Linux-Advocacy Digest #227, Volume #34 Sat, 5 May 01 19:13:04 EDT
Contents:
Re: Alan Cox responds to Mundie (GreyCloud)
Re: Shared library hell (Donn Miller)
Re: The long slow slide to Microsoft.NOT ("Mart van de Wege")
Re: The long slow slide to Microsoft.NOT ("Ayende Rahien")
Re: MS should sue the pants off linux-mandrake (was: Re: Winvocates confuse me -
d'oh!) ("Tom Wilson")
Re: The long slow slide to Microsoft.NOT (GreyCloud)
Re: Alan Cox responds to Mundie (Craig Kelley)
Re: Shared library hell ("Mart van de Wege")
----------------------------------------------------------------------------
From: GreyCloud <[EMAIL PROTECTED]>
Subject: Re: Alan Cox responds to Mundie
Date: Sat, 05 May 2001 15:39:45 -0700
Erik Funkenbusch wrote:
>
> "Craig Kelley" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
> > "Erik Funkenbusch" <[EMAIL PROTECTED]> writes:
> >
> > > "Craig Kelley" <[EMAIL PROTECTED]> wrote in message
> > > news:[EMAIL PROTECTED]...
> > > > "Erik Funkenbusch" <[EMAIL PROTECTED]> writes:
> > > >
> > > > > You, and most other people are confusing GPL and Open Source.
> Mundies
> > > > > comments are particularly against the GPL, not Open Source.
> > > > >
> > > > > While MS made comments about Open Source having a difficult business
> > > model
> > > > > to sustain, its primary beef was with the GPL. Notice that they are
> > > only
> > > > > questioning the business model of Open Source, but are attacking the
> > > GPL's
> > > > > effect on business directly.
> > > > >
> > > > > Don't make the mistake of trying to claim MS is against Open Source.
> > > They
> > > > > could care less if someone gives their code away. What they care
> about
> > > is
> > > > > that the GPL prevents businesses from taking advantage of code paid
> for
> > > by
> > > > > taxpayer dollars.
> > > >
> > > > No, I'm not confusing anything. You're trying to back-peddle on your
> > > > claim that proprietary intellectual property was the *real* motivating
> > > > force behind the internet.
> > >
> > > Indeed it was. Without proprietary IP, the companies involved in
> > > growing and commercializing the internet would have never done it.
> > > The Internet was founded on completely open and public domain
> > > information, which is what allowed companies to make their own
> > > proprietary versions.
> >
> > Who said anything about growing and commercializing it? I am talking
> > about *developing* it.
>
> Guess that depends on what you mean by "developing". It would have still
> been invented, but developing means "growing".
>
> http://www.m-w.com/cgi-bin/dictionary?va=developing
>
> "3 a : to make active or promote the growth of <developed his muscles> b (1)
> : to make available or usable <develop natural resources> (2) : to make
> suitable for commercial or residential purposes c : to move (as a chess
> piece) from the original position to one providing more opportunity for
> effective use"
>
> "4 a : to cause to unfold gradually <developed his argument> b : to expand
> by a process of growth <developed a strong organization> c : to cause to
> grow and differentiate along lines natural to its kind <rain and sun develop
> the grain> d : to have unfold or differentiate within one <developed
> pneumonia>"
>
> > > We would not have the internet as we know it, and probably would
> > > have never even had Linux, had it not been for the fact that the
> > > protocols and code were made publicly available for unrestricted
> > > use.
> >
> > I disagree. Apply natural selection to technology and you *will* end
> > up with open source under general circumstances (ie, not some niche or
> > bleeding-edge projects). The internet could not have been without
> > open source, and it will only continue on those merits.
>
> You are again confusing Open Source with GPL. My argument is only about the
> GPL, and how the Internet could not have become what it is if it had been
> released under GPL semantics.
>
> > Linux is a by-product of this, and would always have been regardless.
