Nekochan Net

Very cool!

That was more a "I know but it's moving slower than a snail nailed to the table and I'll believe it when I see it" post.

I did run it on Debian a year ago but it was unstable because it was out of date.

Hello everybody!
I tried changing ~/.Xdefaults on my Linux box to change Mwm appereance as the topic author did; but it doesn't seems to take any effect. I use Debian GNU/Linux, and Mwm from the motif-clients package (non-free section). What is the trouble?
p.s. solved

I think it would be better to use more license-free LessTif product on FOSS systems and Window Manager bundled in it, but it's on dead side now, and hasn't any UTF-8 and other multi-byte codepages support in last version.
So still stucked on last OpenMotif.

I actually prefer OpenMotif.

What do you use for paging?

Like, multiple desktops? I don't use them nor do I need them.

Hem, you're absolutely right, as LessTif's mwm was freezed at initial state. It's draw window decors on windows that doesn't need them (for example on fltk widgets), it hasn't UTF-8 support, had badly integrated pager, some issues and not 1:1 matching look`n`feel.
Since OpenMotif's mwm can build with png and jpeg it's almost modern in iconofy.

Did you know the problem with duping words on window titles when using gtk2 applications (i hate them, but use pair) and non latin locale? I saw the same issue on *BSD world when used first version of fvwm (fvwm2 hasn't it).

Let me check tomorrow. I have NetBSD with MWM running at work since NetBSD has tons of Motif apps for me to geek on.

I am looking at Abiword. Tell me again what you mean by 'duping words'?

If you set LC_CTYPE to something non-latin but non-mulibyte (for example ru_RU.KOI8-R, ru_RU.CP1251), and set LANG to ru_RU.UTF-8 in your environment and set right font for window titles, all text written in such codepage will be duped like "PoiskPoisk vv Google" (i wrote two first russian words in translit for your understanding, but in real there are in local codepage).

I see what you mean but I have no idea how to fix it.

Will try to fix myself when find some time. Thanks for confirmation!

So back to this problem again.
Here are links to similar troubles:
-> http://www.motifzone.net/forum/general-questions/mwm-title-bug
-> http://www.motifzone.net/forum/xterm-title-some-russian-text

Problem is not rely to concret toolkit with help of the application made, but appears usually when mixing a pair of russian words visually emptied with spaces with word containing other alphabet symbols.
I've tested with Gtk, Qt and Athena widgets xterm
Duping words ISN'T appears when you set LC_CTYPE with one locale to exact application while have other different system wide LC_CTYPE locale and using fonts with codepage for first locale. (e.g. have LC_CTYPE=ru_RU.CP1251 in .xinitrc or .xsession file and running firefox with "LC_CTYPE=ru_RU.KOI8-R firefox" command and using ***-koi8-r font string part in your .Xdefaults).

There are a TONS of ways how to fix it, for example you can avoid WmICCCMToXmString() in WmWinInfo.c and use standart X set (convert XProperty to text list (actually char double-pointer) and then convert it to XmString, don't forget to free memory in time and properly and avoid double frees). But than you should go along other things, as minimized titles use another procedure and produce this bug again.
You can also try to change XmCvtCTToXmString() in Xm/TxtPropCv.c to XmStringCreateLocalized() and produce bug like this. As these titles discussed upper are converted in compound text. So bug is in XmCvtCTToXmString or producures used in it internally. I found the fast yet ugly way in logical low level (see patch). It only fix it in Xm lib and in func used for titles, so there are other ones that has this bug too, but i'm not worry, as i wanted to fix my only problem.

This one is ugly because we dunno exactly what we do, so better way will be finding mwm's locale checking and add exception for own locale.