Forums
I have noticed a problem with ILOBS.
ICEBREAK overwrites at least the first 48 bytes of the ILOB with its own data. This means that when the ILOB is used later in the session the data returned may not be what is expected.
I have a work around, in which I do not use the first 50 bytes of an ILOB.
Is this a 'bug', or is this intentional on behalf of the developers. If intentional, what useful data is available in this part of the ILOB
Regards
Syd
Re: ILOB problem
Hi Syd;
It sounds like a bug.
One thing is that the first byte of an ILOB will always be initilize with x'00 when you create it - but 50 sound odd - I'll have a look !!
Thanks !
Niels Liisberg
Re: ILOB problem
I have observed another problem with ILOBs.
I have a session timeout set at 1440 minutes, yet ILOBs created in a session can disappear after about 5 five minutes - the user space no longer exists on the system.
This causes a problem, because the contents of the ILOB in the session are lost!!
This does not happen all the time, only in some cases. It seems to be erratic.
Is there a reason for this - why should ILOBs disappear before the session has had a chance to time out?
Regards
Syd
Re: ILOB problem
It also sounds like a bug.
I'll post a reply here when I have asked the tech guys.
Can you btw sen me the sourcefile of the first "problem"
Niels
Re: ILOB problem
Appropriate sources attached.
Only the RPGE source is included.
A number of these files are copy source only, used in /INCLUDE instructions
'Documents' is a service program
'Test' Biulds the document index
'Test6' Loads the ILOBs with data from the document index.and then outputs the results to the page with the records in the correct sequence.
The ILOBs used in the session only contain information relating to the folder requested and its children. In most cases the ILOBs will never need to hold the whole index.
Files used are as follows:
DOCLIST is the Physical list of indexed documents
DOCINDEX contains the document ID (derived from IFS information) and the descriptive text of the document
DOCLIST01 - Join Logical view - Folders only
DOCLIST02 - Join Logical view - Documents only - by name
DOCLIST03 - Join Logical view - Documents only - by extension
DOCLIST04 - Join Logical view - Entire index - by name
DOCLIST05 - Join Logical view - Entire index - by extension
DOCLIST06 - Join Logical view - Entire index - by ID
Ultimately, this process must do the following.
1. Automatically index the appropriate portion of the IFS on first use, or when refreshed.
2. Either, Display a folder structure with documents in Alphabetical order by name.
3. Or display folder structure with documents in Alphabetical order by extension
4. A means will be provided to switch between these two differing display sequences in the web page.
5. Each document and folder can have a text description held in the index. This text description will shown beside the document/folder name as part of the folder structure.
6. A means will be provided to edit the document descriptive text
7. Functions to view, copy, create links and delete documents are all required.
Hope this helps
Regards
Syd
Attached Files: Sources.zip (16.2 KB, 96 views)