Linux-Advocacy Digest #639, Volume #27 Thu, 13 Jul 00 02:13:03 EDT
Contents:
Re: Richard Stallman's Politics (was: Linux is awesome! (Leslie Mikesell)
Re: Richard Stallman's Politics (was: Linux is awesome! (T. Max Devlin)
Re: Linsux as a desktop platform ("Christopher Smith")
Re: Richard Stallman's Politics (was: Linux is awesome! (T. Max Devlin)
Re: Linsux as a desktop platform ("Christopher Smith")
Re: Linsux as a desktop platform (Leslie Mikesell)
Re: Tinman digest, volume 2451736 (Slava Pestov)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Leslie Mikesell)
Crossposted-To: gnu.misc.discuss
Subject: Re: Richard Stallman's Politics (was: Linux is awesome!
Date: 13 Jul 2000 00:12:18 -0500
In article <[EMAIL PROTECTED]>,
T. Max Devlin <[EMAIL PROTECTED]> wrote:
>>>All commercial
>>>software is distributed with two different licenses, in much the same
>>>way you describe. They're called "developer license" or "OEM license"
>>>and "End User License Agreements", respectively.
>>
>>But the purpose of applying these licenses is to apply certain
>>restrictions - in perl's case it is to avoid them.
>
>I was hoping for a response that would give some comparison to the two
>license dichotomy, not a value judgement on the motivations of the
>authors in choosing their license strategy.
So read them. An internet search for perl and Artistic License should
scare up a copy if you don't happen to have a perl source distribution
around (hmmm... why was it that you claim to like the GPL anyway?).
>>One is the stock GPL, the other is called the 'Artistic' license
>>which allows just about anything except redistributing modifications
>>without identifying the changed copy as being modified. Embedding
>>in commercial products is also explicitly permitted. This scheme
>>allows it to be combined with GPL'd code without being trapped
>>by it.
>
>Does this, in fact, end up being similar to commercial software, in that
>the GPL is the "end user", and the Artistic license is only of
>restriction or interest to developers?
No, both are irrelevant for an 'end user' except to the extent that
an end user is allowed to obtain perl code built into many things
that would not be possible if only the GPL applied. The GPL
doesn't and can't cover what an end user does with the code. Since
it is based on copyright law it is only binding regarding the
way the code is copied and distributed. However, if only the
Artistic license applied, then it would be impossible to distribute
derived works containing GPL'd code, and if only the GPL applied
it would be prohibited to distribute derived works containing
non-GPL'd code. And there are plenty of both kinds with real
end-user benefits.
Les Mikesell
[EMAIL PROTECTED]
------------------------------
From: T. Max Devlin <[EMAIL PROTECTED]>
Crossposted-To: gnu.misc.discuss
Subject: Re: Richard Stallman's Politics (was: Linux is awesome!
Date: Thu, 13 Jul 2000 01:18:42 -0400
Reply-To: [EMAIL PROTECTED]
Quoting Leslie Mikesell from comp.os.linux.advocacy; 11 Jul 2000
[...]
>I thought everyone involved in the discussion would already
>know the history of the RIPEM case, and would be aware of
>all the extra work the Linux and *BSD volunteers have had
>to do to avoid the restrictions of each other's licenses.
Thanks for the help, Clyde.
[...]
>>It is a problem because it could happen, not because it has. But it
>>has, not as clearly and definitively as you might insist must occur for
>>your to be convinced, but profiteering on software is not a rare thing
>>at all these days, and since all software is based on the work of others
>>as well as the author, all software being GPL is a good thing.
>
>No it isn't. Having a choice as a good thing. The only way
>someone can 'profiteer' on unrestricted software is to add
>enough value that many people choose to their version instead
>of the freely available version in spite of the cost and they
>have to keep doing it in competition with others adding
>to the free base.
Why do you call software I have to agree to a trade secret license to
use "unrestricted"? Yes, yes, I know, you're talking about non-GPL open
stuff. But what you aren't getting is that this is no different in
character from commercial trade secret licenses; it is merely a less
restrictive one. It is by no means unrestrictive, because it does not
*remove* the restrictions that copyright law already puts in place.
