red screen of death (pluggableTextMorph bug?) 
Author Message
 red screen of death (pluggableTextMorph bug?)

most of this has been posted to the dev mailing list but they have stopped
responding to me so i am trying it here. i appologize for line wrap and
lack of coherence, at this point i just want it to work. is there something
i am doing to cause this or is it squeak? the more i mess with it the more
i think it is squeak since it is failing when it shouldn't and since it
is so nondeterministic, but i would prefer it be my code, anyway thank you
for your time and help:

i have a pluggableTextMorph that will randomly display a red box with a
yellow x in it. Upon doing this it throws a messageNotUnderstood:
'adaptToPoint: andSend:' this is called in the Point class. the error seems
to have nothing to do with my code, and appears to be random in its
occurance, happening about 50-70% of the time at random times within the
life of the program. i was referred here by the squeak irc after trying it
on 3.2g and getting the same error. is this something to do with my code
that i am not seeing or is it something to do with squeak? thank you for
your time. the squeakDebug log is at the bottom, what follows is some
suggestions/qustions and the responses:

Quote:
> So in displayOn:using:at:, positionWhenComposed is nil.

> Looking at where this gets set, we see moveBy: , but this couldn't be it,
> because it gets set to the result of #+, which isn't likely to result in a
> nil.

this will only halt (just a self halt in the method no conditional) when
i get the red box and only after i get the displayOn:using:at: error.

here the positionWhenComposed is set to delta, previously positionWenComposed

Quote:
> The other place that this is set is in NewParagraph>>positionWhenComposed:.

> So I'd put a conditional halt in that method to find out who is setting
>it to
> nil:

> NewParagraph>>positionWhenComposed: pos
> pos ifNil: [ self halt ].
> positionWhenComposed _ pos.

this never halted

Quote:
> What is the value of the point ?


x               43
y               145
thiscontext     Point>>-
arg             nil

begin Squeakdebug log:
19 March 2002 3:41:12 pm^MError: MessageNotUnderstood:
+adaptToPoint:andSend:^MUndefinedObject(Object)>>error:^M[] in
+WorldState>>displayWorldSafely:^M[] in
+BlockContext>>ifError:^MMessageNotUnderstood(Exception)>>handlerAction^MMessage+NotUnderstood(Exception)>>signal^MUndefinedObject(Object)>>doesNotUnderstand:^M+Point>>-^MNewParagraph>>displayOn:using:at:^MFormCanvas>>paragraph:bounds:color+:^MTextMorphForEditView(TextMorph)>>drawOn:^MFormCanvas(Canvas)>>draw:^MFormCan+vas(Canvas)>>drawMorph:^M[] in
+TextMorphForEditView(Morph)>>fullDrawOn:^MFormCanvas>>roundCornersOf:during:^MT+extMorphForEditView(Morph)>>fullDrawOn:^MFormCanvas(Canvas)>>fullDraw:^MFormCan+vas(Canvas)>>fullDrawMorph:^M[] in
+TransformMorph>>drawSubmorphsOn:^MArray(SequenceableCollection)>>reverseDo:^M[]+in
+TransformMorph>>drawSubmorphsOn:^MFormCanvas>>transformBy:clippingTo:during:smo+othing:^MTransformMorph>>drawSubmorphsOn:^M[] in
+TransformMorph(Morph)>>fullDrawOn:^MFormCanvas>>roundCornersOf:during:^MTransfo+rmMorph(Morph)>>fullDrawOn:^MFormCanvas(Canvas)>>fullDraw:^MFormCanvas(Canvas)>+>fullDrawMorph:^M[] in
+PluggableTextMorph(Morph)>>drawSubmorphsOn:^MArray(SequenceableCollection)>>rev+erseDo:^M[] in
+PluggableTextMorph(Morph)>>drawSubmorphsOn:^MPluggableTextMorph(Morph)>>drawSub+morphsOn:^M[] in
+PluggableTextMorph(Morph)>>fullDrawOn:^MFormCanvas>>roundCornersOf:during:^MPlu+ggableTextMorph(Morph)>>fullDrawOn:^MFormCanvas(Canvas)>>fullDraw:^MFormCanvas(+Canvas)>>fullDrawMorph:^M[] in
+SystemWindow(Morph)>>drawSubmorphsOn:^MArray(SequenceableCollection)>>reverseDo+:^M[] in +SystemWindow(Morph)>>drawSubmorphsOn:^MSystemWindow(Morph)>>drawSubmorphsOn:^M[+] in
+SystemWindow(Morph)>>fullDrawOn:^MFormCanvas>>roundCornersOf:during:^MSystemWin+dow(Morph)>>fullDrawOn:^MFormCanvas(Canvas)>>fullDraw:^MFormCanvas(Canvas)>>ful+lDrawMorph:^M[] in
+WorldState>>drawWorld:submorphs:invalidAreasOn:^MOrderedCollection>>do:^MWorldS+tate>>drawWorld:submorphs:invalidAreasOn:^MWorldState>>displayWorld:submorphs:
+PasteUpMorph>>privateOuterDisplayWorld^MPasteUpMorph>>displayWorld^M[] in
+WorldState>>displayWorldSafely:^MBlockContext>>on:do:^MBlockContext>>ifError:^M+WorldState>>displayWorldSafely:^MWorldState>>doOneCycleNowFor:^MWorldState>>doO+neCycleFor:^MPasteUpMorph>>doOneCycle^M[] in Project class>>spawnNewProcess^M[]+in BlockContext>>newProcess^M

