#PageStream Crash Bug
22 messages in this thread
I have a reproducible bug in PageStream 2.1 which crashes my machine every
time:
1. Select a text field
2. Do a "Replace Text" and have it not find the text to be replaced
3. Leave the text selected and attempt to select "Replace Text"
again.
BOINNGGG!!!
I have a 2000 which is currently running without acceleration, has a 40 Mb
hard drive and 2.5 mb of memory. Two floppies, Flicker Fixer, Multisync
monitor (not that it matters, but why not be complete?)
Is there a fix for this and if so, what do I have to do to get it?
Mike
I can't reproduce that at all …. however, when you say it bombs out, does
it GURU, or just freeze the machine?
PageStream is forever locking out the machine while it does something, wit
hthe result that the machine dies while it finishes what it was doing (a
particularly unfriendly way of doing things since it effectively does not
multi-task!). Could it be that the lock out is due to this?
I have often found that PStream will go away like this for up to 2 minutes
at a time … it appears to have crashed, but it hasn't.
Mike (Whapping from the UK)
I find through further time- and work product-consuming investigation that
it only occurs when the text to be replaced is longer than the requester in
which it is entered. Normally, I get a guru, but tonight when I apparently
typed in more text than the requester could handle, it just locked up. The
machine couldn't have been doing anything, I hadn't told it to do anything
yet. I don't want to be pushy, but couldn't the machine just beep when an
overflow like that occurs, instead of crashing or locking up?
Mike
Still more investigation resulting in considerable wear and tear on
the guru has led to the conclusion that overfilling the string
gadgets will definitely crash PS 2.1. This occurs with either the
search text or replace text requesters. Apparently, fully filling
the requester the first time (which I had unwittingly done) without
crashing will cause a crash with the second request when the program
tries to offer the gadget with the previous request still in place
and tries to find room for the cursor. That was what I was getting
which led to my initial report. Now I find that if you bring up
either requester and start typing and keep typing, the system will
invariably fail, sometimes with a guru and sometimes by locking up. I
tried massively increasing the stack (for whatever that was worth)
from the stock 4096 (!) to 64 K, with no change in the result. I
didn't bother to count the number of characters it takes to blow
things up, but I was typing "Sunflower Plants Having a High Frequency
of Regenerability from Isolated Protoplasts and Method of Using Same
in the" [BOINNGGG!! — not including the quotes]. I realize that's
kind of a long search string, but I thought it would be quicker than
retyping the text in situ (and apart from the crash, it was.) I
suppose this shows why you test software under extremes of use rather
than under normal conditions. However, I'm repeating my call for
help and a fix. This shouldn't happen with version 2.1 of anything,
and especially not professional quality software.
Mike
Mike: # chars in search text gadgets. Bug noted. I'll put a post-it note
above the boss's desk.
SL
PageStream should never 'go away'. It's true you won't see the busy pointer
or be able to do something during some intensive work (like when I
duplicate 1500 objects 3 t5imes ,,, oops, line noise, 3 times with offsets.
This creates a bit of a bottleneck (and yes, I was doing this yesterday.)
But other than that, PgS should multitask properly. What specific problems
are you having.havou having?
SL Tech
A typical example (off the top of my head) is where a complete document is
"redrawing" and one goes to do something … perhaps click on a scroll bar.
PStream just locks the pointer, and you can do absolutely nothing else
until the redraw is complete (which can be a _very_ long time sometimes.
If you want more specific examples, I could come up with a few … but I'd
need to sit down and make sure that they're reproducable.
Mike (Whapping from the UK)
Several times now, I have had PGS lockup the entire machine when trying to
select the Text tool from the tool box when the document window was still
updating. I 'learned' long ago not to do that with older version of PGS
but thought that maybe it was fixed.
Are you aware of the problem? This is on an A3000 37.175/37.64. Happens
in either custom screen of on top of workbench. -bill – via Whap!
Bill,
I have had that exact problem happen to me in PageStream. I click that
text tool and GURU city. That's been the only thing I can get it to lock
up on but it is very frustrating when it does.
>>DonovaN
Donovan;
Hummm, that sure polite. Indeed a LOT more polite than the things that I
say when it happens to me… seems that it ALWAYS happens when it has been
a long time since I did the last save. -bill – via Whap!
Ahhh Bill,
I have learned the art of saving after every single step in most of my
programs. I makes work slower, but not as slow as it would be if I lost a
major portion of it because of that ##@$!!?! bug. I'm calm, I'm calm…
>>DonovaN
Donovan;
I know, I know, I KNOW!! The bug seldom hits me…. only when I'm in a
real hurry and don't think to save. But then there's no surprise there 🙂
-bill – via Whap!
Yes, the mode switching bug is a serious problem. The bug is not evidenced
in 2.2 so the problem is fixed.
SL Tech
Good, now when do we get 2.2? -bill – via Whap!
I've had PageStream 'go away' in another sense many times. Whenever I
start to print a document and then go do something else, the print
'process' will stop. Seems that I have to have the window active, or it
just gives up. Sure it lets me do other things, but it stops doing things
itself!!
Vic, It doesn't seem to be a problem for me if you actually let the
printing begin before you switch to something else.
-sja
Vic;
Same here but I think that all version prior to 2.1 did stop printing if
changed active windows. -bill – via Whap!
Bill,
I still get it on 2.1, althought I _think_ it only happens at page
change time. We have a Canon bubble jet which changes pages VERY slowly
and I think PS sees a status it doesn't like and just sorta goes to sleep.
Vic,
I have it stop printing sometimes when I go away. If I reactivate the
PageStream print window it starts up again and continues for a while
longer. It seems that the more active I am in the other screens the
quicker PGS stops printing and needs to be reactivated.
btw I'm just using a 24 pin dot matrix printer.
Kevin Ushko
Edmonton, Alberta
Vic, I don't think PGS handles driver returned errors at all (anymore). I
have taken advantage of the fact on occassions. Using the LazerXpress
driver, I can just open the paper feed (which causes a paper out error) and
now, instead of the error requester, PGS just dumps the data to the bit
bucket.
I would have complained but found the 'bug' to be a handy feature for when
you do something stupid like tell PGS to print a 50 page document when you
really only wanted just the current page!
-bill – via Whap!
Vic: this is a known problem. Sorry.
I'll check it out and make sure it's fixed.
SL Tech