New MI maintainer

New MI maintainer

[Daniel Jacobowitz]

I'm glad to announce that GDB finally has an official MI maintainer
again: Vladimir Prus.

-- 
Daniel Jacobowitz
CodeSourcery

2008-02-14  Daniel Jacobowitz  <dan@codesourcery.com>

	* MAINTAINERS: Add Vladimir Prus as MI maintainer.

Index: gdb/MAINTAINERS
===================================================================
RCS file: /cvs/src/src/gdb/MAINTAINERS,v
retrieving revision 1.388
diff -u -p -r1.388 MAINTAINERS
--- gdb/MAINTAINERS	29 Jan 2008 19:15:36 -0000	1.388
+++ gdb/MAINTAINERS	14 Feb 2008 16:34:14 -0000
@@ -370,6 +370,7 @@ language support
   C++			Daniel Jacobowitz	dan@debian.org
   Objective C support   Adam Fedor		fedor@gnu.org
 shared libs		Kevin Buettner		kevinb@redhat.com
+MI interface		Vladimir Prus		vladimir@codesourcery.com
 
 documentation		Eli Zaretskii		eliz@gnu.org
   (including NEWS)

Re: New MI maintainer

[Nick Roberts]

Daniel Jacobowitz writes:
 > I'm glad to announce that GDB finally has an official MI maintainer
 > again: Vladimir Prus.
 > 
 > -- 
 > Daniel Jacobowitz
 > CodeSourcery

I must say I'm surprised by this appointment for several reasons.  When I
offered to act as maintainer, a year or so ago, although I had Eli's
support and other's endorsed it off-list, e.g., Alain Magloire, my offer
was declined.

I've contributed to MI for six years now while Vladimir started under three
years ago.  There are many people on this list who I would say know more about
Gdb than me, but Vladimir is not one of them.  More importantly when I pointed
out last year that his changes broke Emacs use of Gdb as it needs the CLI, his
reply was:

    ...And probably the only way to change the situation is to decide that MI
    is the future, and actively discourage use of CLI for anything, to the
    degree of immediately refusing any request mentioning CLI in relation to
    any frontend.

The process behind these decisions is not open to me but it doesn't escape my
attention that Vladimir is now part of CodeSourcery, and that CodeSourcery have
at least one contract (with Ericsson) to work on Eclipse which uses GDB/MI in
its DSF plugin.  I ask myself in whose interest this appointment is made.  It's
certainly not the GNU project or the FSF.

-- 
Nick                                           http://www.inet.net.nz/~nickrob


 > 2008-02-14  Daniel Jacobowitz  <dan@codesourcery.com>
 > 
 > 	* MAINTAINERS: Add Vladimir Prus as MI maintainer.
 > 
 > Index: gdb/MAINTAINERS
 > ===================================================================
 > RCS file: /cvs/src/src/gdb/MAINTAINERS,v
 > retrieving revision 1.388
 > diff -u -p -r1.388 MAINTAINERS
 > --- gdb/MAINTAINERS	29 Jan 2008 19:15:36 -0000	1.388
 > +++ gdb/MAINTAINERS	14 Feb 2008 16:34:14 -0000
 > @@ -370,6 +370,7 @@ language support
 >    C++			Daniel Jacobowitz	dan@debian.org
 >    Objective C support   Adam Fedor		fedor@gnu.org
 >  shared libs		Kevin Buettner		kevinb@redhat.com
 > +MI interface		Vladimir Prus		vladimir@codesourcery.com
 >  
 >  documentation		Eli Zaretskii		eliz@gnu.org
 >    (including NEWS)

Re: New MI maintainer

[Daniel Jacobowitz]

On Wed, Feb 20, 2008 at 09:18:46AM +1300, Nick Roberts wrote:
> I must say I'm surprised by this appointment for several reasons.  When I
> offered to act as maintainer, a year or so ago, although I had Eli's
> support and other's endorsed it off-list, e.g., Alain Magloire, my offer
> was declined.

Yes.  The global maintainers have been discussing the problem of MI
maintenance for at least that long.  I was hoping to leave this out of
the announcement, but the consensus of that group was that you were
not a suitable choice for a GDB maintainer.  Part of the delay was our
hope that the situation would improve.

> The process behind these decisions is not open to me but it doesn't escape my
> attention that Vladimir is now part of CodeSourcery, and that CodeSourcery have
> at least one contract (with Ericsson) to work on Eclipse which uses GDB/MI in
> its DSF plugin.  I ask myself in whose interest this appointment is made.  It's
> certainly not the GNU project or the FSF.