--
Holland King        



Wed, 15 Sep 2004 01:40:23 GMT  
 red screen of death (pluggableTextMorph bug?)
can no one help with this? if not is there somewhere else i can try? i
really need this to work. thank you.

--
Holland King        



Mon, 20 Sep 2004 00:45:48 GMT  
 red screen of death (pluggableTextMorph bug?)
Hi Joseph,

Quote:

> can no one help with this? if not is there somewhere else i can try? i
> really need this to work. thank you.

> --
> Holland King


No, no one can help you anymore!!!
I don't know what you did or wrote but you must have brought enourmous
shame over you (and all your family relatives...). As you wrote in your
earlier mail, people on the squeak-dev list are no longer responding to
you. This tells me that you must be a really rare and pathologic case of
communicational incompetence, because the squeak-dev mailing list is the
most helpful, willing and forgiving community i ever heard of. I never
read the words "killfile" or "RTFM" on that list.
I don't think this has changed since last i read this list regularly
(like last nov or so). I'm sure that, if you apologize truly to the list
(for whatever you might have done) and ask for help in a polite way
(maybe include some code along with the SqueakDebug.log) you will find a
solution to your problem really soon.
/SCRN

Hope this strategy will lead you to the solution, if you can't get the
sqeuak community back on your side you really are a helpless soul and
will have a really hard time using squeak in the future anyway.

HTH
Torge



Mon, 20 Sep 2004 16:00:00 GMT  
 red screen of death (pluggableTextMorph bug?)
: Hi Joseph,

: you. This tells me that you must be a really rare and pathologic case of
: communicational incompetence, because the squeak-dev mailing list is the
: most helpful, willing and forgiving community i ever heard of. I never
: read the words "killfile" or "RTFM" on that list.
: I don't think this has changed since last i read this list regularly
: (like last nov or so). I'm sure that, if you apologize truly to the list
: (for whatever you might have done) and ask for help in a polite way
: (maybe include some code along with the SqueakDebug.log) you will find a
: solution to your problem really soon.
: /SCRN

if answering question in detail tha were posted to me constitutes offending
the list then i can't see how anything is answered on the list. before
going off on your little pathetic rant you might want to read my email
and possibly the list archives, as i posted most of what was said
and can see no way in which it was offending or 'communicational
incompetent'.

: Hope this strategy will lead you to the solution, if you can't get the
: sqeuak community back on your side you really are a helpless soul and
: will have a really hard time using squeak in the future anyway.

heh, hopefully after a few weeks i won't have to touch squeak again.

--
Holland King        



Mon, 20 Sep 2004 17:45:06 GMT  
 red screen of death (pluggableTextMorph bug?)
Hi Joseph,
You are right, i did do my homework and searched the archives. No
evidence of offence found. I want to present my apologies.
I was jumping to conclusions.
Now that i read the posts i must asssume that everyone's speechlessnes
is due to their cluelessness. Thing is, we're still missing some details
like:
* OS/Version
* VM/Version
* Image vesion
* Your code (which now is rather suspect of causing the trouble)

I've looked into the image and i found no evidence of any existing
Smalltalk-code that might nil out the variable in question. Can you tell
me what you see in your image if you query for "inst var defs" on
positionWhenComposed and then all the senders of these (few) methods ?

I've seen weird behavior along these lines before (variables changing
value (esp. top nil) for no obvious reason), but only with very
experimental VMs (or VMs largely failing to match image version), and
then agian mostly inside the de{*filter*}. It would be nice to know if you
can reproduce the error with some other environments (like different vm
version (esp. with/without jitter,) different OSs)

HTH,
Torge

Quote:


> : Hi Joseph,

[snipped: my sarcastic and pointless ranting]
OK point was getting a reaction - and that i got
Quote:

> if answering question in detail tha were posted to me constitutes offending
> the list then i can't see how anything is answered on the list. before
> going off on your little pathetic rant you might want to read my email
> and possibly the list archives, as i posted most of what was said
> and can see no way in which it was offending or 'communicational
> incompetent'.

