Re: BOGL based frontend for cdebconf

2000-12-20 Thread Michael Sobolev

On Wed, Dec 20, 2000 at 05:42:46PM +0100, Tollef Fog Heen wrote:
 Something like curses?  (I couldn't find bterm either - are you
 talking about jfbterm?)
no.  check utilities/bogl directory in boot-floppies source tree.

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Debian Boot CVS: aph

2000-12-07 Thread Michael Sobolev

On Thu, Dec 07, 2000 at 06:54:33PM +0200, Tapio Lehtonen wrote:
 On Wed, Dec 06, 2000 at 09:02:25PM -0800, Debian Boot CVS Master wrote:
  more missing \n's
 
 Please tell me the minimal build I can execute to get error messages
 from missing or extra \n's. I really would like to commit only stuff
 that the compiler accepts.
You may just execute `make check-fi' in utilities/dbootstrap/po directory.

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Debian Boot CVS: mss

2000-12-04 Thread Michael Sobolev

On Mon, Dec 04, 2000 at 05:06:41AM -0800, Adam Di Carlo wrote:
 Modified files:
   utilities/dbootstrap/langs: ver2.py 
 
 Log message:
 trying to make it independent of the python-xml package version
Please check if it works for you. :)  If it does not, please send the versions
of python-xml and python-base packages.

Thanks,

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Debian Boot CVS: gleydson

2000-12-04 Thread Michael Sobolev

On Mon, Dec 04, 2000 at 11:07:33AM -0500, Adam Di Carlo wrote:
  On Sun, Dec 03, 2000 at 08:31:14AM -0400, Gleydson Mazioli da Silva wrote:
   I've added the language code before of each language option (so it look
   less confuse). A bit example:
   en - You have chosen English. Press Enter to proceed
  
  Hmm...  I would not say it looks less confuse.  To my mind, people are just looking
  for something they understand, and those phrases in their native language should
  be sufficient enough, no?
 
 Well, I like the new scheme better.  But let me re-run it and see what
 I think again.
Hmm...  I just do not understand what is the target audience.

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Debian Boot CVS: gleydson

2000-12-03 Thread Michael Sobolev

On Sat, Dec 02, 2000 at 05:28:25PM -0800, Adam Di Carlo wrote:
 Module name:  boot-floppies
 Changes by:   gleydson00/12/02 17:28:25
 
 Log message:
 Preceeded message selection with it's language code (cosmetic).
Hmm...  I do not quite understand the reasoning behind this change.  Could
you please explain? :)

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Debian Boot CVS: gleydson

2000-12-03 Thread Michael Sobolev

On Sun, Dec 03, 2000 at 08:31:14AM -0400, Gleydson Mazioli da Silva wrote:
 I've added the language code before of each language option (so it look
 less confuse). A bit example:
 en - You have chosen English. Press Enter to proceed

Hmm...  I would not say it looks less confuse.  To my mind, people are just looking
for something they understand, and those phrases in their native language should
be sufficient enough, no?

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: problems comping langs.c in dbootstrap/lang

2000-11-30 Thread Michael Sobolev

On Thu, Nov 30, 2000 at 08:57:33PM +0100, Othmar Pasteka wrote:
 hi,
 
 i tried to do a make langs.c and got the following and yes, i
 installed everything which is mentioned in the README:
 gcc -c -Wall -g -I/usr/include/python1.5 iconv.c
 iconv.c: In function `do_convert':
 iconv.c:72: warning: passing arg 2 of `iconv' from incompatible
 pointer type
 gcc -shared -o iconv.so iconv.o -lpython1.5
 ./ver2.py 0 i386 langs.xml
 Traceback (innermost last):
   File "./ver2.py", line 25, in ?
 from xml.dom import core
 ImportError: cannot import name core
Well, this means that you are building boot floppies on a computer with
python-xml version 0.5.1.  I have created a version of the script for
python-xml version 0.6, but I do not know what is the correct way to
incorporate it.

Do you use woody or potato?

--
Misha

PS  I will also check what this gcc warning is about...