> > The internet succeeded where multi-billion dollar companies failed
> > (Apple's E-World, AOL's QuantumLink/AOL, Microsoft's MSN, Compuserve,
> > The Source, Delphi -- one failure after another, falling down to open
> > source and open standards); to say we "probably would have never even
> > had Linux" is laughable.
>
> My argument is that if the Internet were developed under GPL restrictions,
> it would never have taken off, and if the internet had not taken off, Linux
> would have never come into being, at least not anything beyond Linus's
> plaything.
>
> > > > I agree 100% with your statements above, but that is not what I was
> > > > addressing.
> > >
> > > You seemed to be saying that MS is against open protocols and open
> > > source. Well, any company is. Even so called "open" companies like
> > > Sun. They use openness as a tool to gain more marketshare, and
> > > would just as soon close everything up as soon as they gained a
> > > majority share.
> >
> > I'll belive you when I see the first non-Microsoft implementation of
> > .NET that doesn't need any reverse-engineering to function.
>
> Mind if I save this? We're going to see third party .NET implementations.
>
> > Sun has been VERY benvolent with Java, despite commercial goons'
> > attempts to smear them with FSF-like zeal. It's strange that die-hard
> > GPL fans side with Microsoft whenever Java is the topic of
> > discussion. Regardless, one only needs to see othe Sun projects to
> > find the fault in your claims: NFS, sunrpc, nis, pam, and many many
> > others.
>
> Sun has not been benevolent. They have lied to our faces. They *PROMISED*
> to make it an open standard, and they reniged on that promise, not once, but
> twice!
I'm curious... what were the lies??
>
> > What has Microsoft ever done? Their answers to the above technologies
> > are all closed/proprietary/secret designs which only function with
> > Windows. Their laughable attempt at CIFS was a complete joke, and
> > their current promises for C# are dubious (but nice, if it actually
> > turns out..).
>
> C# has already been submitted to the ECMA. It's done, in fact, 9 months
> ago. And, if CIFS is such a joke, why is SMB such a popular product?
>
> > In short: HOW ARE MICROSOFT TECHNOLOGIES BETTER THAN GPL
> > TECHNOLOGIES?
>
> I never said they were or weren't. My argument is simply about whether
> GPL'd products can be commercially viable over time. I don't think they
> can. The license prevents the common evolution of products that happens
> when companies try to make money off them.
>
> > Answer: They aren't; in fact they carry MORE baggage than their GPL
> > counterparts (Windows).
>
> I wish you would stop trying to steer the argument into something else.
>
> > > > Many (all?) TCP/IP stacks were developed to be BSD-compliant. The
> > > > internet wasn't developed on UNIX, but UNIX made it what it is today
> > > > (you're splitting hairs). Any way you look at it, Microsoft would
> > > > have done everything *worse* than it is now (see SMB, ntrpc, ActiveX
> > > > -- all communication technologies specifically designed to make people
> > > > dependent on Windows).
> > >
> > > ntrpc is DCE/RPC.
> >
> > I'm sure the Samba team would love you to explain that; they've been
> > working to reverse-engineer them for 3 years now.
>
> No, they've been reverse engineering SMB.
>
> > > In any event, My argument has nothing to do with how MS would have
> > > done it. My argument is against Alan's comment, which seems to have
> > > confused Mundie's comments on GPL with Open Source in general. His
> > > comment insinuates that The internet was built on the GPL, which it
> > > wasn't.
> >
> > Translation: Microsoft doesn't have to do anything better, I just
> > want to complain about the response the Linux community offered.
>
> My complaint is that Alan didn't think his response through, and instead
> reacted from gut instinct, which turned out to be the wrong interpretation.
>
> > > > > He's saying that 98, ME, NT, Three versions of 2000 and CE are all
> > > seperate
> > > > > forks. If they are, then Red Hat 7 is a fork, so is 7.1, so is 6.2.
> > > That's
> > > > > not the traditional definition of a fork.
> > > >
> > > > Okay, I'll bite: What is the "traditional definition" of a fork then?