I am not arguing against non-GPL open source. I am arguing against
non-GPL software. Between commercial and open, obviously I'd prefer
open. But that isn't "free enough" for RSM, and it isn't free enough
for me, either. If GPL is "too free" for some developers, then don't
use GPL. Sooner or later, you will find more efficient business models;
you will have to, the market will demand it. But it won't be a
precipitous change. Its going to take many years. But once a typical
user can function equitably on a personal computer with open source
software, then the need for end-user licensing will be gone, because
people will no longer need to accept licensing restrictions to gain
access to software, and they will no longer be willing to do so. Then
all open source might as well be GPL, because the copyright law already
prevents what you feel sure will kill your ability to profit from your
work. The only thing that can stop people from paying someone to "sell
them software" is when the people don't want software. They'll always
want it, so programmers will remain employed.
Will you be able to make billions of dollars on software any more than
on books or magazines or documentaries? No. Does that stop books and
magazines and documentaries from being produced? No.
[...]
>>I concur that making software public domain would be better than GPL. I
>>don't believe that any other open source licenses are "well tested,
>>freely usable bases of code", because they allow someone using that code
>>to place restrictions on others re-use of that code, if they can combine
>>it with secret source code.
>
>But you don't have to use that version unless you have some reason
>to consider it worth the price, and they cannot keep you from
>using the original in any case.
Others have been through this argument several times; take it up with
them. Ask Jedi; he'll explain it to you.
[...]
>I have no idea what you are talking about here. How does, (say)
>a bit of BSD tcp code in a Cisco router, perhaps merged with
>proprietary SNA from another source, harm your ability to
>freely do anything you want with your own copy of TCP from
>the BSD base?
You're tilting at windmills, dude. The answer is "I don't care".
> In my opinion, the existence of the Ciscos
>forming the internet backbone greatly enhances your ability
>to use your own TCP copy rather than taking anything away.
Except Cisco's don't form the internet backbone, and the certainly don't
enhance my ability to use TCP. But your argument disassembles the very
basis of the Internet to begin with, so I won't bother trying to explain
it.
>Or, look at Sun who made proprietary versions but contributed
>back NFS. How were you harmed by that? Had the base code
>been GPL'd, those companies probably could not have started
>up and we'd all be running OSI or SNA now, under the control
>of a single large company who could afford to write everything
>from scratch.
Why do all of your examples require these massive assumptions about what
a market which doesn't exist would "probably" do? I do not *need* to
second-guess the market. Don't ANY of you people understand free
enterprise?
No, you don't. Because you assume that the reason the market is there
is to give you an opportunity to make money.
--
T. Max Devlin
Manager of Research & Educational Services
Managed Services
ELTRAX Technology Services Group
[EMAIL PROTECTED]
-[Opinions expressed are my own; everyone else, including
my employer, has to pay for them, subject to
applicable licensing agreement]-
====== Posted via Newsfeeds.Com, Uncensored Usenet News ======
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
======= Over 80,000 Newsgroups = 16 Different Servers! ======
------------------------------
From: "Christopher Smith" <[EMAIL PROTECTED]>
Crossposted-To: comp.sys.mac.advocacy,comp.os.ms-windows.advocacy,comp.unix.advocacy
Subject: Re: Linsux as a desktop platform
Date: Thu, 13 Jul 2000 15:25:57 +1000
"T. Max Devlin" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Quoting void from comp.os.linux.advocacy; 12 Jul 2000 05:05:24 GMT
> [...]
> >Single-user or not, nobody wants their computer locked up because one
> >application has a serious bug. Operating systems should be resilient
> >against programmer error, because bugs happen and they happen a lot.
>
> No, desktop client operating systems need to be more resilient to user
> errors.
True.
> Programmers are assumed to have done their job correctly.
No. Programmers should never be assumed to have done their job correctly.