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: problems comping langs.c in dbootstrap/lang

2000-11-30 Thread Michael Sobolev

On Thu, Nov 30, 2000 at 09:44:37PM +0100, Othmar Pasteka wrote:
 On Thu, Nov 30, 2000 at 11:30:33PM +0300, Michael Sobolev wrote:
  Well, this means that you are building boot floppies on a computer with
  python-xml version 0.5.1.  I have created a version of the script for
  python-xml version 0.6, but I do not know what is the correct way to
  incorporate it.
 
 i have 0.6.2-2 installed.
yes, that's exactly the problem.  well, i'll try to see what i can do here...

  Do you use woody or potato?
 
 woody.
thank you.  i believe that potato comes with 0.5.1

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: problems comping langs.c in dbootstrap/lang

2000-11-30 Thread Michael Sobolev

On Thu, Nov 30, 2000 at 10:07:14PM +0100, Othmar Pasteka wrote:
  thank you.  i believe that potato comes with 0.5.1
 
 0.5.1-5 ;).
it should be working with this version. :)

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: building arm boot floppies - python error in dbootstrap/lang

2000-11-29 Thread Michael Sobolev

On Tue, Nov 28, 2000 at 09:28:45PM -0500, Adam Di Carlo wrote:
./ver2.py 0 i386 langs.xml
   (I'm suspicous about that i386, but replacing it with 'arm' doesn't
   affect the error and it does seem to have been deliberately overridden- what
   is this about?) The result is;
 
  Well, this i386 is about the contents of *.src (*.xml) files.  So this needs
  only to be changed in case this set of boot floppies is being built for arm
  platform. :)  Actually, it's something that should be changed as soon as
  something reasonable appears in *.src files...
 
 It probably shouldn't be hard-coding the architecture this way, if
 that is what it is doing.
I hope I already (after this very report) changed that. :)

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Debian Boot CVS: mss

2000-11-29 Thread Michael Sobolev

Please check if it's clear enough.

--
Misha

On Wed, Nov 29, 2000 at 06:23:39AM -0800, Adam Di Carlo wrote:
 Modified files:
   .  : README-Translators 
 
 Log message:
 added prelimiary stuff about language chooser


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Debian Boot CVS: mss

2000-11-29 Thread Michael Sobolev

Please check the file before committing:

make check-lang

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: building arm boot floppies - python error in dbootstrap/lang

2000-11-24 Thread Michael Sobolev

On Fri, Nov 24, 2000 at 06:18:01PM +, Wookey wrote:
 Having updated from cvs I am now past the LC problems reported. And then I
 got some python complaints, but after installing task-python-dev (except for
 2 bits (gdb and python-gimp) which aren't compiled for potato arm) I get a
 bit further. Now, when I do make release - When it gets to
 boot-floppies/utilities/dbootstrap/langs dir it compiles iconv.so OK and then
 does:
 
  ./ver2.py 0 i386 langs.xml
 (I'm suspicous about that i386, but replacing it with 'arm' doesn't
 affect the error and it does seem to have been deliberately overridden- what
 is this about?) The result is;
Well, this i386 is about the contents of *.src (*.xml) files.  So this needs
only to be changed in case this set of boot floppies is being built for arm
platform. :)  Actually, it's something that should be changed as soon as
something reasonable appears in *.src files...

 Traceback (innermost last):
   File "./ver2.py", line 30, in ?
 from dumper import dumper
   File "./dumper.py" line 23, in ?
 from iconv import engine
 ImportError: ./iconv.so: R_ARM_PC24 relocation out of range
Hmm..  This error seems to be specific to arm platform.

 This may be an arm-specific problem (I've asked a guru about that), but maybe
 not.
Unfortunately, I am not a debian developer, so I do not have access to those
arm machines available for them (do they exist?).  Maybe if I had an account on
such a computer, I could find what's wrong.  But, it's only needed if we want
lc on arm platform. :)

 What is actually going on in this dir?
The main thing is creating a langs.c file.
 Can I in fact do this part of the compile on an i386 box in order to get
 going? (I tried but got another error, probably due to mixing arm and i386
 files - sigh).
I'd say so.  What version of debian you are using, BTW?

