drawterm qid re-use

drawterm qid re-use

[BurnZeZ]

Using drawterm from the tip of rsc's hg repo on bitbucket (b4e2f62f737e)


meiling% ls -pq /mnt/term/home/burnzez/* | awk '{print $1}' | sort | uniq -c | sort -n | tail
  90 (0000000200007819
 100 (00000002000077a4
 104 (000000020000780c
 115 (00000002000077b1
 119 (00000002000077ff
 128 (00000002000077be
 131 (00000002000077f2
 134 (00000002000077cb
 136 (00000002000077e5
 139 (00000002000077d8


Such unique Qids

Re: [9front] drawterm qid re-use

[cinap_lenrek]

http://drawterm.9front.org/

--
cinap

Re: [9front] drawterm qid re-use

[BurnZeZ]

After disregarding the comment at drawterm/gui-x11/x11.c:/polling/
and changing XA_STRING to utf8string in the following couple lines
to allow for utf8 clipboard shennanigans, I find this iteration of
drawterm quite usable.

I did a bit of playing around and haven't noticed any bugs so far.
Also, do you know off-hand which changes would have dramatically
increased the speed of /mnt/term/* filesystem interaction?

Re: [9front] drawterm qid re-use

[kokamoto]

> http://drawterm.9front.org/

Hmm, you have your version of drawterm...
The muscott icon of it!   Why devil?

Kenji

Re: [9front] drawterm qid re-use

[hiro]

It is tuttle glenda (also known as glenda-t).
http://glenda.cat-v.org/gallery/tuttleglenda.png

On 2/23/16, kokamoto@hera.eonet.ne.jp <kokamoto@hera.eonet.ne.jp> wrote:
>> http://drawterm.9front.org/
>
> Hmm, you have your version of drawterm...
> The muscott icon of it!   Why devil?
>
> Kenji
>
>

Re: [9front] drawterm qid re-use

[cinap_lenrek]

> Hmm, you have your version of drawterm...

there was no other way. all the work done for dp9ik
authentication is of no use if i cannot diable the
p9sk1 tickets in the authentication server. as long
as the authentication server issues des encrypted
tickets that can simply be bruteforced, we cannot
have security.

drawterm was the last blocker. now that it can speak
dp9ik, we can start disabling the p9sk1 i the as
by default.

also being able to finally fix some long outstanding
issues like qid collisions and windows drive letters
and x11 graphics glitches is a bonus.

--
cinap

Re: [9front] drawterm qid re-use

[sl]

> The muscott icon of it!   Why devil?

It's not a devil, it's Tuttle Glenda:

	http://glenda.cat-v.org/gallery/tuttleglenda.png

sl

Re: [9front] drawterm qid re-use

[cinap_lenrek]

> After disregarding the comment at drawterm/gui-x11/x11.c:/polling/
> and changing XA_STRING to utf8string in the following couple lines
> to allow for utf8 clipboard shennanigans, I find this iteration of
> drawterm quite usable.

can you be more specific? a diff maybe? i dont have linux at hand.

> I did a bit of playing around and haven't noticed any bugs so far.
> Also, do you know off-hand which changes would have dramatically
> increased the speed of /mnt/term/* filesystem interaction?

no.

--
cinap

Re: [9front] drawterm qid re-use

[BurnZeZ]

[-- Attachment #1: Type: text/plain, Size: 390 bytes --]

On Tue Feb 23 09:40:02 EST 2016, cinap_lenrek@felloff.net wrote:
> > After disregarding the comment at drawterm/gui-x11/x11.c:/polling/
> > and changing XA_STRING to utf8string in the following couple lines
> > to allow for utf8 clipboard shennanigans, I find this iteration of
> > drawterm quite usable.
> 
> can you be more specific? a diff maybe? i dont have linux at hand.

Here you go.

[-- Attachment #2: x11.diff --]
[-- Type: text/plain, Size: 620 bytes --]

diff -r 5bd785ede354 gui-x11/x11.c
--- a/gui-x11/x11.c	Tue Feb 23 01:12:28 2016 +0100
+++ b/gui-x11/x11.c	Tue Feb 23 15:54:42 2016 -0500
@@ -1496,8 +1496,8 @@
 	 * but that would add to the polling.
 	 */
 	prop = 1;
-	XChangeProperty(xd, xdrawable, prop, XA_STRING, 8, PropModeReplace, (uchar*)"", 0);
-	XConvertSelection(xd, clipboard, XA_STRING, prop, xdrawable, CurrentTime);
+	XChangeProperty(xd, xdrawable, prop, utf8string, 8, PropModeReplace, (uchar*)"", 0);
+	XConvertSelection(xd, clipboard, utf8string, prop, xdrawable, CurrentTime);
 	XFlush(xd);
 	lastlen = 0;
 	for(i=0; i<10 || (lastlen!=0 && i<30); i++){

Re: [9front] drawterm qid re-use

[cinap_lenrek]

thanks. now someone with leenuks test this :-)

--
cinap

Re: [9front] drawterm qid re-use

[kokamoto]

\x7f> and x11 graphics glitches is a bonus.

It's also good, we can use scroll mouse as expected.

Kenji

Re: [9front] drawterm qid re-use

[cinap_lenrek]

ok, pushed. the commit message sucks because i have no idea what
clipboard shennanigans are (your bug report sucks), but giving it
a try anyway... :)

--
cinap

Re: [9front] drawterm qid re-use

[BurnZeZ]

On Wed Feb 24 05:31:18 EST 2016, cinap_lenrek@felloff.net wrote:
> ok, pushed. the commit message sucks because i have no idea what
> clipboard shennanigans are (your bug report sucks), but giving it
> a try anyway... :)

Ah, sorry.
Previously, copying unicode from X11 to drawterm resulted in unescaped
unicode ('┱' becomes '\u2531'), or occasionally mangled byte sequences.

Re: [9front] drawterm qid re-use

[sl]

> Previously, copying unicode from X11 to drawterm resulted in unescaped
> unicode ('┱' becomes '\u2531'), or occasionally mangled byte sequences.

Confirmed, I observed this many times.

sl