I expect an apology from you for this offensive statement.  I am the
only member of the GDB steering committee working at CodeSourcery, and
the SC is a group of GDB contributors, not some sort of sock puppet.
The decision was made solely based on the interests of GDB in our best
judgement.

-- 
Daniel Jacobowitz
CodeSourcery

Re: New MI maintainer

[Nick Roberts]

 > > I must say I'm surprised by this appointment for several reasons.  When I
 > > offered to act as maintainer, a year or so ago, although I had Eli's
 > > support and other's endorsed it off-list, e.g., Alain Magloire, my offer
 > > was declined.
 > 
 > Yes.  The global maintainers have been discussing the problem of MI
 > maintenance for at least that long.  I was hoping to leave this out of
 > the announcement, but the consensus of that group was that you were
 > not a suitable choice for a GDB maintainer.  Part of the delay was our
 > hope that the situation would improve.

Apart from Eli's support I had no feedback from that offer so clearly I
don't understand what that means.

 > > The process behind these decisions is not open to me but it doesn't escape
 > > my attention that Vladimir is now part of CodeSourcery, and that
 > > CodeSourcery have at least one contract (with Ericsson) to work on Eclipse
 > > which uses GDB/MI in its DSF plugin.  I ask myself in whose interest this
 > > appointment is made.  It's certainly not the GNU project or the FSF.
 > 
 > I expect an apology from you for this offensive statement.  I am the
 > only member of the GDB steering committee working at CodeSourcery, and
 > the SC is a group of GDB contributors, not some sort of sock puppet.
 > The decision was made solely based on the interests of GDB in our best
 > judgement.

I don't want to escalate the invective but I don't really see why I should
apologise for my perceptions.  If, as you say, the decision was carefully
discussed, it would not have escaped the global maintainers attention that I
was interested in the role.  Given that I have made significant contributions
to GDB/MI over a long period of time then it would then seem appropriate to at
least e-mail me off-list before the announcement to inform me of your
intention.

-- 
Nick                                           http://www.inet.net.nz/~nickrob

Re: New MI maintainer

[Daniel Jacobowitz]

On Wed, Feb 20, 2008 at 09:46:04AM +1300, Nick Roberts wrote:
> was interested in the role.  Given that I have made significant contributions
> to GDB/MI over a long period of time then it would then seem appropriate to at
> least e-mail me off-list before the announcement to inform me of your
> intention.

Yes.  I didn't handle the announcement well, and I should have avoided
the (untrue) appearance of conflict of interest by asking someone else
to make the announcement.  Joel has kindly volunteered to make
announcements in the future if this situation comes up again.

This appointment isn't supposed to be a slap in the face for anyone
not appointed.  It's supposed to get us out of bad situation: Vladimir
can now approve MI patches, including yours, so you won't have to sit
around waiting for me to look at them.  I have every confidence that
he will be better about it than I was, and that he will listen when
there's a disagreement about the right choices.  If I wasn't
confident, I would have objected to the appointment.

-- 
Daniel Jacobowitz
CodeSourcery

Re: New MI maintainer

[Bob Rossi]

On Tue, Feb 19, 2008 at 03:28:53PM -0500, Daniel Jacobowitz wrote:
> On Wed, Feb 20, 2008 at 09:18:46AM +1300, Nick Roberts wrote:
> > The process behind these decisions is not open to me but it doesn't escape my
> > attention that Vladimir is now part of CodeSourcery, and that CodeSourcery have
> > at least one contract (with Ericsson) to work on Eclipse which uses GDB/MI in
> > its DSF plugin.  I ask myself in whose interest this appointment is made.  It's
> > certainly not the GNU project or the FSF.
> 
> I expect an apology from you for this offensive statement.  I am the
> only member of the GDB steering committee working at CodeSourcery, and
> the SC is a group of GDB contributors, not some sort of sock puppet.
> The decision was made solely based on the interests of GDB in our best
> judgement.

A free and open society protects its users by making public its
actions and intents. By allowing the SC to deliberate its actions
behind closed doors gives me only a reason to doubt it. I do not
support this process, and therefor default to understanding Nicks
response. A natural consequence of such secret deliberations is
frustration and a lack of trust by the users. So, how could you expect
an apology? If I were you, I would expect more of what you just got.