 Can I just skip this stuff as yet more language-specifc jiggery-pokery that I
 almost certanly don't need?
Hopefully so, and actually you should not see anything about langs directory if
you have not set USE_LANGUAGE_CHOOSER=true in config file.

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Bug#77455: i18n] sets wrong keyboard for i386 - kind of solved

2000-11-24 Thread Michael Sobolev

On Fri, Nov 24, 2000 at 12:49:02AM +0100, Marcin Owsiany wrote:
  So it looks like one should normalize the list of available keyboards in one of
  the places...
 
 Maybe the arch prefixes in langs/*.src should be cut off at
 build time?
Looking at utilities/dbootstrap/kbdconfig.c I would say this is going to
work...  Maybe we just get couldget rid of thevariable prefix (in that file)
and full (in some sense) paths?

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Bug#77455: i18n] sets wrong keyboard for i386 - kind of solved

2000-11-23 Thread Michael Sobolev

On Thu, Nov 23, 2000 at 11:41:24PM +0100, Marcin Owsiany wrote:
 No, this is simpler then I thought at first. LC code in
 main_menu passes "i386/qwerty/pl" to configure_keyboard, which
 then compares it against list of strings like "qwerty/pl" (no
 arch prefix). It doesn't find a match, and prints an error.
 Manual selection works later, because it uses strings without
 arch prefix.
So it looks like one should normalize the list of available keyboards in one of
the places...

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Anyone's building bf on a woody system?

2000-11-22 Thread Michael Sobolev

On Wed, Nov 22, 2000 at 02:21:44PM -0700, Erik Andersen wrote:
 FYI, I just checked in this update to the header files, to be sure
 that the correct dev_t is used for the loop stuff...
Yes, I saw. :)  Thank you.

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: 2.2.18 + LANG_CHOOSER has bad bugs

2000-11-21 Thread Michael Sobolev

On Fri, Nov 17, 2000 at 11:57:36PM -0400, Gleydson Mazioli da Silva wrote:
 Unfortunatelly the Portuguese accentuation support with utf looks be 
 little problematic :( 

 Some accentuation codes of Portuguese language aren't working (the
 original
 accentuation character is replaced by one space). For now, these are 
 the problematic letters (in  Html compatible macros for global
 compatibility 
 and both in upper and lower case):
Hmmm...  That should not be. :)  Well, on a serious note, I checked the
Portuguese texts that are in boot floppies, and they look very Portuguese. :)
(Unfortunately, I don't speak Portuguese, so I do not see if it's ok, or not)

 Aacute
 aacute
 agrave
 Agrave
 iacute
 Iacute
 igrave
 Igrave
 oacute
 Oacute
 ograve
 Ograve
But, as I speak French a bit, I do see the French texts ok...  And they make
heavy use of agrave, for example...

 Some others characters are translated to it normal (without
 accentuation): 
This should never happen.

Could you please send me (only me!) the file all.utf (utilities/all.utf) that
you got on your computer?  I'm going to compare it with the one I got one mine
and see what's happening...

--
Misha

PS  Sorry for late response, I am pretty busy at the moment..


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: 2.2.18 + LANG_CHOOSER has bad bugs

2000-11-15 Thread Michael Sobolev

On Wed, Nov 15, 2000 at 10:40:03AM -0400, Gleydson Mazioli da Silva wrote:
 Michael Sobolev wrote:
  
  On Wed, Nov 15, 2000 at 09:18:34AM -0400, Gleydson Mazioli da Silva wrote:
   The accentuation codes for Portuguese language with utf language chooser
   isn't working too.
  That's a bug.  It should be working.
 
 Thanks, I'm starting another test now...
On second thought, the reduced font may miss needed characters...

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: 2.2.18 + LANG_CHOOSER has bad bugs

2000-11-15 Thread Michael Sobolev

Well, I modified the way all.utf file is generated (those files from
dbootstrap/langs directory are now included).  Please check if Portuguese looks
ok now.

