Re: IBM Linux/390 patches

Re: IBM Linux/390 patches

[Mike Stump]

> Date: Mon, 27 Dec 1999 22:00:19 +0000 (GMT)
> In-Reply-To: < 9912230144.AA35080@marc.watson.ibm.com > from "David Edelsohn" at Dec 22, 99 08:44:06 pm
> From: Alan Cox <alan@lxorguk.ukuu.org.uk>

First, one disclaimer I don't know of any of the details, other than
what has been said here....

[ ...let me pull an rms... ]

> That suggests incompetence in the GCC steering committee.

No.

I don't know why everyone is jumping up and down.

People can work on whatever ports they want, and inform others of that
work or not at their discretion, further, they can submit (donate) the
work back to the FSF.

While it would be nice if everyone submitted things in a timely
manner, and gave advance notice so that others don't duplicate work,
this isn't required.  Just because people don't do this, doesn't mean
they should earn insults.  Also, just insult the steering committee
here?  I think you have a much enlarged view of what they do, and I
don't think their scope reaches as far so as to warrant an insult.

> Richard Henderson knew about the work being done by Linas last May
> when it was discussed a bit during Linux Expo. Linas has a right to
> be pissed off if nobody in the gcc/egcs echelons didnt think to tell
> both teams about each other

Talk about indirect.  First, neither party is obligated to inform the
other party about their existence.  So, in fact, there is no blame,
regardless of who knew what and when.  Now, if the people involved
want to be _good_ net citizens, the policy would be to inform the list
_themselves_ of their intended area of work, to read the lists, to
notice each other, and to work together to complete the work and
minimize any duplication of effort.  This is nice, and preferred, but
we don't require this, nor should a party that fails to do this suffer
insults.  Each is donating their hard earned time to contribute to the
state of the art in high performance cross platform industry standard
compilation systems.  That warrants praise and respect, their own time
is theirs to waste as they see fit.  They will be held accountable to
themselves, not us.

If someone wants to ensure their work isn't a complete waste, they
should work with the maintainers, get the work into the tree as soon
and as often as possible.  This way, cvs serves to inform others that
might be working in the same area, that someone else is already
working in that area.  If they fail to do this, and their work winds
up wasted because of duplication, then it was their choice (to not
submit), and it is their pain (waste of time if they then choose the
official version of the code).  If we do anything, it should be to
laugh, and sinker at them, privately, and learn from their mistakes
(if we are to so value them).

I have witnessed and seen evidence of this effect many times.  Has
happened in the past, and will happen in the future.  Some
duplications are tiny, and some are large.

(And all that from a person that loves a good flame war.)

Re: IBM Linux/390 patches

[linas]

It's been rumoured that Mike Stump said:
> I don't know why everyone is jumping up and down.

If the gcc Steering Committee knows of duplicate efforts, then I think
it is duty bound to steer the two efforts towards each other, and to
actively oversee & coordinate these efforts.  I'd think that the i370
and the s390 efforts are large enough (year long, multi-programmer)
to merit this kind of attention.

If this is outside of the purview of the Steering Committee, then maybe
it should change its name to more accurately reflect its duties and
activities?

It would seem that someone on the committee not only knew of both efforts
for a number of months, and did nothing about it, but also capped it off
with a series of solidly rude and insulting messages to the i370 team.

--linas

P.S. I am not enjoying the role I have inadvertently assumed.  I am
getting tired of pointing my finger and trying to sort out the various
social issues and ethical lapses.  I had nightmares last night about
this last set of email exchanges.  Please, can we move on to some sort
of more productive conversations?
                                                                

Re: IBM Linux/390 patches

[Jeffrey A Law]

In message < 199912290515.AAA02086@shadygrove.linas.org >you write:
  > It's been rumoured that Mike Stump said:
  > > I don't know why everyone is jumping up and down.
  > 
  > If the gcc Steering Committee knows of duplicate efforts, then I think
  > it is duty bound to steer the two efforts towards each other, and to
  > actively oversee & coordinate these efforts.  I'd think that the i370
  > and the s390 efforts are large enough (year long, multi-programmer)
  > to merit this kind of attention.
  > 
  > If this is outside of the purview of the Steering Committee, then maybe
  > it should change its name to more accurately reflect its duties and
  > activities?
This is out of the scope of the kinds of things the steering committee is
responsible for.