I think the closed process pushes us all apart, instead of together,
which is where we should be.

Sincerely,
Bob Rossi

Re: New MI maintainer

[Richard Stallman]

    A free and open society protects its users by making public its
    actions and intents.

The GNU Project is not a government, and it does not aim to represent
the people in general, or the users.  Our goals go far beyond "making
users happy".  We campaign for your freedom to change and redistribute
your own version of a program.

That does not include your having the right to be consulted by us
about what we do in developing a program.  Because the program is free,
you are not bound by our decisions anyway.

Re: New MI maintainer

[Bob Rossi]

On Wed, Feb 20, 2008 at 03:59:25AM -0500, Richard Stallman wrote:
>     A free and open society protects its users by making public its
>     actions and intents.
> 
> The GNU Project is not a government, and it does not aim to represent
> the people in general, or the users.  Our goals go far beyond "making
> users happy".  We campaign for your freedom to change and redistribute
> your own version of a program.

OK, to that end, let me say, Thank You. I've been on the receiving end of
the free software movement for a very long time. I also happen to be
reasonably happy with what I have been given.

> That does not include your having the right to be consulted by us
> about what we do in developing a program.  Because the program is free,
> you are not bound by our decisions anyway.

I do however disagree with you here. I'm simply suggesting that when you
keep the process closed, you divide people instead of bring them
together. You have witnessed an instance of this with your current
policy.

Either way, I hope this new appointment is an improvement to GDB, and to
the MI interface. I look forward to seeing some improvements.

Bob Rossi

Re: New MI maintainer

[Richard Stallman]

    The process behind these decisions is not open to me but it
    doesn't escape my attention that Vladimir is now part of
    CodeSourcery, and that CodeSourcery have at least one contract
    (with Ericsson) to work on Eclipse which uses GDB/MI in its DSF
    plugin.

If a company is willing to pay for work on MI, that is a good thing.
I don't think we should treat that as something suspect.

However, it does sound like the SC should say something to Vladimir
about the vital importance of backward compatibility and compatibility
with other GNU packages.

Re: New MI maintainer

[Nick Roberts]

 > If a company is willing to pay for work on MI, that is a good thing.
 > I don't think we should treat that as something suspect.

That is surely a good thing, just as Red Hat paid for the initial development
of MI, but the circumstances are different here.

 > However, it does sound like the SC should say something to Vladimir
 > about the vital importance of backward compatibility and compatibility
 > with other GNU packages.

Clearly Vladimir's appointment should be seen as an endorsement of his views.
You can't really expect checks and balances to be put in place.

Anyway, I should respect the decision of the global maintainers, although I
suspect most didn't have an opinion in the matter.  It looks like I've been
outmanoeuvred and perhaps I shouldn't be such a sore loser.  I guess it's
checkmate and game over.

-- 
Nick                                           http://www.inet.net.nz/~nickrob

Re: New MI maintainer

[Eli Zaretskii]

> From: Nick Roberts <nickrob@snap.net.nz>
> Cc: drow@false.org, gdb@sourceware.org
> Date: Thu, 21 Feb 2008 08:05:12 +1300 (NZDT)
> 
> Clearly Vladimir's appointment should be seen as an endorsement of his views.

No, it is not.  GDB global maintainers are not appointed based on
their views, they are appointed based on their coding and social
skills.  Unfortunately for you, your reaction to Vladimir's
appointment puts you farther from a similar appointment, not closer.

> It looks like I've been outmanoeuvred

You make it sound like there were some kind of courtyard intrigues.
As someone who was part of the discussions, let me assure you that
nothing could be farther from the truth.  I hope at least my word is
still worth enough in your eyes to convince you that no hostile intent
or attitude were involved or expressed at any time, by anyone.

> and perhaps I shouldn't be such a sore loser.  I guess it's
> checkmate and game over.

You are not a loser, and this is not a game.  This is not some kind of
final verdict in high court without right to appeal, either.  If you
can accept friendly constructive criticism and change, there's no
reason not to invite you to become a global maintainer at some later
date; no one says that MI must be maintained by a single individual.
For all Vladimir's expertise, the experience you have gathered by
developing and maintaining the Emacs GDB UI front end is unique and
important, both to GDB and to Emacs.  We will not be willing to give
up your experience and knowledge lightly.

Thanks, and keep up the good work.

