|
DanteX Apprentice
Joined: 13 Aug 2007 Posts: 166
|
Posted: Sat Jan 30, 2010 12:53 pm
[3.13] Coloring problem |
Since upgrading to 3.13, I've started to see some really strange coloring. The mud output gets some random colors, differing from what it's supposed to be. The scrolling also feels slower, where lines appear to not scroll properly, sometimes "lagging behind" making the same line appear twice in a row. If I press CTRL+Z the double-lining disappears, but the coloring gets permanent.
|
|
|
|
Zugg MASTER
Joined: 25 Sep 2000 Posts: 23379 Location: Colorado, USA
|
Posted: Sat Jan 30, 2010 6:12 pm |
You need to open the Script Debugger and turn on the Raw Input/Output message and then capture the text displayed by the MUD that has the incorrect colors. Post the output within [code] tags here. It would also help to see a screenshot of something showing the correct colors (like the CMUD 2.37 version) compared to the 3.13 screen shot so I can see exactly what the problem is.
I'm not aware of any changes made to the text color system, so I have no idea what is causing it. Ctrl-Z toggles the scrollback, so I have no idea on the lines not scrolling properly. But again, I'd need to see the raw dump to try and reproduce it. Also, please always post what version of Windows you are using.
Scrolling speed has not changed for me here. In fact, comparing 3.12 with 3.13 with my timing test script, the 3.13 is actually a bit faster. |
|
|
|
DanteX Apprentice
Joined: 13 Aug 2007 Posts: 166
|
Posted: Sun Jan 31, 2010 9:07 am |
This seems to occur randomly. I cannot say how to reproduce it. I will see different miscoloring ff I type "eq" at various occations...
Below is the same item, it should have the same coloring always... and it does have it mostly but randomly...
Sick coloring:
Code: |
<ESC>[0;32m[ <ESC>[0;36mWorn on torso <ESC>[0;32m]<ESC>[1;37m:<ESC>[0;37m <ESC>[1;31m(K)<ESC>[1;34m(M)<ESC>[1;37m(G)<ESC>[1;36m(H) <ESC>[0;37m<ESC>[1;31mAardwolf <ESC>[1;33mBreastplate of magic resistance<ESC>[0;37m <ESC>[1;37m(<ESC>[1;32m291<ESC>[1;37m)<ESC>[0;37m<LF><CR>
|
Healthy coloring:
Code: |
<ESC>[0;32m[ <ESC>[0;36mWorn on torso <ESC>[0;32m]<ESC>[1;37m:<ESC>[0;37m <ESC>[1;31m(K)<ESC>[1;34m(M)<ESC>[1;37m(G)<ESC>[1;36m(H) <ESC>[0;37m<ESC>[1;31mAardwolf <ESC>[1;33mBreastplate of magic resistance<ESC>[0;37m <ESC>[1;37m(<ESC>[1;32m291<ESC>[1;37m)<ESC>[0;37m<LF><CR>
|
I use Win XP with SP3. I had 3.12 installed, and just ran the 3.13 installer to update. |
|
|
|
DanteX Apprentice
Joined: 13 Aug 2007 Posts: 166
|
Posted: Sun Jan 31, 2010 9:22 am |
I have sent an email to support@zuggsoft.com with a picture that demonstrates the problem.
|
|
|
|
charneus Wizard
Joined: 19 Jun 2005 Posts: 1876 Location: California
|
Posted: Sun Jan 31, 2010 8:11 pm |
Yeah, I've been experiencing the same thing on Aardwolf. Things will just get randomly colored for no apparent reason. It's purely random, though, and I can't find a way to reproduce it or what the problem is...
It's random things like this that really irritate me because I can appreciate how detailed you need things to be as a programmer. You can't fix a problem when randomness is involved because you have several lines of code to look through.
If there were anything that you changed regarding color, that might be the problem... it might have something to do with making it more 'Windows 7 compatible' or whatever... I just hope it gets fixed because screwing up with colors just bites. One thing someone might try (and I'll probably be the one to do it) is create a completely new session, no triggers, no aliases, everything standard and connect with #DEBUGFILE running. Then play through until you see the colors messing up. Then send the entire log to Zugg... *shrug* If I had a better internet connection right now, I'd do that, but that won't be for a few days.
Charneus |
|
|
|
nexela Wizard
Joined: 15 Jan 2002 Posts: 1644 Location: USA
|
Posted: Mon Feb 01, 2010 1:23 am |
yeah I am getting this too completely random can't narrow it down. sometimes it will go away for a while then creep back up
3.13 Win7 64bit |
|
|
|
Zugg MASTER
Joined: 25 Sep 2000 Posts: 23379 Location: Colorado, USA
|
Posted: Mon Feb 01, 2010 6:32 pm |
The debug logs for "sick coloring" and "healthy coloring" are exactly the same! It's going to take a while to track this down. Am I correct in assuming that this problem only starts happening after being connected to the MUD for a while? I can't think of anything that changed in the coloring code from 3.12 to 3.13. About the only fix was a crash in the ClearAttrs routine, so maybe I messed it up so it doesn't properly clear and reuse attribute records correctly.
|
|
|
|
DanteX Apprentice
Joined: 13 Aug 2007 Posts: 166
|
Posted: Mon Feb 01, 2010 8:03 pm |
Ok Zugg, I thought that I wrote the wrong code, but I didn't. The outputs of a corrent and a wrong item are identical when I double check now. And no. This problem occurs quite emediatelly after connecting. I closed cmud. Started it again. Connected. Typed "eq" to check my eq, and emediatelly saw miscoloring.
|
|
|
|
Zugg MASTER
Joined: 25 Sep 2000 Posts: 23379 Location: Colorado, USA
|
Posted: Tue Feb 02, 2010 12:13 am |
If you can reproduce this as soon as you log in, then create a debug file for me. Connect to your MUD, then immediately type:
#DEBUGFILE test.txt test.raw
and then type the "eq" commands that cause the bad text colors. Then close CMUD and email me both files as attachments to sales@zuggsoft.com and I will try to replay them here. |
|
|
|
DanteX Apprentice
Joined: 13 Aug 2007 Posts: 166
|
|
|
|
Zugg MASTER
Joined: 25 Sep 2000 Posts: 23379 Location: Colorado, USA
|
Posted: Tue Feb 02, 2010 6:27 pm |
Got the file. Will try to test it soon before the next patch.
|
|
|
|
Zugg MASTER
Joined: 25 Sep 2000 Posts: 23379 Location: Colorado, USA
|
Posted: Thu Feb 04, 2010 12:39 am |
I found the bug. It was a side effect of a crash fix that I added in 3.13. Definitely a very serious problem that is going to cause all sorts of color problems, and potentially is causing the slowdown and memory leak problems also reported. CMUD was basically never clearing any attributes on a line and was moving all attributes from one line to the next. Bad bad bad. Fixed for the next update later this week.
|
|
|
|
|
|