The primary responsibility of the steering committee is to deal with higher
level, usually politically sensitive issues.  This includes stuff like making
sure that no individuals or groups dominate the project, appointing release
managers, helping deal with disruptive personalities, etc.

It's job is not to coordinate efforts between developers.  That is the job
of the developers themselves.

jeff

Re: IBM Linux/390 patches

[David Edelsohn]

>>>>> linas  writes:

linas> It would seem that someone on the committee not only knew of both efforts
linas> for a number of months, and did nothing about it

	Appearances can be deceiving.  Just because people did not do what
you wanted them to do or expected them to do does not mean that they are
bad or mean or trying to harm you or undermine your efforts.  Have you
ever heard of the concepts of business confidential information and
bureaucracy?

	A lot of people expended a great deal of effort to get IBM to
release this work for the benefit of the entire community.  I find it sad
that you and Peter choose to respond by complaining in what I consider an
unconstructive and defensive manner.  There always is room for
improvement, but that's life.  Why not try to appreciate what has been
provided and look forward to what will come in the future?

linas> P.S. I am not enjoying the role I have inadvertently assumed.  I am
linas> getting tired of pointing my finger and trying to sort out the various
linas> social issues and ethical lapses.

	Then stop self-righteously pointing your finger at everyone else
and try some introspection.

Sincerely,
David

P.S. My comments express my own views and opinions, and are in no way
meant to represent IBM.
===============================================================================
David Edelsohn                                      T.J. Watson Research Center
dje@watson.ibm.com                                  P.O. Box 218
+1 914 945 4364 (TL 862)                            Yorktown Heights, NY 10598

Re: IBM Linux/390 patches

[David Edelsohn]

>>>>> linas  writes:

linas> It would seem that someone on the committee not only knew of both efforts
linas> for a number of months, and did nothing about it

	Appearances can be deceiving.  Just because people did not do what
you wanted them to do or expected them to do does not mean that they are
bad or mean or trying to harm you or undermine your efforts.  Have you
ever heard of the concepts of business confidential information and
bureaucracy?

	A lot of people expended a great deal of effort to get IBM to
release this work for the benefit of the entire community.  I find it sad
that you and Peter choose to respond by complaining in what I consider an
unconstructive and defensive manner.  There always is room for
improvement, but that's life.  Why not try to appreciate what has been
provided and look forward to what will come in the future?

linas> P.S. I am not enjoying the role I have inadvertently assumed.  I am
linas> getting tired of pointing my finger and trying to sort out the various
linas> social issues and ethical lapses.

	Then stop self-righteously pointing your finger at everyone else
and try some introspection.

Sincerely,
David

P.S. My comments express my own views and opinions, and are in no way
meant to represent IBM.
===============================================================================
David Edelsohn                                      T.J. Watson Research Center
dje@watson.ibm.com                                  P.O. Box 218
+1 914 945 4364 (TL 862)                            Yorktown Heights, NY 10598

Re: IBM Linux/390 patches

[linas]

It's been rumoured that Mike Stump said:
> I don't know why everyone is jumping up and down.

If the gcc Steering Committee knows of duplicate efforts, then I think
it is duty bound to steer the two efforts towards each other, and to
actively oversee & coordinate these efforts.  I'd think that the i370
and the s390 efforts are large enough (year long, multi-programmer)
to merit this kind of attention.

If this is outside of the purview of the Steering Committee, then maybe
it should change its name to more accurately reflect its duties and
activities?

It would seem that someone on the committee not only knew of both efforts
for a number of months, and did nothing about it, but also capped it off
with a series of solidly rude and insulting messages to the i370 team.

--linas

P.S. I am not enjoying the role I have inadvertently assumed.  I am
getting tired of pointing my finger and trying to sort out the various
social issues and ethical lapses.  I had nightmares last night about
this last set of email exchanges.  Please, can we move on to some sort
of more productive conversations?
                                                                

Re: IBM Linux/390 patches

[Mike Stump]

> Date: Mon, 27 Dec 1999 22:00:19 +0000 (GMT)
> In-Reply-To: < 9912230144.AA35080@marc.watson.ibm.com > from "David Edelsohn" at Dec 22, 99 08:44:06 pm
> From: Alan Cox <alan@lxorguk.ukuu.org.uk>

First, one disclaimer I don't know of any of the details, other than
what has been said here....