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Debian Boot CVS: aph

2000-11-14 Thread Michael Sobolev

On Tue, Nov 14, 2000 at 08:12:49PM +0100, Marcin Owsiany wrote:
 They are no more independent. This will be going mainstream
 really soon now, and the build process depends on them. Adam
 agreed with me that it is a good idea to have that in utilities
 in order to make a good source package.

On Tue, Nov 14, 2000 at 06:13:55PM -0500, Adam Di Carlo wrote:
 There is reason to merge them, because as  of a day or so or now, you
 will need the sources in order to build boot-floppies.

Well, so you want people to download that huge boot-floppies in case they just
want a custom version of boot floppies without language chooser at all?

I still do not think that this is a good idea to merge these two modules.
Let's better create a separate package (called boot-floppies-lc), and make
boot-floppies suggest it.

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Each team member keep own Notes.name in CVS.

2000-10-20 Thread Michael Sobolev

I'd say that a separate directory would be better...

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Font reduction problems

2000-10-15 Thread Michael Sobolev

On Fri, Oct 13, 2000 at 11:22:01PM +0200, Marcin Owsiany wrote:
 (It doesn't work with either font - reduced or not).
 
 bterm: bogl-vga16.c:223: bogl_vga16_clear: Assertion `y1 = 0  y1 
 bogl_yres' failed.
 Aborted
Well, I rember having such a problem with another driver.  The "fix" was to
comment out the assertions.  Could you please check if it works for you? (Just
comment out the lines 220-225.)

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Font reduction problems

2000-10-13 Thread Michael Sobolev

On Fri, Oct 13, 2000 at 01:37:28AM +0200, Marcin Owsiany wrote:
 $ cd utilities/dbootstrap/po
 $ make all-utf
 $ for x in utf/??.trm;do make `echo $x|sed -e 's/trm/po/'`;done
 $ cat utf/*.po  ../../po.utf
 $ cd ../../../scripts/rootdisk/messages
  edited Makefile so that RN_TARGET_DIR=.
 $ for x in ??;do make ./release_notes.$x;done
 $ cat release_notes.??  ../../../utilities/rn.utf
 $ cd ../../../utilities
 $ cat rn.utf po.utf  all.utf
Hmm...  It looks ok.  I checked on one of my computers at work and it behaves
as you showed in your previous message.  I'll try to investigate what's
happening.  And when I am back at home I'll see if the situation at home
computer is the same.

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Font reduction problems

2000-10-13 Thread Michael Sobolev

Marcin,

It looks I completely lost. :-(  It works on my home computer and it does not
on my both computers at work.

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Font reduction problems

2000-10-13 Thread Michael Sobolev

On Fri, Oct 13, 2000 at 10:43:36PM +0200, Marcin Owsiany wrote:
 On Sat, Oct 14, 2000 at 12:37:30AM +0400, Michael Sobolev wrote:
  Marcin,
  
  It looks I completely lost. :-(  It works on my home computer and it does not
  on my both computers at work.
 
 Maybe try cmp-ing the all.utf and fonts and/or stracing the reduction process?
I created the file on my computer at work and used it for experiments.
Moreover I changed a bit the program (see CVS) and it now outputs the position
and the string where the error occured.  Do you know what I see on my computer
at work?  The first characters (space space space) are illegal!  Very funny.
:-|

Well, I am going to try a debugger or something... :)

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Font reduction problems

2000-10-13 Thread Michael Sobolev

Hmm...  It looks like I found the problem (please see and check CVS version).
I am just not sure I did it right way... :(

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Language chooser - patch

2000-10-12 Thread Michael Sobolev

On Wed, Oct 11, 2000 at 07:32:03PM -0500, Adam Di Carlo wrote:
 The only grotty thing about it, really, is that it required us to have
 bf-utf8 in the same parent dir as the boot-floppies area is, which is
 pretty much the best we can do in a bad situation, IMHO.
That sounds good.  One could check if bf-utf is available in parent directory
and disable LANG_CHOOSER if it is not.