> : Hope this strategy will lead you to the solution, if you can't get the
> : sqeuak community back on your side you really are a helpless soul and
> : will have a really hard time using squeak in the future anyway.

> heh, hopefully after a few weeks i won't have to touch squeak again.

I don't see that this mind set is an exquisite prerequisite, but hey...
Quote:

> --
> Holland King




Mon, 20 Sep 2004 19:27:12 GMT  
 red screen of death (pluggableTextMorph bug?)

Quote:
> heh, hopefully after a few weeks i won't have to touch squeak again.

        This is for school? If so, were you able to get help from the
instructor, teaching assistants, or fellow classmates?

-C

--
Craig Latta
composer and computer scientist

www.netjam.org

To quote is human, to paraphrase divine.



Tue, 21 Sep 2004 00:16:23 GMT  
 red screen of death (pluggableTextMorph bug?)

sorry for mailing you, meant to post here, but hit the wrong key here is
what i replied with.

: Now that i read the posts i must asssume that everyone's speechlessnes
: is due to their cluelessness. Thing is, we're still missing some details
: like:
: * OS/Version

win 2k, freebsd 4.4, linux 2.4

: * VM/Version
: * Image vesion

when i click on the squeak flap and about this system button i get this
in both freebsd and win2k:
squeak 3.0 of 4 february 2001 latest update : 3545

when i right click on the title bar of the entire window in win 2k i get this:
squak vm 3.1 alpha build

this however is the same as what you get when you download from the squeak
site in the main windows section.

: * Your code (which now is rather suspect of causing the trouble)

i really hope it is my code. i have posted my code at:
www.cc.gatech.edu/classes/AY2002/cs3911_spring/Projects/Team9/moos-mu...
the section that is causing problems is the MainScreen class. in this
more specifically the winmorph: method and the mainOutput pluggableTextMorph

: I've looked into the image and i found no evidence of any existing
: smalltalk-code that might nil out the variable in question. Can you tell
: me what you see in your image if you query for "inst var defs" on
: positionWhenComposed and then all the senders of these (few) methods ?

i will post this later today when i get a chance

: I've seen weird behavior along these lines before (variables changing
: value (esp. top nil) for no obvious reason), but only with very
: experimental VMs (or VMs largely failing to match image version), and
: then agian mostly inside the de{*filter*}. It would be nice to know if you
: can reproduce the error with some other environments (like different vm
: version (esp. with/without jitter,) different OSs)

i have two othe people working with me on a variety of platforms (albeit
fairly standard ones). we have tried it under 3.2g also as suggested on the
irc channel. it does presist across new images and vms.

thank you for your help.

--
Holland King        



Tue, 21 Sep 2004 00:23:42 GMT  
 red screen of death (pluggableTextMorph bug?)
On Thu, 4 Apr 2002 16:23:42 +0000 (UTC), Joseph Holland King

Quote:

> i really hope it is my code. i have posted my code at:
> www.cc.gatech.edu/classes/AY2002/cs3911_spring/Projects/Team9/moos-mu...

Just a side note, I think this URL is wrong.  Try:
www.cc.gatech.edu/classes/AY2002/cs3911_spring/Projects/Team9/moose-m...

        -Chris

--

        If you believe that I'm not out there,
        Then think, and breathe again.
                -"Epic", Ridgely



Tue, 21 Sep 2004 00:54:24 GMT  
 red screen of death (pluggableTextMorph bug?)


Wed, 18 Jun 1902 08:00:00 GMT  
 red screen of death (pluggableTextMorph bug?)
: [snipped: my sarcastic and pointless ranting]
: OK point was getting a reaction - and that i got
: >

hum there are other more productive ways of getting reactions.

: > heh, hopefully after a few weeks i won't have to touch squeak again.
: I don't see that this mind set is an exquisite prerequisite, but hey...

no its not i agree but after using squeak on two midsize projects (both
about 15 weeks long and 2-3 people who start out somewhere between inexperienced
to adept at it, i have found that squeak is very unsupported in both
documentation and community. you made a reference to forums and communities
where you get 'rtfm' and 'killfile' thrown around a lot, and i tend to find
those much more helpful since they will at least respond and point people in
the right direction even if it is a bit (ok so a lot) rudely and crude.
this has been three weeks now that i have been trying to get help with this
current problem. about the only way that i have found somewhat useful is the
irc channel because i can sit there and bug the people and then have them
step me through it. the documentation for the language is far less than
adequate. any other language i have had the pleasure or pain to use has at
least had something to refer to. the howtos and tutorials on the squeak site
couldn't be less helpful. if you want to see what i mean by good documentation
see python, java, c, lisp etc. i don't mean to bash the language. as a language
i really like it, (hate the ui but oh well). however, when i have to weigh the
cost of the ui, support etc, i find it lacking greatly when compared to other
languages that i could substitute it for. anyway off to class with me, sorry
for the long rant.