> > >
> > > A fork is when you take a single code base and seperate it into two
> > > code bases with seperate development "tips". Most often, with
> > > different political and/or technical objectives.
> > >
> > > Linux kernel 2.4 is not a seperate fork from 2.2, it's a branch. A
> > > fork is when the projects diverge (as in a fork in the road). An
> > > example of forking would be BSD Light forking into OpenBSD, NetBSD,
> > > and FreeBSD. Three completely seperate projects or Emacs/XEmacs.
> > >
> > > Technically, 16 bit based Windows (9x/ME) and NT are two entirely
> > > different code bases without a common ancestor, but we can call them
> > > effective forks since they both implement the majority of the same
> > > API's.
> >
> > Fine, but it doesn't change Alan's argument a bit: You don't have
> > the luxury of forking Windows code to suit your project.
>
> That wasn't his argument. His argument was to respond to the comment about
> Linux forking by claiming that forking is not uniqe to Unix/Linux, but that
> Windows has tons of forks itself, then goes on to exagerate the problem by
> claiming that newer versions of the same tree are a fork.
--
V
------------------------------
From: Donn Miller <[EMAIL PROTECTED]>
Subject: Re: Shared library hell
Date: 5 May 2001 17:37:18 -0500
Pete Goodwin <[EMAIL PROTECTED]> wrote:
> So linux does not suffer from shared library hell?
> Symbolic linking gets round this problem, I'm told.
> However, say I have libqt.so.2.2.4 and libqt.so.2.3.0. Both would be
> symbolically linked as libqt.so.2 - so unless an application links directly
> to each individual version, it may get the wrong version.
At least under Linux you can use ldd to find out which shared libs the app is
linked to. Under Windows, even with Cygwin, I don't think there is the
equivalent of ldd to find out which shared libs the app needs. I was
frustrated by that shortcoming (and many others).
If you run ldd on the particular app in question, it might show you something
like "not found" for some of the libs, and then you know where the problem
lies.
====== 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: "Mart van de Wege" <[EMAIL PROTECTED]>
Subject: Re: The long slow slide to Microsoft.NOT
Date: Sun, 06 May 2001 00:44:49 +0200
Crossposted-To: comp.os.ms-windows.nt.advocacy,alt.linux,alt.destroy.microsoft
In article <9_XI6.22452$[EMAIL PROTECTED]>, "Erik Funkenbusch"
<[EMAIL PROTECTED]> wrote:
>
> An exploit is an exploit. Someone that allows security to lapse in one
> area over the other is simply shifting priorities. This shows that Red
> Hat and Linux in general tend to be more focused on remote exploits, to
> the detriment of local exploits. Which means, that all it it takes is
> to get a local account, any local account to gain root access.
>
> A common technique some people use is to set up a web site you must
> create an account for. Often, people will create the same account and
> password they use on their local machine. Suddenly, you have an account
> and password for the machine they logged in from, you gain root. Not a
> big deal.
>
>
I get your point. I am just getting into stuff like this, but is this not
a problem with UNIX standard security? Ie, the common problem that even
on a well protected machine, all it takes is one bug to give someone root
privileges and all hell breaks loose?
Of course when I say local exploit, I mean someone actually sitting down
at the machine to be exploited. For obvious reasons, it is very hard to
*totally* lock down a machine this way, as giving someone physical access
already implies some level of trust, whereas of course anyone accessing the
machine over a network is not to be trusted by default.
I might be very naive here, so please correct me if necessary.
Mart
--
Gimme back my steel, gimme back my nerve
Gimme back my youth for the dead man's curve
For that icy feel when you start to swerve
John Hiatt - What Do We Do Now
------------------------------
From: "Ayende Rahien" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy,alt.linux,alt.destroy.microsoft
Subject: Re: The long slow slide to Microsoft.NOT
Date: Sun, 6 May 2001 01:46:44 +0200
"Chad Everett" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> On Sat, 5 May 2001 19:54:17 +0200, Ayende Rahien <[EMAIL PROTECTED]> wrote:
> >
> >"JVercherIII" <[EMAIL PROTECTED]> wrote in message
> >news:ADVI6.297$[EMAIL PROTECTED]...