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Language chooser

2000-10-12 Thread Michael Sobolev

On Thu, Oct 12, 2000 at 12:40:44AM +0200, Marcin Owsiany wrote:
 Now there's the problem with fonts. The smallest I could find is font.bgf.
 IIRC Takesoshi Sano said this one is mostly OK (as for what characters it
 contains). But it's too big to fit on a floppy. One option seems to be
 font reduction, but the program 'reduce-font' doesn't seem to work for me:
 
 # ./reduce-font font.bgf  ../allpos.utf
 reduce-font: mbrtowc.c:87: __mbrtowc: Assertion _mbsinit (data.statep)'
 failed.
 Aborted
Hmm...  To begin with, you should use .bdf font.  Then LC_CTYPE should have a
correct value, and the last, libutf8_plug.so should be ld-preloaded.

If it does not work, let me know.  In this case, I'd need the font you are
trying to use. :)

 allpos.utf is a file which i got by catening all 'po' files in UTF - is it
 suitable for 'text'?
It's not sufficient. You should also have a look at utf versions of release_notes
that come to boot disk!

-
Misha


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Font reduction problems

2000-10-12 Thread Michael Sobolev

Well, I've been using unifont.bdf as it comes in bf-utf.

How do you create you all.utf file?

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Font reduction problems

2000-10-12 Thread Michael Sobolev

On Fri, Oct 13, 2000 at 02:38:15AM +0400, Michael Sobolev wrote:
 Well, I've been using unifont.bdf as it comes in bf-utf.
I checked the program with ucs.bdf font, but unfortunately, it comes with
symbolic names for characters (what is not supported by this simple program).

unifont.bdf uses U+ notation only, so it should work.

 How do you create you all.utf file?
And the answer to this question is really important as error code -1 means `An
invalid multibyte sequence encountered'.  I'd guess that the file all.utf is
not in utf-8...

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Language chooser

2000-10-10 Thread Michael Sobolev

Marcin,

I just got an idea! :)  You are trying to use graphical interface, are you not?
And not just bterm program?

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Language chooser

2000-10-10 Thread Michael Sobolev

On Tue, Oct 10, 2000 at 02:01:12PM +0400, Michael Sobolev wrote:
 On Tue, Oct 10, 2000 at 11:18:19AM +0200, Marcin Owsiany wrote:
  Um.. I don't think so... I compiled dbootstrap with LANGUAGE_CHOOSER=true,
  as well as bterm and after setting these all variables ran 
  
  ../bogl/bterm -f ../bogl/font.bgf ./dbootstrap
 This seems to be correct.  Let me check when I am back to home (in about 8
 hours). :)
Last note (before leaving to home :).  What files have you tried to browse?
Have you converted them into UTF-8? (Wild guess, but who knows? :)

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Language chooser

2000-10-10 Thread Michael Sobolev

On Tue, Oct 10, 2000 at 08:25:27PM +0400, Michael Sobolev wrote:
 On Tue, Oct 10, 2000 at 02:01:12PM +0400, Michael Sobolev wrote:
  On Tue, Oct 10, 2000 at 11:18:19AM +0200, Marcin Owsiany wrote:
   Um.. I don't think so... I compiled dbootstrap with LANGUAGE_CHOOSER=true,
   as well as bterm and after setting these all variables ran 
   
   ../bogl/bterm -f ../bogl/font.bgf ./dbootstrap
  This seems to be correct.  Let me check when I am back to home (in about 8
  hours). :)
 Last note (before leaving to home :).  What files have you tried to browse?
 Have you converted them into UTF-8? (Wild guess, but who knows? :)

Well, I tried to run the stuff on my computer and It Works (tm). :(

So what I did.  According to README.BOGL-bterm, I downloaded and compiled
libutf8 plug.  Then compiled bterm.  My commands are:

$ export LD_PRELOAD=`pwd`/libutf8/libutf8_plug.so
$ ./bterm -f ucs.bgf -l en_US.UTF-8 /bin/bash

Then, in bterm, I executed

$ less f

where f is a concatenation of release notes (old versions) in all available
languages.  English, French, and Russian seems to be OK.  Japanese looks
unusual, but it's full of those nice symbols. :)

Could you please execute

$ locale

after you start bterm.  And, another try, what fonts do you use?  From bf-utf?