Re: New MI maintainer

[Nick Roberts]

 > > Clearly Vladimir's appointment should be seen as an endorsement of his
 > > views.
 > 
 > No, it is not.  GDB global maintainers are not appointed based on
 > their views, they are appointed based on their coding and social
 > skills.  Unfortunately for you, your reaction to Vladimir's
 > appointment puts you farther from a similar appointment, not closer.

Hopefully their views are taken into account.  It doesn't sound like
I was very close, in the first place.

Having refused to revert changes that broke Emacs use of Gdb (albeit with
"undocumented assumptions") for the last three months, I find it disingenuous
for him to suddenly offer to fix parts now.  I'm not optimistic about the
future, when the heat has died down.

 >... 
 > > and perhaps I shouldn't be such a sore loser.  I guess it's
 > > checkmate and game over.
 > 
 > You are not a loser, and this is not a game.  This is not some kind of
 > final verdict in high court without right to appeal, either.  If you
 > can accept friendly constructive criticism and change, there's no
 > reason not to invite you to become a global maintainer at some later
 > date;

Thanks for your support but after six years of contributing and peer reviewing
patches the thought that "if I keep trying, I too might one day become a
maintainer" doesn't really excite me.

-- 
Nick                                           http://www.inet.net.nz/~nickrob

Re: New MI maintainer

[Joel Brobecker]

> Having refused to revert changes that broke Emacs use of Gdb (albeit with
> "undocumented assumptions") for the last three months, I find it disingenuous
> for him to suddenly offer to fix parts now.  I'm not optimistic about the
> future, when the heat has died down.

For the record, the reason why Vladimir offered to help is because
after thanking him for spotting the testsuite failures, I told him that
if you weren't for any reason able to take care of this soon, either
he or I should just finish the work. We do this for two reasons: because
we care about emacs, and because we'd like to get started on 6.8.

-- 
Joel

Re: New MI maintainer

[Frank Ch. Eigler]

Nick Roberts <nickrob@snap.net.nz> writes:

> [...]  Thanks for your support but after six years of contributing
> and peer reviewing patches the thought that "if I keep trying, I too
> might one day become a maintainer" doesn't really excite me.

Please don't be disappointed by mere politics.  If your technical
contributions continue, and are robustly constructed and justified,
any reasonable maintainer will have no choice but to go along.  A
powerful contributor can outproduce a maintainer, and in the end
that's what matters.


- FChE

Re: New MI maintainer

[Nick Roberts]

 > Please don't be disappointed by mere politics.  If your technical
 > contributions continue, and are robustly constructed and justified,
 > any reasonable maintainer will have no choice but to go along.  A
 > powerful contributor can outproduce a maintainer, and in the end
 > that's what matters.

Thanks but I'm probably not really a powerful contributor.  I just wanted to be
able to commit patches in a timely fashion and somehow thought if I put in the
hard yards that one day I would be able to do so.

-- 
Nick                                           http://www.inet.net.nz/~nickrob

Re: New MI maintainer

[Thomas Dineen]

Gentle People:

   Yes but with the conditions listed below:

   "robustly constructed and justified"

   any contributor can easily be blown off for
any political reason!

   From my observations the attitudes expressed
in this thread are NOT positive for the Open Source
Movement. I would suggest that your movement would
benefit from more openness and democracy.

Thomas Dineen



Nick Roberts wrote:

> > Please don't be disappointed by mere politics.  If your technical
> > contributions continue, and are robustly constructed and justified,
> > any reasonable maintainer will have no choice but to go along.  A
> > powerful contributor can outproduce a maintainer, and in the end
> > that's what matters.
>
>Thanks but I'm probably not really a powerful contributor.  I just wanted to be
>able to commit patches in a timely fashion and somehow thought if I put in the
>hard yards that one day I would be able to do so.
>
>  
>

Re: New MI maintainer

[Eli Zaretskii]

> Date: Fri, 22 Feb 2008 09:20:56 -0800
> From: Thomas Dineen <tdineen@ix.netcom.com>
> CC: "Frank Ch. Eigler" <fche@redhat.com>,  gdb@sourceware.org
> 
> Gentle People:
> 
>    Yes but with the conditions listed below:
> 
>    "robustly constructed and justified"
> 
>    any contributor can easily be blown off for
> any political reason!

While theoretically possible, I'm quite sure such conspiracies can
never happen in practice in this case, since the people involved come
from several different continents, cultural backgrounds, and political
views.  In fact, just reaching an agreement is sometimes a formidable
job for us.