[ ...let me pull an rms... ]

> That suggests incompetence in the GCC steering committee.

No.

I don't know why everyone is jumping up and down.

People can work on whatever ports they want, and inform others of that
work or not at their discretion, further, they can submit (donate) the
work back to the FSF.

While it would be nice if everyone submitted things in a timely
manner, and gave advance notice so that others don't duplicate work,
this isn't required.  Just because people don't do this, doesn't mean
they should earn insults.  Also, just insult the steering committee
here?  I think you have a much enlarged view of what they do, and I
don't think their scope reaches as far so as to warrant an insult.

> Richard Henderson knew about the work being done by Linas last May
> when it was discussed a bit during Linux Expo. Linas has a right to
> be pissed off if nobody in the gcc/egcs echelons didnt think to tell
> both teams about each other

Talk about indirect.  First, neither party is obligated to inform the
other party about their existence.  So, in fact, there is no blame,
regardless of who knew what and when.  Now, if the people involved
want to be _good_ net citizens, the policy would be to inform the list
_themselves_ of their intended area of work, to read the lists, to
notice each other, and to work together to complete the work and
minimize any duplication of effort.  This is nice, and preferred, but
we don't require this, nor should a party that fails to do this suffer
insults.  Each is donating their hard earned time to contribute to the
state of the art in high performance cross platform industry standard
compilation systems.  That warrants praise and respect, their own time
is theirs to waste as they see fit.  They will be held accountable to
themselves, not us.

If someone wants to ensure their work isn't a complete waste, they
should work with the maintainers, get the work into the tree as soon
and as often as possible.  This way, cvs serves to inform others that
might be working in the same area, that someone else is already
working in that area.  If they fail to do this, and their work winds
up wasted because of duplication, then it was their choice (to not
submit), and it is their pain (waste of time if they then choose the
official version of the code).  If we do anything, it should be to
laugh, and sinker at them, privately, and learn from their mistakes
(if we are to so value them).

I have witnessed and seen evidence of this effect many times.  Has
happened in the past, and will happen in the future.  Some
duplications are tiny, and some are large.

(And all that from a person that loves a good flame war.)

Re: IBM Linux/390 patches

[Jeffrey A Law]

In message < 199912290515.AAA02086@shadygrove.linas.org >you write:
  > It's been rumoured that Mike Stump said:
  > > I don't know why everyone is jumping up and down.
  > 
  > If the gcc Steering Committee knows of duplicate efforts, then I think
  > it is duty bound to steer the two efforts towards each other, and to
  > actively oversee & coordinate these efforts.  I'd think that the i370
  > and the s390 efforts are large enough (year long, multi-programmer)
  > to merit this kind of attention.
  > 
  > If this is outside of the purview of the Steering Committee, then maybe
  > it should change its name to more accurately reflect its duties and
  > activities?
This is out of the scope of the kinds of things the steering committee is
responsible for.

The primary responsibility of the steering committee is to deal with higher
level, usually politically sensitive issues.  This includes stuff like making
sure that no individuals or groups dominate the project, appointing release
managers, helping deal with disruptive personalities, etc.

It's job is not to coordinate efforts between developers.  That is the job
of the developers themselves.

jeff

Re: IBM Linux/390 patches

[Linus Torvalds]

In article < E120weu-00009s-00@pet >,
Peter Schulte-Stracke <Peter.Schulte-Stracke@t-online.de> wrote:
>
>(1)    The Linux/390 patches published by IBM come with their own 
>       binutils, gcc, libc, prior and public work notwithstanding.

So? They thought they could do a better job starting from scratch.

I don't know if they did, but it's not unreasonable.

>(4)    I request the IBM personnel to abide by the rules and to submit
>       their patches to the respective maintainer, and to merge them 
>       where appropiate with our work; in particular for binutils and
>       gcc I see no reason not to do so.

"abide by the rules"?

They did. The rules you mention you made up yourself.