> >> Civility people! I use both Linux and Windows, and both have their
places
> >> (IMHO). I make a living right now writing VB programs so I'm kind of
> >living
> >> off the Microsoft gravy train. That being said, they do some things
which
> >> are very unpleasing. My main complaint with Microsoft is that they
stifle
> >> innovation. They never have come up with an original idea.
> >
> >Bullshit, and a big one.
> >
>
> Man, you'd better brush up on your history.
>
> >To name a few of the top of my head:
> >COM
> >COM+
>
> These are not original ideas, they came out of OMG and CORBA
COM+, most certainly not, I'm not sure about COM.
> >IE (No other browser can come even close, Mozilla can't render yahoo.com
> >properly, and crash when you try to send a bug report)
> >
>
> Talk about big BullCraps:
>
> Are you trying to tell us that IE was an original idea?
Yes. An HTML render, no. But the things it can do? Most certainly so.
Check the feature list.
------------------------------
From: "Tom Wilson" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: MS should sue the pants off linux-mandrake (was: Re: Winvocates confuse
me - d'oh!)
Date: Sat, 05 May 2001 22:51:04 GMT
"Ayende Rahien" <[EMAIL PROTECTED]> wrote in message
news:9d1eed$6db$[EMAIL PROTECTED]...
>
> "T. Max Devlin" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
> > Said Ayende Rahien in comp.os.linux.advocacy on Fri, 4 May 2001 23:25:54
> > >"T. Max Devlin" <[EMAIL PROTECTED]> wrote in message
> > >news:[EMAIL PROTECTED]...
> > >> Said Ayende Rahien in comp.os.linux.advocacy on Wed, 2 May 2001
> 20:30:02
> > >> >"Tom Wilson" <[EMAIL PROTECTED]> wrote in message
> > >> [...]
> > >> >When was Ada the next big thing?
> > >> >I agree that the syntax can use improving, but the ideas on the
basis
> of
> > >Ada
> > >> >are *very* good.
> > >>
> > >> A couple decades ago, the U.S. Government defined ADA as a
standardized
> > >> programming language. All work done for the gov't was to be done in
> > >> ADA, the 'hardware neutral' programming language, the 'next big
thing'.
> > >
> > >I know what Ada is, and I know how badly the mandate was accepted.
> > >I also know that there were many loopholes in it.
> >
> > Then you should have simply said "Ada was never 'the next big thing'",
> > if you wanted to disagree with Tom's phrasing, rather than pretending to
> > not understand it.
>
> I thought he might have more knowledge of the subject then I am.
> Considerring that he said that there was a time that Ada was the next big
> thing, and that I knew of no such time, I asked about it.
> I couldn't tell whatever he is right or wrong before knowing what he spoke
> about.
Essentially, what Max is saying is correct. ADA was almost universally
trumpeted by the government back in the eighties. It was dragged out of
academia (Where, IMHO, it belonged) and promoted as the cure-all and end-all
computer language. As with anything embraced by a bureaucracy, ADA became
over-developed. A language just can't be made to be everything to everybody.
That is where my "Next Big Thing" statement originated. A LOT of noise was
made about it back then.
I agree with your earlier assessment of ADA having some sound ideas. It just
became too unwieldy.
------------------------------
From: GreyCloud <[EMAIL PROTECTED]>
Subject: Re: The long slow slide to Microsoft.NOT
Date: Sat, 05 May 2001 15:54:16 -0700
JVercherIII wrote:
>
> That's where the interface deal and the I guess the arguable part comes
> in... COM components use that IUnknown interface and a couple other things
> (I'd have to go back through my MCSD course materials and read it again) but
> the way it seemed to be presented was that it was implementing an interface,
> which as I said can be either argued as oop or not I guess depending on your
> views. I agree that the code doesn't necessary have to use objects, but the
> question comes in when you put the interface on top of it does that become
> object oriented. (BTW this discussion got way off track, didn't it?) If you
> were to write something else implementing another interface you wrote... Hmm
> I'm not even really sure about that. Since interfaces are abstractions and
> are never really instantiated themselves (obviously) then are they really
> objects?
