[GRASS5] Re: GRASS i18n

[Markus Neteler]

On Fri, Feb 16, 2001 at 03:34:59PM +0600, artemis wrote:
> Markus Neteler wrote:
> > Hi Artemis,
> > 
> > On Thu, Feb 15, 2001 at 04:42:05PM +0600, artemis wrote:
> > > Hi, Markus
> > > If you remember, I'm the guy stucked with i18n of GRASS.
> > > There are several difficulties that I experience now:
> > > 
> > > First, the GUI. Now, I'm updating the modules in tcltkgrass to the current
> > > state of the programs already in russian, and think it would be nice if it
> > > had something like gettext. Imagine, how easy the localization of the GUI
> > > could be. I can help with this, as soon as I finish some of the other
> > > work.
> > I agree, gettext would be the best. It's a major issue and we have scheduled
> > it for GRASS 5.1:
> > http://www.geog.uni-hannover.de/grass/grass51/
> > 
> > That time help will be appreciated. Even with implementation of gettext as I
> > am unexperienced here.
> > But maybe we could use tcltkgrass to get familiar with gettext? If you
> > have ideas how to implement this, please let me know.
> > 
> 
> I've recently seen tcl-gettext library, that I'm going to explore, 
> as I'm not familiar with gettext well as well.

Great! Probably you get some ideas. Please forward the tcl-gettext URL to
me.

> > > Second, the maps. I work in governmental organization, and we have rules
> > > for map making. E.g. when I need to make a colortable in ps.map, I need to
> > > edit the category labels, but modcats/G_edit_cats etc. doesn't allow
> > > entering 8bit characters. Perhaps, this is important in the software, as I
> > > found very many (about 30!) checks like 'getchar() & 0177' and so on.
> > > (btw, the problem is to print russian labels on the map).
> > > 
> > > So I edit the cats file by hand and everything seems to work. I even got
> > > the nice russian colortables on the final map. The query programs work
> > > too (haven't tried nviz yet)
> > > 
> > > Why are there such restrictions? When we tried to modify those at our site
> > > (commented out all the '& 0177', '< 0177' and so on in Vask) it didn't
> > > work.
> > > 
> > > The total patch (all such entries in the whole tree) didn't work too.
> > > I wonder where the code that handles the 8bit char lives then. Or is it
> > > a kind of importance to GRASS programs and we shouldn't modify that at 
> > > all?
> > > In this case, I've already started writing a category editor, because we
> > > need to label everything in russian.
> > 
> > The current font support is a mess. From my side there are no restrictions
> > to improve the situation. I think the bad coding is due to the history of
> > GRASS: The U.S. army was simply not interested in supporting other
> > languages. As GRASS is an internation project now, we have to improve the
> > situation.
> 
> The problem is not in the fonts (but it would be if we would use p.map).
> GRASS 'cyrilc' font has only 26 letters and is unusable in the case, but
> type-1 fonts work well with ps.map.
Yes. Consider to contact Alex Shevlakov
 sixote alin <sixote at yahoo.com>
he has written the GRASS 'cyrilc' font. Maybe he has some hints for you.

> But the problem is in the limitations on enetering extended chars in the
> CERL code. I suppose, it should be completely rewritten with new curses
> routines, if possible. We found about 30 checks like '*ANSIPTR<0177' and 
> so.
> As I've already told, correcting it didn't work. The problem was in the
> type 'signed char' in the gis and vask libraries and a guy at our bureau
> solved it already. He promised to post a patch.
Great, it would be nice to improve that. Most of the curses stuff is very
old in GRASS and needs to be updated.

> > Please post this on the GRASS 5 developers list. There are other developers
> > who might have an answer or recommendations for you. I definitly want to
> > support *any* font!
> > The list is here:
> > http://www.geog.uni-hannover.de/grass/grassdevel.html
> > (maybe you have already subscribed)
> > 
> > Let's discuss the topic there.
> 
> I have no ability to write to the list due to some problems with mail 
> here.
> As they are solved, I'll do.
>
[...]
>Artemis

Fine. Until then, please write to me, I'll forward to GRASS 5 list.
 
Regards

 Markus

---------------------------------------- 
If you want to unsubscribe from GRASS Development Team mailing list write to:
minordomo at geog.uni-hannover.de with
subject 'unsubscribe grass5'