I don't think you realize how hard it is for a large company to do
something like what they did - getting enough blessing to release AT
ALL. 

>(5)    The disdain demonstrated by the IBM employees cannot but hurt
>       our community, where the most exacting duties are executed for
>       nothing but just the consideration which is so utterly absent
>       from their conduct.  One surely cannot expect IBM or any
>       organization like it to act spontanously according to the
>       rules of quite a differing game; they need to be told and learn
>       to do so. This is as inevitable as it is often painful.

IBM seems to have done a good job, and you're complaining?

Open source is about people being able to do their own projects. It's
about NOT having to "go through channels" and have to worry about issues
like that. It's about empowering people to do what they want.

INCLUDING the big guys. 

		Linus

Re: IBM Linux/390 patches

[Alexandre Oliva]

On Dec 22, 1999, Peter.Schulte-Stracke@t-online.de (Peter Schulte-Stracke) wrote:

> (4)    I request the IBM personnel to abide by the rules and to submit
>        their patches to the respective maintainer, and to merge them 
>        where appropiate with our work; in particular for binutils and
>        gcc I see no reason not to do so.  

Just to make something clear: even though it is usually appreciated if
someone who modifies GPLed code contributes the code back, this is not
required by the GPL.

-- 
Alexandre Oliva http://www.ic.unicamp.br/~oliva IC-Unicamp, Bra[sz]il
oliva@{lsd.ic.unicamp.br,guarana.{org,com}} aoliva@{acm,computer}.org
oliva@{gnu.org,kaffe.org,{egcs,sourceware}.cygnus.com,samba.org}
** I may forward mail about projects to mailing lists; please use them

Re: IBM Linux/390 patches

[David Edelsohn]

	I am a member of the GCC Steering Committee and I initiated
discussions about IBM's S/390 GNU toolchain improvements many months ago
with other GCC core maintainers.  You obviously do not know about that and
do not know about other details of the S/390 GNU toolchain and Linux port
that continue to evolve.

	I would appreciate if you would step down from your self-righteous
plateau and stop making condescending, insulting comments and demands
about this work and IBM's efforts without knowing the facts.  I find your
note and attitude much more contemptable than any supposed slight made to
the "community" by IBM employees.  IBM executives and employees are well
aware of the free software community and have been cooperating with the
community since its earliest beginnings.  IBM is strongly engaged with the
GNU/Linux community.

	If you wish to spread FUD and incite hostility, I think that you
will do much more to undermine GNU/Linux than any corporation that feels
threatened by this movement.

Sincerely,
David Edelsohn

P.S. My comments express my own views and opinions, and are in no way meant
to represent IBM.
===============================================================================
David Edelsohn                                      T.J. Watson Research Center
dje@watson.ibm.com                                  P.O. Box 218
+1 914 945 4364 (TL 862)                            Yorktown Heights, NY 10598

IBM Linux/390 patches

[Peter Schulte-Stracke]

(1)    The Linux/390 patches published by IBM come with their own 
       binutils, gcc, libc, prior and public work notwithstanding.

(2)    Work on a GCC for the IBM mainframes started in the 1970s as
       witnessed by the gnu-architecture string `i370', predating the
       current ESA/390 subarchitecture. We were ourselves surprised to
       learn that the ELF architecture identifier is EM_S370 = 9,
       registered in connection with Amdahl's UTS long ago.
       
(3)    When he started working on the Linux/390 project, Linas Vepstas
       ( http://linas.org/linux/i370 ) took over the machine
       dependant portions of binutils and gcc, investing much work
       into it. 
       
(4)    I request the IBM personnel to abide by the rules and to submit
       their patches to the respective maintainer, and to merge them 
       where appropiate with our work; in particular for binutils and
       gcc I see no reason not to do so.  

(5)    The disdain demonstrated by the IBM employees cannot but hurt
       our community, where the most exacting duties are executed for
       nothing but just the consideration which is so utterly absent
       from their conduct.  One surely cannot expect IBM or any
       organization like it to act spontanously according to the
       rules of quite a differing game; they need to be told and learn
       to do so. This is as inevitable as it is often painful.