> Again I've seen people argue both sides of that...
>
> "Erik Funkenbusch" <[EMAIL PROTECTED]> wrote in message
> news:8qYI6.22458$[EMAIL PROTECTED]...
> > COM is really a presentation layer. It doesn't matter how they are
> > implemented, or how they're used. One side could implement them as
> classes
> > and objects, while the other used them as simple functions, or vice versa.
> >
> > "JVercherIII" <[EMAIL PROTECTED]> wrote in message
> > news:rcYI6.428$[EMAIL PROTECTED]...
> > > Perhaps I stated that incorrectly. Generally the com components are
> > classes
> > > using interfaces. Sure they can be used by non-oop programs/languages
> but
> > > they are object oriented in the face that are classes.
> > >
> > > "Erik Funkenbusch" <[EMAIL PROTECTED]> wrote in message
> > > news:07YI6.22454$[EMAIL PROTECTED]...
> > > > "JVercherIII" <[EMAIL PROTECTED]> wrote in message
> > > > news:9UXI6.404$[EMAIL PROTECTED]...
> > > > > No it is an extension - it uses OOP. I write in Visual C++, Java (a
> > > > little),
> > > > > and VB... I know what I'm talking about... The basic concept is that
> > you
> > > > are
> > > >
> > > > Just because you use VC and Java doesn't mean you know what you are
> > > talking
> > > > about, and in this case you don't know what you're talking about.
> > > >
> > > > COM is not an extension to OOP. There's no such thing as "extending"
> > OOP.
> > > > Either it is, or it isn't.
> > > >
> > > > For instance, COM is perfectly useable from C without a single object
> in
> > > > sight. Component based programming and Object based programming are
> > > > related, but not even close to being the same thing.
> > > >
> > > >
> > > >
> > > >
> > >
> > >
> >
> >
The getting off topic happens a lot in COLA. But then again, getting
derailed off topic is one of Erics modus operandi. Stick by your guns!
--
V
------------------------------
From: Craig Kelley <[EMAIL PROTECTED]>
Subject: Re: Alan Cox responds to Mundie
Date: 05 May 2001 16:54:20 -0600
"Erik Funkenbusch" <[EMAIL PROTECTED]> writes:
> "Craig Kelley" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
> > "Erik Funkenbusch" <[EMAIL PROTECTED]> writes:
> >
> > >
> > > Indeed it was. Without proprietary IP, the companies involved in
> > > growing and commercializing the internet would have never done it.
> > > The Internet was founded on completely open and public domain
> > > information, which is what allowed companies to make their own
> > > proprietary versions.
> >
> > Who said anything about growing and commercializing it? I am talking
> > about *developing* it.
>
> Guess that depends on what you mean by "developing". It would have
> still been invented, but developing means "growing".
>
> http://www.m-w.com/cgi-bin/dictionary?va=developing
[snip semantic argument]
> > > We would not have the internet as we know it, and probably would
> > > have never even had Linux, had it not been for the fact that the
> > > protocols and code were made publicly available for unrestricted
> > > use.
> >
> > I disagree. Apply natural selection to technology and you *will* end
> > up with open source under general circumstances (ie, not some niche or
> > bleeding-edge projects). The internet could not have been without
> > open source, and it will only continue on those merits.
>
> You are again confusing Open Source with GPL. My argument is only
> about the GPL, and how the Internet could not have become what it is
> if it had been released under GPL semantics.
>
> > Linux is a by-product of this, and would always have been regardless.
> > The internet succeeded where multi-billion dollar companies failed
> > (Apple's E-World, AOL's QuantumLink/AOL, Microsoft's MSN, Compuserve,
> > The Source, Delphi -- one failure after another, falling down to open
> > source and open standards); to say we "probably would have never even
> > had Linux" is laughable.
