[9fans] drawterm prints blanks instead of characters

David Tolpin

unread,

Mar 11, 2004, 6:53:40 AM3/11/04

to

Hi,

Sometimes, drawterm begins to print blanks instead of
glyphs. This happens with X11 drawterm with some hardware
and drivers.

It happens because Xfree86 drivers for some video cards assume that
masks (stipples) must have height equal to a multiple of 8. This
causes drawterm to display blanks instead of glyphs with some fonts.
To use drawterm with these drivers, the patch at

http://ftp.davidashen.net/misc/drawterm-stipple.diff

can be applied to the source (/sys/src/cmd/unix/drawterm).
drawterm will become slightly slower, but will not loose glyphs
any more.

Use ape/patch to apply the diff.

David Tolpin
http://davidashen.net/

andrey mirtchovski

unread,

Mar 11, 2004, 9:38:42 AM3/11/04

to

David,

The patch/* code should be used to submit patches to the Plan 9
developers. see patch(1) for more information.

I see how you could've missed it -- the man page isn't indexed so lookman
doesn't show it, but it's really simple to find binaries if you just list
recursively (scripts available in the archives) /bin.

With the disappearance of the gajillion [s]bin directories the need for
programs like 'locate' becomes non-existent :)

andrey

David Tolpin

unread,

Mar 11, 2004, 9:53:46 AM3/11/04

to

> The patch/* code should be used to submit patches to the Plan 9
> developers. see patch(1) for more information.

I saw it a long time ago. I clearly understand its purpose and
its purpose does not meet my intents.

> I see how you could've missed it

I didn't. I am not submitting a patch to the Plan 9 developers.
Plan 9 developers should figure a way to adjust the pixmap size so
that it works with all drivers properly, and check for Xlib status.

What I have sent is a fully working, but temporary solution, aimed
at temporary disabling buggy code. It should not be included into
the source tree, and should not become the only solution to the
problem.

When/if I'll find time and desire to re-write pixmap-related code
so that it is actually working, I'll submit the patch using appropriate
tools.

Sincerely,
David Tolpin
http://davidashen.net/

Russ Cox

unread,

Mar 12, 2004, 12:58:36 AM3/12/04

to

> The patch/* code should be used to submit patches to the Plan 9
> developers. see patch(1) for more information.

buh? Last I checked this still didn't work since there
is no publically writable Plan 9 server.

Russ

andrey mirtchovski

unread,

Mar 12, 2004, 1:15:49 AM3/12/04

to

nemo has been submitting patches there, i thought you've enabled it:

home% 9fs sources
post...
home% cd /n/sources
home% ls -l /n/sources
...
d-rwxrwxrwx M 60 glenda sys 0 Mar 3 09:51 patch
...
home% ls -l patch
d-rwxrwxr-x M 60 nemo sys 0 Mar 3 00:39 patch/defaultmkfile
d-rwxrwxr-x M 60 nemo sys 0 Mar 3 09:51 patch/httplength
d-rwxrwxr-x M 60 nemo sys 0 Jan 15 08:45 patch/manfix
d-rwxrwxr-x M 60 nemo sys 0 Jan 14 02:48 patch/stats
d-rwxrwxr-x M 60 nemo sys 0 Jan 23 00:43 patch/test
home%

andrey

David Tolpin

unread,

Mar 12, 2004, 1:37:40 AM3/12/04

to

>
> nemo has been submitting patches there, i thought you've enabled it:

nemo has write access to sources; lc /n/sources/nemo

David Tolpin

unread,

Mar 12, 2004, 1:40:41 AM3/12/04

to

just

9fs sources
bind -bc /tmp /n/sources/patch

then the patch archive is in /tmp and can be delivered to Bell Labs
through other paths.

andrey mirtchovski

unread,

Mar 12, 2004, 1:44:40 AM3/12/04

to

>>
>> nemo has been submitting patches there, i thought you've enabled it:
>
> nemo has write access to sources; lc /n/sources/nemo

no way! when did that happen?

Fco.J.Ballesteros

unread,

Mar 12, 2004, 2:24:38 AM3/12/04

to

When I was using the bitsy, to make it easier to merge the
changes. I though it was easier to use patch w/o binding a
local directory (as David suggested).