problem with codebase 6.2

[jgw]

I am trying to fix a problem in a program I wrote about 10 years ago. It has been working without a problem for 10 years but, now, I am getting the out of memory thing when I try and append a record. I have checked my memory and its just fine (lots of it). That being the case I am sure that I have, as far as I can tell, a problem in a file. This program write 5 different files. I have disabled the writes to one file at a time and find that one file write will bring up the error whilst all other files are being written to with no problem. I have examined the each of the fields of the file in question and they all look good. I have also simply tried to write records with blank fields to see if something is wrong that I can't see but it still flunks out. This led me to believe that, possibly, writes to the other files were bad so I examined all of those records too but they too are just fine.

All this makes me believe that I have a memory problem someplace. I have watched my memory, as the program is running and its well within safe boundaries.

I have no idea if anybody is following this forum but, hopefully, somebody is and can help me.

I can be reached at jwhites...@yahoo.com

Thank you..........

[Geoff Barnard]

Hello

I'm still using Codebase 5 for my various projects.

I assume you're referring to error -920, e4memory, 'Out of Memory', which
I'd suspect is just the same in my version of Codebase.

I've never seen anything like you describe, so on the very limited info you
provide, I'm just guessing.

Does the file you refer to have Memos? Does it use Index file(s)? Both
these things are more likely to cause your problem than a simple record
append.

What are the sizes of the files, and associated Memo/Index files?

Your computer may have lots of memory, but at the point of the error, just
how much memory is the prog trying to grab? I'd guess the prog is trying
to grap a LOT, for some reason, and it's trying to grab more than is
available. Regarding an Index, if the index key contains a lot of very
similar data, then the updating of that Tag might be needing more memory
to resolve?

Another question, if you use something like DBU (Clipper) or say the CID
utility that comes with Codebase, do the files behave correctly then? If
it works with DBU but fails with CID then the problem will be in the
Codebase code. If it works with both, then something in your prog. If
it fails with both, then look more at the datafiles (incl any Index).

By all means offer more info as suggested above.

Geoff

for reply remove zero from addr

[jgw]

I see no zero anyplace for me to zero out. I am, obviously, a little confused here. Anyway, I thank you for the response as I was holding little hope. I do hope this gets posted.


type error is exactly the same as you describe. I I am not sure how to check keys. The files are not all that large. The largest would be slightly over 8mb This is a vanilla installation, just dbase (although fox will work on it). There are no memos. The largest cdx file is also slightly more than 1mb. I did check all the file fields necessary to create indexes, in the files that I cannot write to. I suspect the problem was earlier. I have no idea how to check an index file, other than to check the fields the index is based on. As far as I can see they are all working. This is a real mystery to me. You can also contact me at jwhites...@yahoo.com