>
> My argument is that if the Internet were developed under GPL restrictions,
> it would never have taken off, and if the internet had not taken off, Linux
> would have never come into being, at least not anything beyond Linus's
> plaything.
Ahem, Linux *is* GPL software.
> > > > I agree 100% with your statements above, but that is not what I was
> > > > addressing.
> > >
> > > You seemed to be saying that MS is against open protocols and open
> > > source. Well, any company is. Even so called "open" companies like
> > > Sun. They use openness as a tool to gain more marketshare, and
> > > would just as soon close everything up as soon as they gained a
> > > majority share.
> >
> > I'll belive you when I see the first non-Microsoft implementation of
> > .NET that doesn't need any reverse-engineering to function.
>
> Mind if I save this? We're going to see third party .NET implementations.
Be my guest, as long as microsoft isn't contributing secret parts to
the implementation.
> > Sun has been VERY benvolent with Java, despite commercial goons'
> > attempts to smear them with FSF-like zeal. It's strange that die-hard
> > GPL fans side with Microsoft whenever Java is the topic of
> > discussion. Regardless, one only needs to see othe Sun projects to
> > find the fault in your claims: NFS, sunrpc, nis, pam, and many many
> > others.
>
> Sun has not been benevolent. They have lied to our faces. They *PROMISED*
> to make it an open standard, and they reniged on that promise, not once, but
> twice!
This is exactly what I was talking about.
Why don't you show this revulsion for *everything else* Microsoft has
done?
> > What has Microsoft ever done? Their answers to the above technologies
> > are all closed/proprietary/secret designs which only function with
> > Windows. Their laughable attempt at CIFS was a complete joke, and
> > their current promises for C# are dubious (but nice, if it actually
> > turns out..).
>
> C# has already been submitted to the ECMA. It's done, in fact, 9
> months ago. And, if CIFS is such a joke, why is SMB such a popular
> product?
I think you know the answer to that.
Why is SMB blocked at every boarder router, if it *isn't* such a joke?
> > In short: HOW ARE MICROSOFT TECHNOLOGIES BETTER THAN GPL
> > TECHNOLOGIES?
>
> I never said they were or weren't. My argument is simply about
> whether GPL'd products can be commercially viable over time. I
> don't think they can. The license prevents the common evolution of
> products that happens when companies try to make money off them.
So you'd rather complain than support. Typical.
> > Answer: They aren't; in fact they carry MORE baggage than their GPL
> > counterparts (Windows).
>
> I wish you would stop trying to steer the argument into something
> else.
Ditto.
> > > > Many (all?) TCP/IP stacks were developed to be BSD-compliant. The
> > > > internet wasn't developed on UNIX, but UNIX made it what it is today
> > > > (you're splitting hairs). Any way you look at it, Microsoft would
> > > > have done everything *worse* than it is now (see SMB, ntrpc, ActiveX
> > > > -- all communication technologies specifically designed to make people
> > > > dependent on Windows).
> > >
> > > ntrpc is DCE/RPC.
> >
> > I'm sure the Samba team would love you to explain that; they've been
> > working to reverse-engineer them for 3 years now.
>
> No, they've been reverse engineering SMB.
No, they've been reverse engineering ntrpc calls that NT uses for
domain control. Try to keep up here. Here's what Allison has to say
to Mundie:
Jeremy Allison Subject: Shared Source ( May 3, 2001, 17:33:32 )
Well I'm all for shared source. I wonder if
Microsoft would share some of their source to make
interoperability with Microsoft Windows 2000 domain
controllers easier for Samba ? I'm eagerly waiting
the first "shared source" release of that code. I
wouldn't even need to change any of it - just read
it :-).
Or is it all about control of intellectual
property.......
No - surely Microsoft wouldn't be so disingenuous :-)
:-). After all, look at their open and honest dealings
with the Kerberos community about their modifications
to the TGT...
Oh. Wait...... "
:-).
Regards,
Jeremy Allison,
Samba Team.