>    From my observations the attitudes expressed
> in this thread are NOT positive for the Open Source
> Movement. I would suggest that your movement would
> benefit from more openness and democracy.

Sorry, but that's pure demagoguery.  "Open Source" and "Free Software"
mean that the sources are available.  They don't mean that every
aspect of human life is open to the public.  Even the most open
democracies always have closed deliberations about certain issues.
You cannot lead a group in any significant human endeavor without a
certain distance between the leadership and the rest, and without
closed deliberations about some sensitive issues.

Re: New MI maintainer

[Christopher Faylor]

On Fri, Feb 22, 2008 at 10:22:41PM +0200, Eli Zaretskii wrote:
>> Date: Fri, 22 Feb 2008 09:20:56 -0800
>> From: Thomas Dineen <tdineen@ix.netcom.com>
>> CC: "Frank Ch. Eigler" <fche@redhat.com>,  gdb@sourceware.org
>> 
>> Gentle People:
>> 
>>    Yes but with the conditions listed below:
>> 
>>    "robustly constructed and justified"
>> 
>>    any contributor can easily be blown off for
>> any political reason!
>
>While theoretically possible, I'm quite sure such conspiracies can
>never happen in practice in this case, since the people involved come
>from several different continents, cultural backgrounds, and political
>views.  In fact, just reaching an agreement is sometimes a formidable
>job for us.

Yes! What seems to be getting lost is the fact that all of the
development of gdb IS *very* open.  When someone submits a patch,
everyone gets to see it and everyone gets to see any technical
objections.  How long would the project last if there were hidden
agendas behind every communication?

(Well, ok, experience has shown that it could last several years
like that but still, it's not like that *now*)

>>From my observations the attitudes expressed in this thread are NOT
>>positive for the Open Source Movement.  I would suggest that your
>>movement would benefit from more openness and democracy.
>
>Sorry, but that's pure demagoguery.  "Open Source" and "Free Software"
>mean that the sources are available.  They don't mean that every aspect
>of human life is open to the public.  Even the most open democracies
>always have closed deliberations about certain issues.  You cannot lead
>a group in any significant human endeavor without a certain distance
>between the leadership and the rest, and without closed deliberations
>about some sensitive issues.

Amen.

It's interesting how often people try to tack on bigger concerns to the
simple concept of Free Software.  Free Software isn't supposed to be
solving global warming and it isn't supposed to be a demonstration of a
New World Order with feel-good cum ba yah.  It's just a guarantee that
you get the source code for the software that you're using.

Managing any project where people are involved means that sometimes the
people in charge have to have frank, private conversations.  The
alternative, as Stan Shebs, notes is to essentially do performance
reviews in the open.

Some projects *do* work that way but they are hardly bastions of
civilized discourse.

cgf

Re: New MI maintainer

[Bob Rossi]

On Fri, Feb 22, 2008 at 05:09:27PM -0500, Christopher Faylor wrote:
> On Fri, Feb 22, 2008 at 10:22:41PM +0200, Eli Zaretskii wrote:
> >Sorry, but that's pure demagoguery.  "Open Source" and "Free Software"
> >mean that the sources are available.  They don't mean that every aspect
> >of human life is open to the public.  Even the most open democracies
> >always have closed deliberations about certain issues.  You cannot lead
> >a group in any significant human endeavor without a certain distance
> >between the leadership and the rest, and without closed deliberations
> >about some sensitive issues.
> 
> Amen.
> 
> It's interesting how often people try to tack on bigger concerns to the
> simple concept of Free Software.  Free Software isn't supposed to be
> solving global warming and it isn't supposed to be a demonstration of a
> New World Order with feel-good cum ba yah.  It's just a guarantee that
> you get the source code for the software that you're using.
> 
> Managing any project where people are involved means that sometimes the
> people in charge have to have frank, private conversations.  The
> alternative, as Stan Shebs, notes is to essentially do performance
> reviews in the open.
> 
> Some projects *do* work that way but they are hardly bastions of
> civilized discourse.

This rhetoric annoys me. I was bringing up a sincere concern that I
have. I suggested a more open model because I've been wondering for 
well over 5 years why it (used to?) takes 6 months to get a code review.

Because the time to get a review was so long, I stopped bothering to submit
patches. I remember that I planned my rehersal dinner for my
wedding faster than I could get a code review, and thought that was
wierd.

