Howdy,

I am exporting data out of a CHRIS system using Advanced Revelation (roughly 200 tables) and exporting using the ASCII format. I get the following errors on maybe 15% of the tables that I export.

1) Line 1 Broke

2) The Program Stack is Full

3) "object of some sort" cannot be executed - it does not exist

4) Object code for "table" could not be found

I am the most interested in the Line 1 Broke issue. I have had a hard time finding topics regarding that one and most of the tables that we do need have that issue.

Any help or ideas to go in would be great.

The problem relates to code in CHRIS.

This is either

(a) severe recursion in programs which call each other

(b) a severe lack of RAM available to the application

© some sort of memory corruption due to virus or other software

if you enter the application, hit F5 (bring up TCL), type WHO, what memory stats and version stats do you have?

Stevem

Thanks for the response.

The same errors happen on various computers here in our office that we test it on in addition to computers on my clients location. The computers here in our office are also high end machines.

Here are the WHO Stats

Release: 3.12

Type: Novell Netware

Math Processor: 80,487

Available Memory: 387,644

Expanded Memory Allocated: 524,288

Expanded Memory Used: 499,712

Thanks - Ben

Steve - Sorry about the name "Stevem" I got fat fingers and the "m" and "," are really close together!

"Line 1" is what is reported when you compile a program without linemarks for speed purposes - meaning the system no longer actually knows what the line is.

Is this a one off export or something that needs to be done on a regular basis?

[email protected]

The Sprezzatura Group Web Site

World Leaders in all things RevSoft

This use usually because of recursive programs getting load into the stack - the export program uses the braces / xlate / calculate functions so each dictionary item gets loaded as a program in the stack and doesn't flush until the export is done.

Usually the only fix is to split the export into multiple ascii files:

ascii1.txt=fields 1 thru 100

ascii2.txt=fields 101 thru 200

etc.

You have to play around and see how many fields you put in your exports as it depends on how much base memory you have available in ARev.

I wrote a quick and dirty program that would split the dictionary up into import.export template records. I don't have a copy of the code but it is a simple enough to write.

Either that or you write your own export program that doesn't use the braces / xlate function.

OK - nothing appears too bad in terms of memory config.

Another way around the problem may be to

(a) run an expendable shell around CALCULATEX

(b) Mark all programs as "EXPENDABLE" in the source

© Run frequent garbagecollect statements in your code.

(Assuming that you have the source code for CHRIS)

Steve

Warren is of course correct. Your best bet is to create a simple export routine.

If this is a problem let us know

[email protected]

DSigs Radio Free Oregon

Phone: 971-570-2005

I am just doing a one time export of all of the data. As far as I know from our client they have never used the export tool before.

As much as is possible I would like to avoid reworking source code since I have never before this used the Advanced Revelation or CHRIS system - but of course I will do what has to be done.

So… as politely as we can put this…

You are moving a client AWAY from the environment in which we all make a living.

Perhaps paid consultancy would be a better bet??

[email protected]

The Sprezzatura Group Web Site

World Leaders in all things RevSoft

[email protected]

DSigs Radio Free Oregon

Phone: 971-570-2005

View this thread on the forum...