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Language chooser

2000-10-10 Thread Michael Sobolev

On Tue, Oct 10, 2000 at 11:05:57PM +0200, Marcin Owsiany wrote:
  Last note (before leaving to home :).  What files have you tried to browse?
  Have you converted them into UTF-8? (Wild guess, but who knows? :)
 
 Hmm.. I cat-ed some files like 'UTF-8-demo.txt' which was I guess in libutf8
 directory... They looked OK with the font i'm using. Actually I don't know
 where did this font come from exactly. Anyway, if it displays Polish chars i
 enter in bterm from the keyboard as well as runes and greek, it is probably
 OK.
Aha.  I've got a question for you. :)  Does you keyboard layout generate UTF-8
sequences?  If so, it's OK.  If not, then you are trying to do something in a
wrong way.

 But if the dbootstrap's messages are not automatically converted to UTF8
 during build of the .trm files when LANGUAGE_CHOOSER==true, then that's
 probably the problem. I'll check that in a second.
There is a special target (all-utf) that allows to create UTF-8 versions
of messages...

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Language chooser

2000-10-09 Thread Michael Sobolev

On Mon, Oct 09, 2000 at 04:28:52PM +0200, Marcin Owsiany wrote:
 However, on each line only the first ASCII characters are displayed.
 
 Do you get them to display properly?
I believe, you missed two things (as you did not say that you had read the
README file (the one called utilities/bogl/README.BOGL-bterm :))

LC_CTYPE=en_US.UTF-8

The part before '.' is not important (unless you have messages for this very
locale), the important part is 'UTF-8'.

The second thing is

LD_PRELOAD=libutf8_plug.so

This one is a bit more difficult. :)  You should either get it where README
file points to, or get bf-utf package (where I put it).

If my supposition is wrong (you actually read the README file :), then I'd need
more information to help you get it working.

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: mdash in boot-floppies/documentation

2000-09-30 Thread Michael Sobolev

On Sat, Sep 30, 2000 at 02:44:12AM +0200, Van Buggenhaut wrote:
  According to Section 24 of HTML 4.01 specification, mdash is a valid entity. :)
 I'm using Netscpae 4.75 with Potato and mdash isn't resolved when i read HTML pages
As I already said, Netscape is known to not support all those entities.  That's very
bad. :-(

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: unable to make Russian PDFs still

2000-09-30 Thread Michael Sobolev

On Sat, Sep 30, 2000 at 12:57:12PM -0400, Adam Di Carlo wrote:
 
 Even using woody TeX (which has some serious problems, btw), I still
 get tons and tons of errors making install.ru.pdf
 
   ! LaTeX Error: Command \cyrt unavailable in encoding T2A.
 
 I don't understand how other people are producing this PDF properly.
 It's going to have to be disabled until someone can explain to me what
 to setup to get this file building.
Just a guess.  I am using Russian environment, that means that I have

LANG=ru_RU.KOI8-R

I am going to check if it's a correct guess.

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: mdash in boot-floppies/documentation

2000-09-30 Thread Michael Sobolev

On Sat, Sep 30, 2000 at 01:38:34PM -0400, Adam Di Carlo wrote:
  Yes. This brings us back to what I mention at the beggining of this
  thread. Shoudn't the concerned file be adapted then ?
 
 I guess so.  Since we aren't likely (I don't think!) to get an updated
 debiandoc-sgml, can someone check to see whether we could perhaps
 define the mdash entity, when making HTML, to '-' or '--', perhaps in
 defaults.ent?
Well, my knowledge of SGML is not very solid, so I could be wrong, but AFAIK
just defining mdash somewhere in the document does not help, as those ISO
entities are processed earlier, and, as we all know, in SGML world the first
wins. :)

 If someone can get this fixed (I don't have time) and just commit it
 in CVS that would pretty much fix the problem.
Maybe we could wait a bit before Ardo pays some attention to #72550?