I hope things are better now that Vladimar has taken his position. If
so, I'll be joining you singing, cum ba yah.

Bob Rossi

RE: New MI maintainer

[Dave Korn]

On 22 February 2008 22:21, Bob Rossi wrote:

> This rhetoric annoys me. I was bringing up a sincere concern that I
> have. I suggested a more open model because I've been wondering for
> well over 5 years why it (used to?) takes 6 months to get a code review.

  Because there are very few of us here and we all have day jobs.  Not because
everyone's arranged behind the scenes to cut you out.


  For my part, I simply cannot understand where this entire conversation is
coming from.  It seems to be predicated on the assumption that a
maintainership is some kind of position of prestige, status and power - as
opposed to the onerous duty and ongoing burden only willingly accepted out of
a sense of debt and responsibility that it actually represents.

    cheers,
      DaveK
-- 
Can't think of a witty .sigline today....

RE: New MI maintainer

[Alpár Jüttner]

I think, the real issue is not the openness of the project management,
but this:

> Given that I have made significant contributions
> to GDB/MI over a long period of time then it would then seem appropriate to at
> least e-mail me off-list before the announcement to inform me of your
> intention.

As far as I see, nobody has commented on this sentence of Nick. I think
it would be worth doing that, at least before blaming him and saying
that a global maintainer must have a higher level of ability of being
able to avoid heated personal debates.

As a complete outsider, my feeling is that if Nick had really been
considered as a valuable and important member of the community rather
than someone who should be "outmaneuvered", then exactly the same
decision would have been communicated toward him in a much more friendly
way.

Although Nick's responses in this thread are offensive and his
statements may even be false, I understand his bad feeling and I must
say it - at least partially - justifies his reactions.

Some people expected an apology in this thread but nobody seemed to be
thinking about whether he should apologize. More than one person made a
mistake in this awkward story. I strongly believe that a frank 'sorry'
from the one who made the first one would be followed by others.

After all, this is a great community of great people. With no exception.

All the best,
Alpar

Re: New MI maintainer

[Jim Blandy]

On Fri, Feb 22, 2008 at 4:01 PM, Alpár Jüttner <alpar@cs.elte.hu> wrote:
>  Although Nick's responses in this thread are offensive and his
>  statements may even be false, I understand his bad feeling and I must
>  say it - at least partially - justifies his reactions.

I'm comfortable with the decision we made, and how we arrived at it,
but I do regret not explaining to Nick the reasons we decided not to
invite him to evaluate patches for MI.  At the time, explaining seemed
unlikely to be constructive, but perhaps I was also trying to avoid a
possibly unpleasant conversation.  Explaining the decision would have
at least allowed Nick to know for a fact what had happened, instead of
being forced to guess, and tempted to guess the worst.

In situations where someone is expecting a response, and the group
decides not to invite them to take on a responsibility, I think the
group ought to explain their decision to the person in private, even
when doing so is uncomfortable.

RE: New MI maintainer

[Nick Roberts]

 > Some people expected an apology in this thread but nobody seemed to be
 > thinking about whether he should apologize. More than one person made a
 > mistake in this awkward story. I strongly believe that a frank 'sorry'
 > from the one who made the first one would be followed by others.

Thanks.  I don't really want an apology from anyone (just as I would rather
not apologise) but I appreciate you pointing out other contributory factors.

I now know more than I did at the start of the thread and the only thing I
would add is that I think such appointments are announced by a member of the
GDB Steering Committee who are "not generally involved in day-to-day
development" (not the release manager) starting with something like:

"After careful consideration the GDB Steering Committee are glad to announce..."

to make it clear that the appointment has been made collectively.

I'm starting to feel uncomfortable with this thread now so I'd like it to
stop (just as the majority of silent subscribers probably would too).

One chapter ends and a new one begins.

-- 
Nick                                           http://www.inet.net.nz/~nickrob

Re: New MI maintainer

[Eli Zaretskii]

> From: =?ISO-8859-1?Q?Alp=E1r_J=FCttner?= <alpar@cs.elte.hu>
> Cc: gdb@sourceware.org
> Date: Sat, 23 Feb 2008 01:01:16 +0100
> 
> I think, the real issue is not the openness of the project management,
> but this:
> 
> > Given that I have made significant contributions
> > to GDB/MI over a long period of time then it would then seem appropriate to at
> > least e-mail me off-list before the announcement to inform me of your
> > intention.
> 
> As far as I see, nobody has commented on this sentence of Nick.