They are the single most fallible part of the development process.
> That
> this doesn't always happen is not the issue, as lock-ups do occasionally
> occur (at least X dies, which is good enough for most end users) even on
> PMT systems.
They're a lot more common when a program that gets caught in a while(1) can
lock up the entire OS due to a brain dead scheduler.
> >Can you find a way to make cooperative multitasking as robust as
> >preemptive multitasking?
>
> No. Can you find a way to make PMT as user-responsive as CMT?
It already is more so. The main *point* of pre-emptive multitasking is to
improve response times, at the expense of some (irrelevant these days)
performance.
> Then all
> you're doing is implementing CMT.
You nearly emulate CMT in almost all PMT systems by upping the process
priority as high as it will go. Much like CMT, this tends to lead to
undesirable effects.
> The result is the value, not the
> process. I don't care *how* you do the scheduling. As long as whatever
> program I'm working in has, as far as I am concerned 100% of the
> available time to keep up with me, even if it spends a lot of that time
> waiting. Of course, background processes shouldn't be ignored, but they
> only rarely have true priority.
On a PMT system, for the vast, vast majority of people, the CPU spends most
of its time twiddling its thumbs waiting for something to do. Personally
I'd rather the CPU spend its thumb twiddling time downloading web pages in
the background, printing things etc.
> Again, I'm over-simplifying the case by assuming that everything is user
> applications, and I/O devices (network, drive) don't screw things up
> because they were locked out.
You're not over-simplifying, you're just plain wrong because you have no
idea what you're talking about.
> I'm not sure how those details are
> handled, so perhaps I am merely arguing for a PMT system that pretends
People a lot smarter than you went through this sort of thing decades ago.
It's been a fairly well understood system for a *very* long time. The
details have already been handled, and handled well.
> to be CMT, now that we are no longer limited to the stand alone
> "austere" environment which the Mac was developed in. But having you
> guys argue against the logic so hard and only concentrate on the
> engineer's view is a bit disconcerting.
There is no logic in what you are arguing. Early PC systems used CMT
scheduling primarily because of the hardware overhead in anything more
complex.
> >Anyway, the type of multitasking and the behavior of the GUI are not as
> >tightly coupled as you think.
>
> Yes, that's what I've finally realized. Yet I suspect they are not as
> unrelated as theory indicates. I would still always like to have the
> GUI have preference in multitasking.
Every PMT GUI-based OSes I know of boosts the priority of whichever app is
in the foreground. They've been doing this for *years*.
Like I said, this has been understood and addressed for a *long* time.
> >No, it's lousy. Even when written with the best intent in the world,
> >programs screw up and fail to yield the processor. This scheme also
> >results in less efficient use of processor time, which in the end means
> >that things take longer.
>
> So get rid of them and get other programs.
You can't, when there *are* no other programs.
This is an inherent problem in any CMT system that is used for a general
purpose, because app creators *cannot know* what other programs will be
running with their apps. This is why CMT is a Bad Idea for a general
purpose OS.
> I don't care if "things"
> take longer; I'm the only one with important things to do. The computer
> is just a desktop, not a Cray doing nuclear weapons testing.
>
> >These things degrade the user experience, drastically in the first case,
> >subtly in the second.
>
> PMT is not subtle in its annoyance value for the user, any more than
> modal dialogs are.
I have yet to see anyone be "annoyed" by PMT.
You can often watch CMT-enduced stress rise by the minute, however.
> >I disagree, and so do Apple, Microsoft, Be, and anyone else who might be
> >producing new desktop operating systems.
>
> Yes, and I'm worried it means that's not a single person who can be an
> engineer and still maintain an end-users perspective. I guess I'm
> probably wrong, but you guys have done little to raise my comfort value.
They are maintaining and end-users perspective. PMT systems are designed to
be responsive above all else - precisely what an end user wants.
> >But on unix, the background task will only slow very slightly, while
> >interactive apps are still nice and responsive. You should read up on
> >the algorithm used to do this, it's quite clever.