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: mdash in boot-floppies/documentation

2000-09-30 Thread Michael Sobolev

On Sat, Sep 30, 2000 at 03:36:06PM +0200, Van Buggenhaut wrote:
  As I already said, Netscape is known to not support all those entities.  That's 
very
  bad. :-(
 
 Yes. This brings us back to what I mention at the beggining of this thread.
 Shoudn't the concerned file be adapted then ?
Well, I submitted bug #72550.  This is a wishlist, though.  Or do you think
that nonreadable mdashes are more than just inconvenience?  If so, you could
submit a bug against boot-floppies...

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: mdash in boot-floppies/documentation

2000-09-29 Thread Michael Sobolev

On Fri, Sep 29, 2000 at 05:20:11PM +0300, Antti-Juhani Kaijanaho wrote:
 In HTML, there is no portable way to represent an em-dash - except by
 using a plain old "-".
You mean two of them, do you not? :)

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: mdash in boot-floppies/documentation

2000-09-29 Thread Michael Sobolev

On Wed, Sep 27, 2000 at 04:22:01PM -0500, Ardo van Rangelrooij wrote:
 I'll look into this.  The HTML translation map in debiandoc-sgml maps mdash to 
mdash
 which according to the HTML/XHTML book of O'Reilly doesn't exist.  Probably we need 
to
 use #151; but conformance is not guaranteed.  I'll check the whole table and let you
 know what's up.
According to Section 24 of HTML 4.01 specification, mdash is a valid entity. :)

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: mdash in boot-floppies/documentation

2000-09-26 Thread Michael Sobolev

On Tue, Sep 26, 2000 at 07:10:35PM -0400, Adam Di Carlo wrote:
  On Sat, Sep 23, 2000 at 07:40:41PM +0200, Van Buggenhaut wrote:
   On Sat, Sep 23, 2000 at 04:54:50PM +0100, Martin Keegan wrote:
On Sat, 23 Sep 2000, Van Buggenhaut wrote:

 I just checked the web pages and apparently (on my Nestcape, at least) 
mdash; isn't converted at all, which is quite ugly !

Well it should depend entirely on your DTD ...
 
  Actually, the correct answer here is: it should depend on your sgml
  processor.  :) Netscape is known to not process correctly certain
  ISO entities. IE does the job much better in this very case.
 
 Hmm.  Sounds like a bug on debiandoc-sgml...   Can anyone look into
 that?
Do you really think it's a bug?

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: mdash in boot-floppies/documentation

2000-09-23 Thread Michael Sobolev

On Sat, Sep 23, 2000 at 03:50:37PM +0200, Van Buggenhaut wrote:
 There's an apparently new variable introduced in recent versions of canonical
 files, it's mdash; I'm desperately trying to find out what it refers to. Has
 the french translation of it been included. Can anyone help ?
It refers to a dash of length of the letter `m'. :)  A long dash.  There is
also ndash and, I believe, others.

 Also where are the lists of the other "variables" and their related local
 translations ?
The file you are looking for is /usr/lib/sgml/entities/ISOpub.

The idea behind mdash is that in different cases you may need different
constructions to show appropriate symbols.  Two good examples come to my mind:
HTML and TeX.

Hope this helps,

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: mdash in boot-floppies/documentation

2000-09-23 Thread Michael Sobolev

On Sat, Sep 23, 2000 at 07:40:41PM +0200, Van Buggenhaut wrote:
 On Sat, Sep 23, 2000 at 04:54:50PM +0100, Martin Keegan wrote:
  On Sat, 23 Sep 2000, Van Buggenhaut wrote:
  
   I just checked the web pages and apparently (on my Nestcape, at least) mdash; 
isn't converted at all, which is quite ugly !
  