       With kind regards,
	    
	    Yours, sincerly
		     
		     Peter Schulte-Stracke		     

Re: IBM Linux/390 patches

[Alan Cox]

> 	I am a member of the GCC Steering Committee and I initiated
> discussions about IBM's S/390 GNU toolchain improvements many months ago
> with other GCC core maintainers.  You obviously do not know about that and
> do not know about other details of the S/390 GNU toolchain and Linux port
> that continue to evolve.

That suggests incompetence in the GCC steering committee. Richard Henderson 
knew about the work being done by Linas last May when it was discussed a bit
during Linux Expo. Linas has a right to be pissed off if nobody in the gcc/egcs
echelons didnt think to tell both teams about each other

> the "community" by IBM employees.  IBM executives and employees are well
> aware of the free software community and have been cooperating with the
> community since its earliest beginnings.  IBM is strongly engaged with the
> GNU/Linux community.

Which IBM is this. The IBM that tried to get people fired, the IBM that tried
to stop me working on the lanstreamer driver, the IBM that helped me work
on the lanstreamer driver, the IBM that worked on the excellent raid drivers
and donated them or the IBM that tried to stop the S/390 release getting out
of IBM ?

This is aimed at everyone. IBM is _huge_, bits of IBM are on both sides of 
every war. In fact the only thing IBM departments seem to hate more than
Microsoft are each other.

Linas, please bear in mind that the folks who got the 390 code and tools out
are on our side, the ones who made it so damned difficult to get out are not.
Both sides work for IBM, but that doesn't mean IBM is good or IBM is bad, IBM
is simply too large to fit on one side of most fences.

The code is now out there, it apparently works passably and is a great basis
to build good mainframe Linux support. Don't blame the folks who wrote the code
for the duplication of effort. You had to fight lack of access, and a steep
learning curve, they had to fight ibm  8)

Alan

Re: IBM Linux/390 patches

[Alan Cox]

> 	I am a member of the GCC Steering Committee and I initiated
> discussions about IBM's S/390 GNU toolchain improvements many months ago
> with other GCC core maintainers.  You obviously do not know about that and
> do not know about other details of the S/390 GNU toolchain and Linux port
> that continue to evolve.

That suggests incompetence in the GCC steering committee. Richard Henderson 
knew about the work being done by Linas last May when it was discussed a bit
during Linux Expo. Linas has a right to be pissed off if nobody in the gcc/egcs
echelons didnt think to tell both teams about each other

> the "community" by IBM employees.  IBM executives and employees are well
> aware of the free software community and have been cooperating with the
> community since its earliest beginnings.  IBM is strongly engaged with the
> GNU/Linux community.

Which IBM is this. The IBM that tried to get people fired, the IBM that tried
to stop me working on the lanstreamer driver, the IBM that helped me work
on the lanstreamer driver, the IBM that worked on the excellent raid drivers
and donated them or the IBM that tried to stop the S/390 release getting out
of IBM ?

This is aimed at everyone. IBM is _huge_, bits of IBM are on both sides of 
every war. In fact the only thing IBM departments seem to hate more than
Microsoft are each other.

Linas, please bear in mind that the folks who got the 390 code and tools out
are on our side, the ones who made it so damned difficult to get out are not.
Both sides work for IBM, but that doesn't mean IBM is good or IBM is bad, IBM
is simply too large to fit on one side of most fences.

The code is now out there, it apparently works passably and is a great basis
to build good mainframe Linux support. Don't blame the folks who wrote the code
for the duplication of effort. You had to fight lack of access, and a steep
learning curve, they had to fight ibm  8)

Alan

Re: IBM Linux/390 patches

[Linus Torvalds]

In article < E120weu-00009s-00@pet >,
Peter Schulte-Stracke <Peter.Schulte-Stracke@t-online.de> wrote:
>
>(1)    The Linux/390 patches published by IBM come with their own 
>       binutils, gcc, libc, prior and public work notwithstanding.

So? They thought they could do a better job starting from scratch.

I don't know if they did, but it's not unreasonable.

>(4)    I request the IBM personnel to abide by the rules and to submit
>       their patches to the respective maintainer, and to merge them 
>       where appropiate with our work; in particular for binutils and
>       gcc I see no reason not to do so.

"abide by the rules"?

They did. The rules you mention you made up yourself.