>
> I have no use for clever algorithms! You speak in theoretical cases.
No, he speaks of what every other desktop OS except MacOS does, and does
well. This isn't theory, it's well understood and executed *practice*.
> When I've got five program instances running, I want the one I'm *using*
> to be the one taking up almost all of the computers time.
Foolish. Your downloads will time out, your print jobs will take orders of
magnitude longer and your CPU will be needlessly heating up and wasting
power.
> When I go to
> thirty eight instances (I've done it; just me browsing the web, and
> ignoring the many other processes), I don't want the front one to be any
> slower, not just "only very slightly" slower.
It won't be on any remotely modern CPU, they've all got cycles to burn for
this sort of usage.
If you can perceive time increments measured in nanoseconds well, life must
be pretty boring.
> >No. There is no benefit that comes with this waste. There are
> >scheduling algorithms that can take advantage of unused cycles without
> >slowing down interactive processes noticeably.
>
> My point is that cycles spent waiting for the user on a desktop client
> system are not *wasted*. They are *spent*. Waiting for me.
No, they are wasted. They are wasting power, creating heat, shortening the
lifespan of your system for *no* *noticable* *change*.
> Engineers
> have a warped, not inaccurate, but different, view of "noticeable" than
> end users do.
>
> >>It wouldn't surprise me if it was a Linux which allowed adjustment to
> >>just how pre-emptive the multi-tasking is.
> >
> >That's because you overestimate the role of multitasking in determining
> >how the GUI functions.
>
> No, I overestimate the role of approach in how engineering gets done.
> I'm very ruthless in this regard: the user counts; the engineer's
> theory's don't.
The engineer's theory have been being put into practice for a very long
time.
[chomp]
Max, it's blatantly obvious you have no knowledge, understanding or
experience whatsoever of the issues involved here.
------------------------------
From: T. Max Devlin <[EMAIL PROTECTED]>
Crossposted-To: gnu.misc.discuss
Subject: Re: Richard Stallman's Politics (was: Linux is awesome!
Date: Thu, 13 Jul 2000 01:19:35 -0400
Reply-To: [EMAIL PROTECTED]
Quoting Austin Ziegler from comp.os.linux.advocacy; Tue, 11 Jul 2000
23:38:54 -0400
>On Tue, 11 Jul 2000, T. Max Devlin wrote:
>> Masterfully summarized, Jedi. I now officially claim that people who
>> think Free Software isn't Free are people who think freedom for some
>> people is Freedom, and that makes them as bad as Hitler.
>
>And you were complaining about those of us who are rationally opposed to
>the lies behind the GPL as being trolls?
Yes, you're right. I lied. I called you Hitler, and then kept posting.
Mea culpa.
Goodby. If anyone wants me, I'll be in alt.destroy.microsoft.
--
T. Max Devlin
Manager of Research & Educational Services
Managed Services
ELTRAX Technology Services Group
[EMAIL PROTECTED]
-[Opinions expressed are my own; everyone else, including
my employer, has to pay for them, subject to
applicable licensing agreement]-
====== Posted via Newsfeeds.Com, Uncensored Usenet News ======
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
======= Over 80,000 Newsgroups = 16 Different Servers! ======
------------------------------
From: "Christopher Smith" <[EMAIL PROTECTED]>
Crossposted-To: comp.sys.mac.advocacy,comp.os.ms-windows.advocacy,comp.unix.advocacy
Subject: Re: Linsux as a desktop platform
Date: Thu, 13 Jul 2000 15:29:39 +1000
"Leslie Mikesell" <[EMAIL PROTECTED]> wrote in message
news:8kiqe5$30et$[EMAIL PROTECTED]...
> In article <8ki26f$9am$[EMAIL PROTECTED]>,
> Christopher Smith <[EMAIL PROTECTED]> wrote:
> >
> >> AFAIK, the only problems that Linux has with IRQs is people who don't
> >> understand them, whereas Windows seems to mostly have problems with
> >> people who do understand them, as it seems to resist dealing with them
> >> simply and easily. I'm not saying they are a simple issue; I'm well
> >> aware of the problems which occur on any OS using the archaic and
arcane
> >> IRQs which are part of that platform.