  Well it should depend entirely on your DTD ...
Actually, the correct answer here is: it should depend on your sgml processor.  :)
Netscape is known to not process correctly certain ISO entities. IE does the
job much better in this very case.

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: add_modules_from_floppy() in configure_drivers()

2000-09-01 Thread Michael Sobolev

On Thu, Aug 31, 2000 at 12:33:05PM +0200, Martin Schulze wrote:
I guess that's already done.  I saw it today that 'No' was preselected.
I'm not sure which which dialog that was but I guess it was 2.0 ext2 fs
compatibility or swap.
 
 Sorry, can't find it anymore.  Maybe I was dreaming and separating reality
 and dream didn't work this time.
Maybe we could just rephrase the question to make the answer "yes" the one that
skips the step.

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: add_modules_from_floppy() in configure_drivers()

2000-08-29 Thread Michael Sobolev

On Tue, Aug 29, 2000 at 06:35:01PM +0200, Guillaume Morin wrote:
 So I plan to remove this call temporaly in the CVS. But I'd like this to
 stay in the bf. Have anyone a solution to integrate properly ?
#if 0

? :)

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: Language Chooser

2000-08-28 Thread Michael Sobolev

On Sat, Aug 26, 2000 at 09:29:57AM +0200, Martin Schulze wrote:

 who is in charge of the language chooser?  The head of choose_lang.c says
 that it's haggie but he says, it's somebody else.
I'd say that I am in charge of language chooser.  Please be advised that
choose_lang.c has nothing common with language chooser.

 Next question: Why isn't that thing enabled?
because it was not well tested and it's incomplete (see below).

 Next question: How is it supposed to work?
well, there is a program (bterm) that implements an UTF-8 aware terminal.
Second step is to compile dbootstrap with language chooser support (add
USE_LANGUAGE_CHOOSER=true), then you can run something like

bterm -f nice-font-with-lots-of-characters /path/to/dbootstrap

It should be working. :)  At least, it works for me (tm).

 Next question: What kind of work needs to be done to get it working?  I've
been told that there were some problems with UTF8 fonts c and that
its development involves deep understnding in utf8, dbootstrap and python?
Is that true?  If so, why?
Let's start with false statements. :)  It does not need any understanding of
python nor utf-8.  I used python as it seemed to be a good choice for small
tasks.  I used it to create data files needed for language chooser.

What's done: it does permit to choose a language (along with locale name and
keyboard layout) and makes dbootstrap showing all messages in chosen language.

What's to be done. This information (the chosen language) is not stored
anywhere, as result base-config cannot use it.  This and the fact that we do
not create any configuration file that would allow to create a correctly
localized environment for the user are the major issues to be resolved and
implemented.

I also created a program that takes a set of files, a bdf font and creates a
font that allows to display US-ASCII + everything in these files.  While
experimenting (sp?) with release_notes and message files for dbootstrap I found
that the resulting font is not big at all.

Unfortunately, I was short of time this summer, hopefully I will have more now.
Anyway, I am going to work on this.

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




[i18n] language choser impact

2000-08-28 Thread Michael Sobolev

Just to open a discussion what and how, I'd like to advise on what impact
language chooser has (on my pentium):

dbootstrap (w/o language chooser)   124540
dbootstrap (w/ language chooser)193180

release_notes for all available
languages in UTF-8  32K
dbootstrap messages for all
available languages in UTF-81044K
libutf8_plug (to make UTF-8
really working) 36661
bterm program (which supports
UTF-8 stuff)30556
a reduced (from unifont) font
that displays all characters in
release notes and messages  57059

what gives + about 1265K.  So what can we do with that?  Please comment.

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]




Re: [i18n] language choser impact

2000-08-28 Thread Michael Sobolev

On Tue, Aug 29, 2000 at 03:26:12AM +0400, Michael Sobolev wrote:
 what gives + about 1265K.  So what can we do with that?  Please comment.
This raises two related questions:

how much space do we have on the image that comes to CD (2.88?)
is there any way to determine where have we booted from

--
Misha


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]