> > > In any event, My argument has nothing to do with how MS would have
> > > done it. My argument is against Alan's comment, which seems to have
> > > confused Mundie's comments on GPL with Open Source in general. His
> > > comment insinuates that The internet was built on the GPL, which it
> > > wasn't.
> >
> > Translation: Microsoft doesn't have to do anything better, I just
> > want to complain about the response the Linux community offered.
>
> My complaint is that Alan didn't think his response through, and
> instead reacted from gut instinct, which turned out to be the wrong
> interpretation.
Well you've come across as a bonefied Microsoft Cheerleader instead.
> > > A fork is when you take a single code base and seperate it into two
> > > code bases with seperate development "tips". Most often, with
> > > different political and/or technical objectives.
> > >
> > > Linux kernel 2.4 is not a seperate fork from 2.2, it's a branch. A
> > > fork is when the projects diverge (as in a fork in the road). An
> > > example of forking would be BSD Light forking into OpenBSD, NetBSD,
> > > and FreeBSD. Three completely seperate projects or Emacs/XEmacs.
> > >
> > > Technically, 16 bit based Windows (9x/ME) and NT are two entirely
> > > different code bases without a common ancestor, but we can call them
> > > effective forks since they both implement the majority of the same
> > > API's.
> >
> > Fine, but it doesn't change Alan's argument a bit: You don't have
> > the luxury of forking Windows code to suit your project.
>
> That wasn't his argument. His argument was to respond to the
> comment about Linux forking by claiming that forking is not uniqe to
> Unix/Linux, but that Windows has tons of forks itself, then goes on
> to exagerate the problem by claiming that newer versions of the same
> tree are a fork.
I'm not going to quote it again, if you can't understand English,
that's not my problem.
Anyway, my closing thoughts:
o BSD licences are superior to GPL
o GPL licenses are *far* superior to Microsoft licenses
If you want to limit the scope of this discussion to the first point,
then by all means -- but I won't participate under those constraints.
Microsoft complaining about onerous licenses is a bit too much to
stomach, really.
--
It won't be long before the CPU is a card in a slot on your ATX videoboard
Craig Kelley -- [EMAIL PROTECTED]
http://www.isu.edu/~kellcrai finger [EMAIL PROTECTED] for PGP block
------------------------------
From: "Mart van de Wege" <[EMAIL PROTECTED]>
Subject: Re: Shared library hell
Date: Sun, 06 May 2001 00:56:48 +0200
In article <8JYI6.4578$[EMAIL PROTECTED]>, "Pete Goodwin"
<[EMAIL PROTECTED]> wrote:
> So linux does not suffer from shared library hell?
>
> Symbolic linking gets round this problem, I'm told.
>
> However, say I have libqt.so.2.2.4 and libqt.so.2.3.0. Both would be
> symbolically linked as libqt.so.2 - so unless an application links
> directly to each individual version, it may get the wrong version.
>
> You would think that libqt.so.2.2.4 is older than libqt.so.2.3.0 -
> however, there's a difference. One was built with an _older_ version of
> gcc - 2.3.0! This fact alone is enough to break some applications - they
> might work with 2.3.0 but they fail due to the incompatability with gcc.
>
> So how does Linux cope with this?
>
Well Pete,
Since libqt is programmed in C++, and the different versions of gcc are
not binary-compatible when it comes to C++, you would have a point.
AFAIK, the only solution would be to recompile the library you need
yourself. Are you quite sure that both these libraries came from the same
source (distributor)? I would say that a distributor is responsible for
making sure that all C++ code in a distribution is compiled with the same
compiler, in order to maintain binary-compatibility. Again AFAIK, Debian
does this, don't know about Mandrake or SuSE.
Disclaimer: I am not a C++, nor a KDE developer, so take my words with
about a whole sea of salt please.
Mart
--
Gimme back my steel, gimme back my nerve
Gimme back my youth for the dead man's curve
For that icy feel when you start to swerve
John Hiatt - What Do We Do Now
------------------------------
** 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
******************************