--
Holland King        



Tue, 21 Sep 2004 01:00:40 GMT  
 red screen of death (pluggableTextMorph bug?)


Wed, 18 Jun 1902 08:00:00 GMT  
 red screen of death (pluggableTextMorph bug?)

Quote:

> i really hope it is my code. i have posted my code at:
> www.cc.gatech.edu/classes/AY2002/cs3911_spring/Projects/Team9/moos-mu...
> the section that is causing problems is the MainScreen class. in this
> more specifically the winmorph: method and the mainOutput pluggableTextMorph

I gather that's moose-mud.st

I'd like to try it, but it seems that (at least) classes #MagicBox and #MOOSEPluggableTextMorphNI are missing.

I gather that's a MUD client. Do I have to run it against a special server?

Have you noticed that Torge sent a possible workaround to the Squeak list under the subject of "[FIX][hack]NewParagraph problem"?

Joern



Tue, 21 Sep 2004 01:08:26 GMT  
 red screen of death (pluggableTextMorph bug?)

: > heh, hopefully after a few weeks i won't have to touch squeak again.

:       This is for school? If so, were you able to get help from the
: instructor, teaching assistants, or fellow classmates?

this is a senior design project class. the project and implementation are
independent of the class. so, in our case the prof teaching the course and
the prof we are working for do not know squeak well enough to help, in the
class the only two other people who are willing to use squeak are in my group
and have helped me out tremendously but were also stumped by this.

--
Holland King        



Tue, 21 Sep 2004 02:14:35 GMT  
 red screen of death (pluggableTextMorph bug?)


: >
: > i really hope it is my code. i have posted my code at:
: > www.cc.gatech.edu/classes/AY2002/cs3911_spring/Projects/Team9/moos-mu...
: > the section that is causing problems is the MainScreen class. in this
: > more specifically the winmorph: method and the mainOutput pluggableTextMorph

: I gather that's moose-mud.st

: I'd like to try it, but it seems that (at least) classes #MagicBox and #MOOSEPluggableTextMorphNI are missing.

these belong to other people i will have to get their permission before posting
that code.

: I gather that's a MUD client. Do I have to run it against a special server?

unfortunatly the server we use is limited access we might be able to test it
on another server, but i will have to look into it.

: Have you noticed that Torge sent a possible workaround to the Squeak list under the subject of "[FIX][hack]NewParagraph problem"?

no i will look into it though thank you very much.

--
Holland King        



Tue, 21 Sep 2004 02:20:18 GMT  
 red screen of death (pluggableTextMorph bug?)

Quote:
> this is a senior design project class. the project and implementation are
> independent of the class. so, in our case the prof teaching the course and
> the prof we are working for do not know squeak well enough to help, in the
> class the only two other people who are willing to use squeak are in my group
> and have helped me out tremendously but were also stumped by this.

Is there some reason you're using Squeak rather than something a bit more
stable, like VWNC or Dolphin?  My impression (which may be wrong, of course)
is that Squeak is great for Smalltalk experts who want to experiment and play
around, but not so great for those of us without a lot of Smalltalk
experience.  If you just want to get a project done, without wondering about
whether or not the VM and/or vanilly image is buggy, I'd go with VWNC.

--



Tue, 21 Sep 2004 02:34:34 GMT  
 
 [ 16 post ]  Go to page: [1] [2]

 Relevant Pages 

1. NT Blue Screen of Death

2. Bugs, Death and everything

3. Bugs, Death and everything

4. BUGS, BUGS, BUGS, BUGS, C4 BUGS

5. screen display bug

6. Fix a graphic bug (refresh screen)

7. Screen bug in Win95 Wish8.0p2 (console overwrites itself)

8. Bug with interpreter names on multiple screens in Tk (8.0 and earlier)

9. ?Folling bios (dos) for screen saver (screen related)??

10. Imminnent Death of the 'Imminent Death of the Net' Postings

11. Finding Screen Size to centre top-level window on Screen

12. Windows XP: DOS death-knell?

 

 
Powered by phpBB® Forum Software