GPF on MDE compilation

[SiJP]

I'm having trouble creating an MDE, whereby access will give a general
protection fault each time I try.

Background:

Access 97 Frontend
Access 97 backend's (multiple) using linked tables
all db's secured by mdw

Scenario 1:

Front end version 1

1) DECOMPILE FRONT END
2) Open frontend
3) Compile and save all modules
4) Compact database
5) Create MDE

- works no problem

Scenario 2:

Front end version 2, modified with a couple of new forms, and some module
coding.

1) Open frontend (no decompile this time)
2) Compile and save all modules
3) Compact database
4) Create MDE

- works no problem

Scenario 3:

Font end version 2 again

1) DECOMPILE FRONT END
2) Open frontend
3) Compile and save all modules
4) Compact database
5) Create MDE

- GPF occurs. I have to revert back to version 1.


This is causing a few issues as the new forms/modules have been created by a
member of our dev team, and to re-do the work would take a bit of time. Even
so, Access is compiling ok, so syntatically it is correct. So why would it
gpy on creation of an MDE after a decompile?

I have gone through object by object adding things in etc, and there is no
particular form or module that will cuase the 'bug' to occur - it will happen
on any created object.

- Is there a known bug?
- If not, what can I try to isolate the problem.

Thanks for any help offered - I'm really stuck, and to be honest bricking
it, as the software is due for client distribution shortly!

 

[Rick Brandt]

[snip]

Font end version 2 again

1) DECOMPILE FRONT END
2) Open frontend
3) Compile and save all modules
4) Compact database
5) Create MDE

- GPF occurs. I have to revert back to version 1.

This is causing a few issues as the new forms/modules have been
created by a member of our dev team, and to re-do the work would take
a bit of time. Even so, Access is compiling ok, so syntatically it
is correct. So why would it gpy on creation of an MDE after a
decompile?

My main Access app has had this behavior for about a year. I can use it to
create MDEs as long as I NEVER do a Compile and Save All. I spent quite a bit
of time on it several months ago, but have subsequently given up on fixing it.
My routine for deploying now is...

1) Compact development copy on top of a second copy I keep in my deployment
source folder.

2) Create an MDE in the deployment source folder using the MDB in that same
folder

3) Compact MDB in deployment folder on top of my development copy. I know it's
"clean" because I was able to successfully make an MDE with it.

4) Distribute new version of MDE from deployment folder and begin next cycle of
development.

Back when I spent the time on it I tried EVERY corruption fix idea that I could
find in these groups and on the internet and none of them solved the problem.
That doesn't mean that you should not try the same, but I just wanted to let you
know that not being able to fix it is at least a possibility. As long as you
have at least one routine that will allow you to successfully create MDEs then
you should not be in too bad of shape.

BTW you should never lose any work because of this as you should be creating
copies of your file at the start of every cycle of development including going
through any of the processes you are describing.

 

[SiJP]

Hi Rick, thanks for the response.

I read the post you mentioned over the weekend, along with several others,
along this same sort of lines. One of them being:
http://groups-beta.google.com/group...shes+create+mde&rnum=1&hl=en#31103dea704f1c41


My two main issues I feel are:

- Getting 'a' copy ready for distribution - I will certainly have a bash at
what you've suggested.

- Knowing why this happens. I'm not comfortable in the knowledge that this
can sporadically occur, without knowing why the gpf occurs when creating the
MDE.

The later issue I would love some more responses to.

Cheers,