Minutes from X.Org Architecture Call for 30 August 2004

Daniel Stone daniel at freedesktop.org
Sun Sep 5 05:07:00 PDT 2004


On Tue, Aug 31, 2004 at 09:36:21PM +0200, Roland Mainz wrote:
> Keith Packard wrote:
> > > One simple question: The option of putting XawPrintShell into a seperate
> > > library has already been _REJECTED_.
> > 
> > Yes, I know you are strongly opposed to such an architecture, but I still
> > haven't seen anything but rhetoric in support of that position.
> 
> I and others in the Xorg arch phone conference did list more than one
> non-rhethorical reason why this should be avoided. But it seems that you
> are simply ignoring that part and stamp even those technical issues as
> "rhetorics". Oh well... ;-(

Just as with myself and the rest of us, Keith doesn't have an infallible
memory; it may be valuable to clearly and *calmly* restate these
arguments in a technical and unemotional way on the list.
		
> > I assert
> > that we will be able to find a clean and straightforward way to split out
> > this functionality into a separate library. 
> 
> This may be your plan for that but I am NOT going to support that. I
> hoped I made that CLEAR in the phone conference.
> And I explicitly said that the choices for the Xorg arch board election
> should NOT include this option. *POINT*

Yes, and this is where your beliefs differ.  Both sides need to give a
little and concede some ground here -- saying 'the vote must not include
this option' is, IMO, not an option, nor do I consider it playing nicely
with others.

> > > So why are you now coming up with that issue again ? Was the whole phone
> > > conference for /dev/null ? ;-(
> > 
> > I think we all understand your position in this matter, and I want you to
> > know that I will not railroad my own ideas through the X.org release
> > process.  Given your strong objections to this architecture, I expect
> > we'll need a credible implementation that satisfies your goals along with
> > those of the rest of the community.
> 
> The only objections are coming from your position, with a little bit
> support from Redhat staff.

The good people that work at Red Hat have their own opinions, just as
you do.  They've worked on X, and they have their ideas, and you cannot
discredit them just by saying 'omg they work at rhat' and waving your
arms around muttering something about a conspiracy.

Does the fact that open source work (most of which is unrelated to
X.Org, likely) earns them their food disqualify them from having an
opinion on X?  Is getting paid for working on open source a crime?  Does
it make them more conspiritorial?  I don't see how membership of any
organisation -- whether it pays your bills or not -- automatically
qualifies you for the part of a conspiracy.

By all means, feel free to throw theories around about how the
anti-Xprint cabal fly out in their black helicopters and silence dissent
and generally get rid of Xprint, but please let's stop with this
distributor conspiracy (expletive removed).

Disclosure of bias: I work for a distributor, too.

> BTW: We are talking about a "Sample Implementation". Why has a sample
> implementation that compliciated ?

Sample is not simple. Window systems are hard, let's go shopping.

> It wasn't Mozilla. It was NS4.x's plugin architecture which was
> reimplemented into Mozilla. And we are talking about libXaw.so and NOT
> libXt.so. And we are also NOT talking about making the whole tree depend
> on Xprint - we are talking ONLY talking about making libXaw.so depend on
> libXp.so, one of the smallest X11 extension libraries around.

>From a distributor's and hacker's point of view, I still find this
dependency unacceptable.  Making it optional -- you can disable Xprint
support, at the cost of losing printing in Xaw -- is great.  If the
fools want to make their system irrelevant, stupid, and whatever else
it's been called, let them.

But for the love of god, don't make it mandatory.

> In the bump from libXaw.so.6 to libXaw.so.7 the Athena widget library
> was made depend on the SHAPE extension parts in the libXext.so library.
> And now we are in the progress to make it depend on libXp.so (for the
> logical consequence of adding printing support to libXaw.so) - which is
> NOT differently from the issue with the SHAPE extension.

It's been bumped to 8 (see below for my poor explanation of why this
mail is so badly late) already.  Which is, IMO, a poor choice, but the
decision of the architecture board which I have to accept, obviously.

> > Changes to libraries of this age should be done with extreme
> > care and respect to the hundreds of ancient and unrepairable applications
> > which use them.
> 
> Right. But the last changes to libXaw.so are only one year old - and
> therefore this argumentation really does not apply to libXaw.so.

Does that make libXaw not old?  I thought you were just saying how many
old applications used Xaw, and how deeply entrenched it was; either it
is, or isn't.

> > > The number of people who do not WANT Xprint due political reasons can be
> > > counted here with one hand.

I don't want it due to technical reasons.

What political reasons do I have?  My attitude to Xprint is
well-documented, and it's not been formed for political reasons (some of
which could be: I hate HP, I hate Sun [for the record, neither is true],
I hate European people [neither is this]).  It's been formed from
looking at Xprint and making a technical judgement, as someone who is
capable of making technical judgements.

Or does the fact I work for a distributor preclude me from making
technical judgements -- political judgements only?

> Having a history of what particular issue ? Xprint, libXp.so or turning
> the Xorg Foundation into the Keith-Packard-one-man-show ? =:-)

I really hope I'm not Keith.  I like my music taste the way it is,
thanks. :) (In all seriousness, poor point, dude.)

> > > And we are NOT talking about Xprint in general: We are talking about
> > > building libXp.so, a shared library which is build by DEFAULT in the
> > > Xorg releases since 1994 and mandatory for todays systems.
> > 
> > The same status was afforded both XIE and PEX for a long time, but is no
> > longer true as those systems never gained widespread support.  In fact,
> > Mozilla was linked against XIE for the longest time, causing vendors to
> > ship that library even though the functionality could never be used.
> 
> You are talking about PEX and XIE which are really dead since a long
> time. But Xprint isn't. And Xprint will be around for many many years,
> regardless whether you like it or not. And permanenetly annoucing it's
> "death" in your various talks won't change that either.

Death is subjective; you announcing its life won't change anything,
either.  People have to make their own judgements on Xprint's death or
not, and that's what they're doing (as 'political' choices, presumably).

> > That's what we're trying to avoid here -- an artificial dependency for
> > functionality which will never be used and which we plan on replacing in
> > the near future.
> 
> Replaced by WHAT ? By CAIRO ? Well, then you can only be kidding since
> the current design of CAIRO doesn't seem to have learned the mistakes of
> the past. It's printing code still uses the 20year old design of being
> an unidirectional wrapper for PostScript code (and PDF, too). Please
> read my reply to Owen Taylor in this thread before making such
> "annoucements" (which won't be approved by the Xorg arch board in the
> forseeable either).
> And: Xprint will VERY likely be around in the next 10+ years.

Sweet mother of god, dude.  That's really subjective.

Other predictions which belong in this basket include:
  * Latham to win the Australian election by five seats.
  * Geelong for AFL premiers this year.
  * The shiraz crop down at Blackjack will be particularly good.
  * The train tomorrow will be four minutes and twenty-three seconds
    late.
  * I will die of natural causes at seventy-four years of age.


More information about the release-wranglers mailing list