libgnomecanvas backed gnc_item_edit vs. GtkEntry

Chris Shoemaker c.shoemaker at cox.net
Fri Dec 24 07:00:48 EST 2004


On Thu, Dec 23, 2004 at 06:16:50PM -0500, Josh Sled wrote:
> 
> I believe the current desires are best captured by: "get thing working
> under gnome2 first, then later get things working right."

Ok, is there a list of MUSTFIXes for a g2 release?  I checked the
bugzilla: not much about g2.  And GNOME2_STATUS mostly says where g2
is at now, but says little about where it has to get to.  Is there an
archived email I didn't find?

GNOME2_STATUS mentions quite a handful of register-related minor bugs.
I can infer that these need to be fixed before release.  (or do they?)
It seems that you (Josh), and David are both thinking along the same
lines regarding improving the register.  But, both of you are thinking
about the "restructuring" coming after the g2 release.  Does that mean
you think these bugs will be fixed soon with only minor changes?  or
do you think g2 will be released with these bugs and then fixed by the
"restructuring" after the release?


> 
> | I'm still trying to understand the character of the G2 port, too.  I
> | get the "fix bugs" goal (e.g. cursor won't show), but would making
> | GncItemEdit backed by GtkEntry be in accord with the goals of the g2
> | port?  I'm kinda walking in blind here, so someone let me know if I'm
> | missing something.
> 
> It is fundamentally backed by a GtkEntry; the GnucashItemEdit can wrap a
> couple of gtk widgets [a combo box, for instance], but it primarily wraps
> the GtkEntry.  The reason it's a gnome canvas item is that the register
> is a gnome-canvas, which takes gnome canvas items.

I'm catching on to this.  It seems like there's one GtkEntry for the
entire sheet.  I haven't found it yet, but I expect to find some some
responsible for moving the GtkEntry to where ever the cursor is.

> I've been thinking
> about just how bad it would look/feel if the register-gnome was just "raw"
> gtk layout-managers + widgets ... but any experiments I was thinking of
> performing along those lines would be after the gnome2 port is completed.

Well, I'm not sure it would look that bad, but I think it would be
simpler to maintain/fix.

> 
> The appropriate signals and actions should be proxied to the gtk entry as
> much as possible.  It's a bit less clear to me why the cursor isn't showing,
> as it's generally unclear to me right now how the damaged-area painting is
> supposed to be handed there.

If it's unclear to you, how do you think it is for me?  :)

> 
> Derek's response is right on, though: if you can find a different solution
> that works while both reducing complexity and not regressing functionality,
> that's a good thing.

Well it seems like functionality is already regressed relative to
HEAD.  :( Replacing gnomecanvas in the register is a big deal.  (But,
IMO, probably a good idea.)  I'm sure it wouldn't be perfect the first
time.  The question I have is: are we so close to releasing g2, that
we shouldn't start something so ambitious?


In summary, I think we need a list of specific MUSTFIXes for a g2
release.  Then people like me can decide which bugs to work on.

-chris


More information about the gnucash-devel mailing list