> >
> >IRQs are a pretty simple issue. If you have cards that cannot share
IRQs,
> >then they must each have a unique one. If you have more cards than IRQs,
> >you're stuffed.
>
> They are note quite that simple. ISA cards can't share IRQs, but
> with PCI it is up to the driver software. Linux drivers
> don't like to share them.
Which is exactly what I just said.
> >The average PC has about 5 free IRQs. The logic behind determining
whether
> >or not a given IRQ is used or free, and thus whether or not it can be
> >assigned, is not difficult.
>
> PCI cards have the IRQs assigned by the motherboard bios at
> start-up. You don't get much control beyond being able to
> reserve some for the ISA slots.
PnP PCI IRQs are almost always hard wired to specific slots on the
motherboard, if it really becomes essential that PCI cards not share IRQs.
I can't think of any PCI motherboard for which this is not true.
> >> The relative statement that Linux
> >> 'doesn't have' IRQ problem obviously refers to whatever prevents
Windows
> >> from working correctly.
> >
> >The same thing that will stop Linux working correctly.
>
> This depends on the drivers' ability to share.
I believe this discussion was about non-PnP peripherals that cannot share.
> I recently had
> a PCI ethernet and SCSI card sharing an interrupt under
> Win98 but had to free up another one by disabling the motherboard
> serial port to make it work with Linux.
That is a Linux issue, not an IRQ issue.
> On the other hand,
> now I've swapped the network card and the new one came up
> right under Linux/ipchains but the Win98se internet sharing
> is completely screwed up.
Quite strange. What type of card ?
------------------------------
From: [EMAIL PROTECTED] (Leslie Mikesell)
Crossposted-To: comp.sys.mac.advocacy,comp.os.ms-windows.advocacy,comp.unix.advocacy
Subject: Re: Linsux as a desktop platform
Date: 13 Jul 2000 00:38:08 -0500
In article <8kjjpn$pcv$[EMAIL PROTECTED]>,
Christopher Smith <[EMAIL PROTECTED]> wrote:
>> On the other hand,
>> now I've swapped the network card and the new one came up
>> right under Linux/ipchains but the Win98se internet sharing
>> is completely screwed up.
>
>Quite strange. What type of card ?
I originally had the machine set up for dual boot with
internet sharing on the modem, then recently got
a cablemodem and put in a Farallon card (very new
with a realtek chip). That one worked with win98 sharing
the interrupt with a scsi card and I bound the internet
sharing to it. Then I swapped it for an old DEC-chip
NetGear and opened up another interrupt to get Linux
working. When I switched back to Win98 and reloaded
the card driver the internet sharing still appeared to
be bound to the new one but it didn't work. I haven't
spent much time trying to figure out why - I just bought
one of those cheap LinkSys routers that have a 4-port
switch built in and don't need sharing now.
Les Mikesell
[EMAIL PROTECTED]
------------------------------
From: Slava Pestov <[EMAIL PROTECTED]>
Crossposted-To:
comp.sys.mac.advocacy,comp.os.os2.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: Tinman digest, volume 2451736
Date: Thu, 13 Jul 2000 06:04:18 GMT
tinman wrote:
>
> In article <[EMAIL PROTECTED]>, Marty <[EMAIL PROTECTED]> wrote:
>
> > Tinman wrote:
> >
> > 1> Jumping into conversations again Karl? Cool, have fun!
> >
> > Still posting for entertainment purposes, eh Tinman?
>
> That's tinman. ('
On what basis do you make that claim?
> And why else would I post?
Don't you know?
>
> > Not surprising,
> > considering that you are being digestified.
>
> On the contrary.
Prove it, if you think you can.
> My polycarbonate exterior resists digestification.
What alleged "polycarbonate exterior"?
Slava
------------------------------
** 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
******************************
