general about the Web would tend
> to flatly contradict you.
I really don't see what's so hard to understand here.
Here are some requirements for a killer app:
1. Lots of people have to use it.
2. Lots of people have to know about it.
Obviously, then, to be a potential killer app, a program must appeal
to lots of people.
Does Apache appeal to lots of people?
No. I'm sorry, but this is bloody obvious, and I really don't
understand how anyone can argue with it. The vast - *VAST* - majority
of computer users have not installed Apache, and never *will* install
Apache, no matter that it's the best thing since sliced bread.
There's just no purpose to it.
Having lots of people use Apache in the sense of getting Web pages
through it (like, over the 'net) does not qualify it as a killer app.
The only people who count, numbers-wise, are the ones who administer
it, and that's never going to hit anywhere near a large enough figure
for Apache to be a killer app.
You may make the claim that a killer app only has to draw lots of
people, relative to the audience an OS (or arch, or whatever)
currently enjoys, and that's fine. But Apache is never going to draw
people to Linux, since it's available in stable and up-to-date form on
many other OSes.
--
Eric P. McCoy ([EMAIL PROTECTED])
non-combatant, n. A dead Quaker.
- Ambrose Bierce, _The Devil's Dictionary_
------------------------------
From: [EMAIL PROTECTED] (Leslie Mikesell)
Crossposted-To: comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: Things Linux can't do!
Date: 20 May 2000 17:50:11 -0500
In article <8g6s6u$f0k$[EMAIL PROTECTED]>,
Christopher Smith <[EMAIL PROTECTED]> wrote:
>> My S.O.s laptop (running W2K professional as well) has now
>> bluescreened 4 times, twice while trying to quite RealPlayer.
>> (trying to quit IE was what did it the first time on my desktop
>> machine). Luckily the last couple of bluescreens on that laptop
>> havent been as bad as the first (when I horror of horrors, plugged
>> a USB mouse into the machine) which resulted in a 'NO KERNEL FOUND'
>> (or something very close to that) error apon reboot.
>I always find it amazing that this sort of thing:
>a) *Never* happens to anyone I know
I've had two bad personal experiences with NTFS filesystems. One
was on a box someone else (long gone) at work had configured to
collect news stories from a wire service into files and make
them available via the web. It probably collected a few hundred
small files into a new directory every day. It crashed every
few weeks (don't remember the service pack level) and eventually
started to take longer and longer to chkdisk on the way back.
Finally it didn't finish over a 3 day weekend and I gave up.
The other one was my desktop machine which had a CPU fan fail and
crashed several times in operation. I fixed the fan and it doesn't
crash now, but the file system is broken even though chkdisk/scandisk
says it is OK. If I try to install sp6a, I get file errors reading
the files it just thought it sucessfully unpacked into a temp
directory. I don't know of any way to fix this. I've had files
get lost to a unix fsck after a crash but I've never seen the
filesystem fail after fsck cleaned it up.
>b) Always seems to happen to faceless people on usenet who spend most of
>their time cursing Microsoft.
I'm certainly willing to believe that those other people have good
reasons for cursing too.
Les Mikesell
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Leslie Mikesell)
Crossposted-To: comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: Things Linux can't do!
Date: 20 May 2000 17:58:52 -0500
In article <8g6up4$qb7$[EMAIL PROTECTED]>,
Stephen S. Edwards II <[EMAIL PROTECTED]> wrote:
>
>As Christopher pointed out, I find it odd that nobody that I
>know of that uses either Windows9x, or WindowsNT ever notes
>that they've had any troubles with them (and I'm usually the
>first person my friends come to).
I have, on my desktop, an NT box that will not complete the
installation of sp6a due to disk errors that happen in the
temp files after it unpacks them. Chkdisk say the partition
is OK. I'd appreciate any advice on how to fix this without
having to reinstall all the software loaded on the box.
(Hardware problems caused the crash that corrupted the disk
and this has since been fixed).
>On EFNet WindowsNT, the most you ever see are people with
>configuration questions.
Maybe they've given up and re-installed already.
>On EFNet #Linux, you see a lot
>of people having difficulty getting hardware and software
>to work together properly.
And perhaps this is because they know that most of the
problems have easy solutions.
Les Mikesell
[EMAIL PROTECTED]
------------------------------
** 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 (and comp.os.linux.advocacy) via:
Internet: [EMAIL PROTECTED]
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
******************************