I don't think you realize how hard it is for a large company to do
something like what they did - getting enough blessing to release AT
ALL. 

>(5)    The disdain demonstrated by the IBM employees cannot but hurt
>       our community, where the most exacting duties are executed for
>       nothing but just the consideration which is so utterly absent
>       from their conduct.  One surely cannot expect IBM or any
>       organization like it to act spontanously according to the
>       rules of quite a differing game; they need to be told and learn
>       to do so. This is as inevitable as it is often painful.

IBM seems to have done a good job, and you're complaining?

Open source is about people being able to do their own projects. It's
about NOT having to "go through channels" and have to worry about issues
like that. It's about empowering people to do what they want.

INCLUDING the big guys. 

		Linus

Re: IBM Linux/390 patches

[Alexandre Oliva]

On Dec 22, 1999, Peter.Schulte-Stracke@t-online.de (Peter Schulte-Stracke) wrote:

> (4)    I request the IBM personnel to abide by the rules and to submit
>        their patches to the respective maintainer, and to merge them 
>        where appropiate with our work; in particular for binutils and
>        gcc I see no reason not to do so.  

Just to make something clear: even though it is usually appreciated if
someone who modifies GPLed code contributes the code back, this is not
required by the GPL.

-- 
Alexandre Oliva http://www.ic.unicamp.br/~oliva IC-Unicamp, Bra[sz]il
oliva@{lsd.ic.unicamp.br,guarana.{org,com}} aoliva@{acm,computer}.org
oliva@{gnu.org,kaffe.org,{egcs,sourceware}.cygnus.com,samba.org}
** I may forward mail about projects to mailing lists; please use them

Re: IBM Linux/390 patches

[David Edelsohn]

	I am a member of the GCC Steering Committee and I initiated
discussions about IBM's S/390 GNU toolchain improvements many months ago
with other GCC core maintainers.  You obviously do not know about that and
do not know about other details of the S/390 GNU toolchain and Linux port
that continue to evolve.

	I would appreciate if you would step down from your self-righteous
plateau and stop making condescending, insulting comments and demands
about this work and IBM's efforts without knowing the facts.  I find your
note and attitude much more contemptable than any supposed slight made to
the "community" by IBM employees.  IBM executives and employees are well
aware of the free software community and have been cooperating with the
community since its earliest beginnings.  IBM is strongly engaged with the
GNU/Linux community.

	If you wish to spread FUD and incite hostility, I think that you
will do much more to undermine GNU/Linux than any corporation that feels
threatened by this movement.

Sincerely,
David Edelsohn

P.S. My comments express my own views and opinions, and are in no way meant
to represent IBM.
===============================================================================
David Edelsohn                                      T.J. Watson Research Center
dje@watson.ibm.com                                  P.O. Box 218
+1 914 945 4364 (TL 862)                            Yorktown Heights, NY 10598

IBM Linux/390 patches

[Peter Schulte-Stracke]

(1)    The Linux/390 patches published by IBM come with their own 
       binutils, gcc, libc, prior and public work notwithstanding.

(2)    Work on a GCC for the IBM mainframes started in the 1970s as
       witnessed by the gnu-architecture string `i370', predating the
       current ESA/390 subarchitecture. We were ourselves surprised to
       learn that the ELF architecture identifier is EM_S370 = 9,
       registered in connection with Amdahl's UTS long ago.
       
(3)    When he started working on the Linux/390 project, Linas Vepstas
       ( http://linas.org/linux/i370 ) took over the machine
       dependant portions of binutils and gcc, investing much work
       into it. 
       
(4)    I request the IBM personnel to abide by the rules and to submit
       their patches to the respective maintainer, and to merge them 
       where appropiate with our work; in particular for binutils and
       gcc I see no reason not to do so.  

(5)    The disdain demonstrated by the IBM employees cannot but hurt
       our community, where the most exacting duties are executed for
       nothing but just the consideration which is so utterly absent
       from their conduct.  One surely cannot expect IBM or any
       organization like it to act spontanously according to the
       rules of quite a differing game; they need to be told and learn
       to do so. This is as inevitable as it is often painful.

       With kind regards,
	    
	    Yours, sincerly
		     
		     Peter Schulte-Stracke