(old)Puppy Linux Discussion Forum

A: I didn't find this, thekevinday did. 13 posts above. I just put together a package.

B: It isn't really "fixed", just worked-around. What I did was enable debug-mode. But we shouldn't need to have that enabled.

And C: it seems to be much bigger with debug enabled.

Part of that is that I also enabled xinput, so that people with tablets can take advantage of them without having to recompile GTK (it always used to be enabled in older versions of Puppy, pre 4.0). But I did test it without that first, so I know it's only the enable-debug option that was needed.

Though I don't think that makes a big size difference. I imagine the debug option makes up the bulk of that increase.

EDIT: Actually, by my measurements, xinput adds only about 8k, after stripping. Cool. I thought it would be a bit more than that. Well, that's how much it adds without debug. It probably adds more with debug, because then the debug info for xinput has to be included too. But I'm not about to recompile it again to compare that too. Takes too long. Would rather see if I can find those ZZZzzzzzs I misplaced.

EDIT 2: Something seemed off. You said a substantial size increase, but I'm only seeing a 50k difference between the debug and non-debug versions I compiled. So I dug up the original, and there's a much bigger difference - 120k. So there is something else different. Possibly that I didn't specify an architecture, or maybe he explicitly disabled something. But those ZZZzzzzzzs are getting lonely without me, so I'll work on it another day. (Not tomorrow - must study for exams).

Pizzasgood wrote:A: I didn't find this, thekevinday did. 13 posts above. I just put together a package.

Oops, my mistake. Sorry, thekevinday. I should have read the whole thread more carefully before posting. You'll have to share Pizza's Attaboys.

B: It isn't really "fixed", just worked-around. What I did was enable debug-mode. But we shouldn't need to have that enabled.

All I had to do was click a few times and the problem went away. That's a fix as far as I'm concerned.

Does 4.1.2 incorporate the fix Pizzasgood found for the bug where hitting the Insert key while composing a message in this forum causes Seamonkey to close?

In 4.1.2, I'd used the fix pet (Pizzasgood's gtk+-2.12.1-xinput-debug.pet) but INSERT key still killed seamonkey when I'm posting in forum.

Sorry! Sorry!
I'd made mistake: Isert key won't killed seamonkey after "restart x server"
TK

FWIW I can confirm it is only necessary to replace libgdk, not libgtk
Does anybody know if this fix is making it into 4.2?

Ha! i always remove INS key, because firepup also shuts down when i am posting and I press INS:

Annoying as hell, once i had a 1340 words text written... of course i didn't had it anywhere else, so i simply cryed "WHY GOD??!??! WHY????" before i shut down PC and went to bed (at 7 pm).

No problems in 4.2 hitting the insert key.

Note that this problem also occurs with numlock off using the 0 key on the numbers pad. Must also be typing something in seamonkey such as filling in this form or profile form etc.

The package I posted above doesn't cooperate with SCIM for some reason. I actually encountered the problem months ago, but didn't realize that it was GTK causing the problem.

I'm not sure what is wrong with it to cause the problem. Everything else seems to work.

Anyway, I recompiled it, and it turns out the debug option isn't actually needed - simply recompiling it is enough to stop the insert error from killing things. And this time, it worked with SCIM. I still used the --with-xinput option to support pressure sensitivity on my Wacom tablet.

gtk+-2.12.1-xinput.pet
gtk+_DEV-2.12.1-xinput.pet
gtk+_DOC-2.12.1-xinput.pet
gtk+_NLS-2.12.1-xinput.pet

These have been tested in both 4.1.2 and 4.2.

gtk+-2.12.1-xinput.pet
gtk+_DEV-2.12.1-xinput.pet
gtk+_DOC-2.12.1-xinput.pet
gtk+_NLS-2.12.1-xinput.pet

note that this update/fix also fixes a crash with avidemux 2.4 and searchmonkey

nice one

mike

yes excellent job, now if we just had a updated xorg pet then Firefox 3 series would run right on series4.0 and it would be worth keep 4 series around a bit longer.
ttuuxxx