Actually, that's not true.  Daniel did respond:

> > Date: Wed, 20 Feb 2008 20:41:47 -0500
> > From: Daniel Jacobowitz <drow@false.org>
> > Cc: gdb@sourceware.org, Richard Stallman <rms@gnu.org>
> > 
> > On Wed, Feb 20, 2008 at 09:46:04AM +1300, Nick Roberts wrote:
> > > was interested in the role.  Given that I have made significant contributions
> > > to GDB/MI over a long period of time then it would then seem appropriate to at
> > > least e-mail me off-list before the announcement to inform me of your
> > > intention.
> > 
> > Yes.  I didn't handle the announcement well, and I should have avoided
> > the (untrue) appearance of conflict of interest by asking someone else
> > to make the announcement.

I think that's an appropriate response.

> Some people expected an apology in this thread but nobody seemed to be
> thinking about whether he should apologize. More than one person made a
> mistake in this awkward story. I strongly believe that a frank 'sorry'
> from the one who made the first one would be followed by others.

I think the above comes as close to "sorry" as one can expect in such
situations.

Re: New MI maintainer

[Eli Zaretskii]

> Date: Fri, 22 Feb 2008 17:21:24 -0500
> From: Bob Rossi <bob_rossi@cox.net>
> 
> > It's interesting how often people try to tack on bigger concerns to the
> > simple concept of Free Software.  Free Software isn't supposed to be
> > solving global warming and it isn't supposed to be a demonstration of a
> > New World Order with feel-good cum ba yah.  It's just a guarantee that
> > you get the source code for the software that you're using.
> > 
> > Managing any project where people are involved means that sometimes the
> > people in charge have to have frank, private conversations.  The
> > alternative, as Stan Shebs, notes is to essentially do performance
> > reviews in the open.
> > 
> > Some projects *do* work that way but they are hardly bastions of
> > civilized discourse.
> 
> This rhetoric annoys me. I was bringing up a sincere concern that I
> have. I suggested a more open model because I've been wondering for 
> well over 5 years why it (used to?) takes 6 months to get a code review.

The long review time has nothing to do with the issue at hand, or with
the above rhetoric.

Re: New MI maintainer

[Jim Blandy]

On Fri, Feb 22, 2008 at 9:20 AM, Thomas Dineen <tdineen@ix.netcom.com> wrote:
>    From my observations the attitudes expressed
>  in this thread are NOT positive for the Open Source
>  Movement. I would suggest that your movement would
>  benefit from more openness and democracy.

The procedure for making decisions that GDB has followed here is
common among successful Open Source projects, with minor variations,
and has been for a long time.  For example, the Subversion project,
which many consider to be well-run, operates in exactly this way: the
group of "full committers" watches out for people to invite to become
full committers, discusses possibilities and reaches decisions in
private, makes invititations privately, and announces accepted
invitations publicly.  See also the book _Producing Open Source
Software_, available on-line at http://producingoss.com/.

In Free software, your ultimate protection from demagogues comes from
the license itself: if a project is badly run but the software is
valued, people who can run the project better will fork it, and
compete for developers with the badly-run project.  If the original
project was, in fact, poorly run, it will eventually die off.  In some
cases, the old project gets re-absorbed into the new one, thus
effecting a change in leadership without coercion.

Re: New MI maintainer

[Stan Shebs]

Nick Roberts wrote:
> The process behind these decisions is not open to me but it doesn't escape my
> attention that Vladimir is now part of CodeSourcery, and that CodeSourcery have
> at least one contract (with Ericsson) to work on Eclipse which uses GDB/MI in
> its DSF plugin.  I ask myself in whose interest this appointment is made.  It's
> certainly not the GNU project or the FSF.
>   
One of the qualities we look for in a maintainer is the ability to keep 
a level head, and to tone down the flames rather than escalate when the 
situation is difficult. As we know from bitter experience, technical 
ability and knowledge is not sufficient. Nick, this kind of accusation 
is precisely the sort of thing that we want to keep out of the GDB 
development process - if there were fencesitters on the issue before 
now, do you think that your public insinuation of bad faith and 
incompetence is going to cause them to regard you more favorably? So 
while I don't myself expect an apology, you should think more about your 
goals and how you are most likely to achieve them.

