Re: proposed updateBacgroundColor() fix

From: Joaquin Cuenca Abela (e98cuenc@yahoo.com)
Date: Mon Jul 29 2002 - 11:27:34 EDT

  • Next message: Jesper Skov: "AbiWord Weekly News #102 (2002, week 29) released"

    --- "j.m.maurer" <j.m.maurer@student.utwente.nl>
    wrote:
    > Op ma 29-07-2002, om 17:07 schreef Joaquin Cuenca
    > Abela:
    > >
    > > --- "j.m.maurer" <j.m.maurer@student.utwente.nl>
    > > wrote:
    > > > AFAICT, this is just a hack and doesn't fixes
    > the
    > > > real problem I
    > > > addressed is my mail from a few minutes ago.
    > >
    > > btw, IMO your problem is a completely orthogonal
    > one
    > > to the transparent stuff.
    >
    > True, I see that now
    >
    > > Just to put some background on this problem (to
    > > everybody that has not followed all that closely
    > > enough), when a run has a "transparent" (or white)
    > > background,
    >
    > (But now you are getting to my problem again....)
    > No, a run isn't supposed to have a background or
    > transparent color, it's
    > container may have on, but a Run not

    nope. I run should have a background. If you select
    a word and you emphatize it giving it a yellow
    background, then the word will become a run with a
    yellow background. The yellow background is not a
    property of its container.

    This background is the first that we look when we want
    to draw a run. If it's transparent (or white for
    compatibility with older versions), then we look at
    our container background color.

    > > and somebody asks the run to clear itself,
    > > the run asks the page what is its background,
    >
    > The clear call should be forwarded to it's parent
    > container. The run
    > should not be allowed to do this

    And how is the container going to know the background
    color of the run? (which is important to draw the run
    on screen).

    If this color is transparent, then IMO that should be
    handled by the container (right now it's still handled
    by the run), but before trying to draw the background,
    the run should be asked.

    > and then
    > > we do a lookup which ends showing in the
    > performance
    > > stats (basically because we ask each run to clear
    > > itself a *lot* of times).
    >
    > This will indeed help a for now. Although this
    > should be rewritten
    > sometimes

    It needs some fixes, agreed. And I guess that these
    fixes will change a bit the code, but the performace
    reason Tomas was speaking about has little to know
    with who is responsible of doing what.

    Well, at least I fixed that two days ago (modulo some
    bad surprises that I can discover with more tests) and
    I didn't need any big changes

    Cheers,

    =====
    Joaquin Cuenca Abela
    e98cuenc@yahoo.com

    __________________________________________________
    Do You Yahoo!?
    Yahoo! Health - Feel better, live better
    http://health.yahoo.com



    This archive was generated by hypermail 2.1.4 : Mon Jul 29 2002 - 11:34:04 EDT