Sorry for the confusion but obviously it reflects what I had in my head ;)
It was also after a day of reading documentation online.. hope I am excused.
A short intro:
I come from fortran/emacs(make it yourself stuff) and I decided to learn RAD[1]
(Actually I wanted to learn RAiD[1]): i.e. use at best the work of others.
I got fedora (which I usually trust: Not too fancy, not too strict). Fedora was: glade and
anjuta. (kilos of KDE, Qt and the like. But at the time Qt was still struggling for public licence.).
I wrote an application using anjuta and glade. I didn't had too much time to work on it. It used to be a part
time stuff. Right now it is more interesting to me for personal reasons than official ones.
Anyway here is a pic of the first result:
http://www.scriptsforscience.com/projects/loom/loom_ss.jpgSo I rephrase myself:
>you mean libglade vs gtkbuilder ? go with gtkbuilder, this will allow you
>to dynamically build your treeviews and cell renderers, and give you
>generally more access to gtk+ api via the ui file than libglade did.
>(libglade is pretty much deprecated by now and unmaintained anyway)
With portability and stability I meant it within linux. I might have to be technical.
What is the chance that (an old project):
./configure
make
will work for: the running linux distributions (least probem) over time (5 years?),
any other linux distributions over some time span
(It should work. What about a 5 years old distribtion?)
and some GTK+ libs (Over some time)?
I.e. If I write the application code and you write the UI code.
Shall I rewrite the full code every 1 year?
This is what I mean with: stability and portability.
glade writes the code, libglade uses a libglade file and gtkbuilder uses
it's own XML format. Isn't it better to write down the bare code?
Something we can deal with? or will it come next the gtk+make?
(gtk+make: the thing that makes everything within gtk and your code [problem?].)
and leave to anjuta, the playing around stuff? [anjuta has still some bugs to solve, but ..]
I.e Why not to spend the effort on anjuta?
Are we able to make a standard that will last for 10 years at least?
Sorry for the plural. But I gonna use use this environment for a while.
This application runs on linux since 10 years and is reported to be compiled and run
under win32 since 6 years.
Will it be a nice statment? or.. help me to make it that way.
Glade, glade2, glade3, libglade, bare gtk+, gtkbuilder?
I have an application and I need it to last for a while.
So .. where to?
Within GTK+ of course.
I used gnome-canvas (kind of deprecated). .. so where to?
cheers,
F
[It might be that I have to write these comments on an other mailing list: so moderator feel free.
But still. A bunch of programmers.. opinions?]
NOTES
----------------------------------------------------------------------------------------
[1] We should make a couple of changes:
RAiD might be: Rapid Application interface Development. Because we deal
with Rapid interfaces. But of course the code below is anything but rapid.
RAD: Might be for those applications were you can select a kind of icons and you can put
them together and get an application.
RAID: Should change too. Why would I put my data on a RAID using
"inexpensive" disks? I mean if I use RAID5 to protect my data
I want quality disks too. The expensive ones. ;)
I came from glade when it was writing the code
On Thu, Jan 29, 2009 at 2:51 AM, Tristan Van Berkom
<tvb gnome org> wrote:
2009/1/28 Fabio Mariotti <mariotti fabio gmail com>:
[...]
>
> The question is: Is glade3 stable enough?
My opinion obviously biased, Glade 3 (3.6) when released, should be
stable enough.
Currently I think we are crash free, with only some annoying usability bugs
(and cross project pastes of widget hierarchies with object property references
are buggy in svn, but also should be cleared up for 3.6, which is scheduled
for the GNOME 2.26 release).
> I'll need some graphical tunes: libgnome or gtkbuilder? [or what is the
> difference]
> Because I'll be working on these graphical objects..
you mean libglade vs gtkbuilder ? go with gtkbuilder, this will allow you
to dynamically build your treeviews and cell renderers, and give you
generally more access to gtk+ api via the ui file than libglade did.
(libglade is pretty much deprecated by now and unmaintained anyway).
> Gnomecanvas or drawable area?
> And I am the lonely programmer..
> long life widgets?
>
> I can rewrite the application now. But I will not do it again.
> Any suggestion?
Unfortunately I think there is still no officially selected canvas widget
for gtk+, I would stay away from gnome canvas and at the moment
would suggest you look at goocanvas for a canvas api.
Cheers,
-Tristan