As for the decision process not being open, I note that almost nobody 
volunteers to have their job performance reviews and salary history be 
put out  in public. But if a wouldbe maintainer were to do that, I 
suggest that the SC have the discussion about that person in public as well.

Stan

Re: New MI maintainer

[Nick Roberts]

 > One of the qualities we look for in a maintainer is the ability to keep 
 > a level head, and to tone down the flames rather than escalate when the 
 > situation is difficult. As we know from bitter experience, technical 
 > ability and knowledge is not sufficient. Nick, this kind of accusation 
 > is precisely the sort of thing that we want to keep out of the GDB 
 > development process - if there were fencesitters on the issue before 
 > now, do you think that your public insinuation of bad faith and 
 > incompetence is going to cause them to regard you more favorably? 

I have certainly done myself no favours and did not intend to so.  However, to
not allow the suggestion of a conflict of interest would certainly provide
cover for those who wish to abuse positions of trust.  Others, can presumably
decide for themselves whether those suggestions have validity or not.

 > while I don't myself expect an apology, you should think more about your 
 > goals and how you are most likely to achieve them.

It looks like I won't be able to achieve them so I should probably set
new goals.

 > As for the decision process not being open, I note that almost nobody 
 > volunteers to have their job performance reviews and salary history be 
 > put out  in public. But if a wouldbe maintainer were to do that, I 
 > suggest that the SC have the discussion about that person in public as well.

No, but for Daniel alone, to just announce out of the blue that Vladimir is
now MI maintainer, you must admit, comes over as a bit of a stitch up.

-- 
Nick                                           http://www.inet.net.nz/~nickrob

Re: New MI maintainer

[Bob Rossi]

On Tue, Feb 19, 2008 at 02:12:22PM -0500, Daniel Jacobowitz wrote:
> I'm glad to announce that GDB finally has an official MI maintainer
> again: Vladimir Prus.

Congratulations Vladmir.

Could you tell us if you have any goals as the official MI maintainer?

Thanks,
Bob Rossi

Re: New MI maintainer

[Vladimir Prus]

[-- Attachment #1: Type: text/plain, Size: 458 bytes --]

Bob Rossi wrote:

> On Tue, Feb 19, 2008 at 02:12:22PM -0500, Daniel Jacobowitz wrote:
>> I'm glad to announce that GDB finally has an official MI maintainer
>> again: Vladimir Prus.
> 
> Congratulations Vladmir.

Thanks.

> Could you tell us if you have any goals as the official MI maintainer?

Attached is my current list of MI problems that I find most important
to resolve. The item most close to completion is custom display of
varobjs.

- Volodya





[-- Attachment #2: MI Development(2).txt --]
[-- Type: text/plain, Size: 2283 bytes --]

 

MI Development

The GDB/MI interface as it is today has the following important problems.

Variable objects and RTTI
For C++, variable objects are not able to look at the real type of the object.
Only the static type is shown. We should be able to implement display of
real type, using Apple's branch as reference.


CLI commands bypass MI
When a CLI command is issued, we don't see "^running". As result, frontend can easily think 
that gdb is waiting for commands while inferior is running.

Variable objects and scopes

Variable objects don't care much about C++ scopes. For example, it's not possible to
create a variable object for a given expression in particular scope, which makes it impossible
to accurately implement variable tooltips. Also, it's not possible to list all local variables in the
entire function, which requires extraordinary effort to display all local variables as the enter
scope and leave scope.


Change notification style
Generally, all commands that a frontend might want to issue at each step -- list of breakpoints,
list of threads, list frames, list of local variables -- should have a notification to match, which is
emitted whenever gdb thinks the result of those command can possibly change. Wit this,
the frontend won't have to automatically issue any commands after step -- if anything changes,
gdb will report that itself.


Static fields

The way -var-list-children displays always displays static field is a bit annoying, it can also
result in nice never ending trees, like in https://bugs.eclipse.org/bugs/show_bug.cgi?id=136627 
Need to make it customizable.


Custom display
For many kinds of objects there's some more reasonable representation than pure varobj tree.
For example -- for QString we need the real string. For vector we need content of the vector
to be the children.  This requires some scripting level, that can customize creation of children
of variable object and display as string.  In order for -var-update to work, that customization
level should actually be able to construct arbitrary value and use that value for a variable object.
For -var-update together with vector, we should be able to check if the vector size/storage
has changed. I'm not sure how we'd create new children if a vector is resized.