Frank Rogers
1st May 2002, 18:41
What rules do Baan follow for patch numbering ?

We ask as in some cases we raise the "original" patch is updated twice in two weeks with "different" solution numbers

However we recently have been asked to load a solution number 81246 which was loaded on our systems on 8/6/00 and has been updated as recently as 23/2/2002

Second question is what is correct procedure to install this "revised" patch over existing one ?

Francesco
1st May 2002, 22:11
Funny. That is the first time I see the words Baan, patch and rationale in the same sentence.

ulrich.fuchs
2nd May 2002, 09:54
Some terms and definitions:
Case: A issue you want to have the Baan support solved. You get a specific case number assigned for each issue.

Solution:
1) compiliation of software objects that may be put in your system (normally by PMC nowadays)
2) An Information record in the solution database of Baan.

So, when you log a case that turns out to be caused by a bug, you get a solution (hopefully) with new software objects. If that does'nt solve your problem you get another solution. Each new set of software objects must get a unique number (the solution number) to be identifiable. So, your case isn't "solved" by a solution, it is "closed" when a *particular* solution solves your problem.

However, there is another type of solution: This type is not accompanied by software objects, but just textual information in the Baan support solution database. THere are solutions of this type that are regulary updated (eg: Latest information on Euro conversion or for a service pack) to show you up-to-date information on specific critical issues.

An old solution may also be updated if it turns out to be hazardous. Then Baan support (they're really getting better!) put's a respective note on the solution.

Hope that clears the things a bit.

Uli

Frank Rogers
2nd May 2002, 10:18
Hi Uli

Thanks for the clarification

Are you able to advise on the installation procedure in the second example where solution number is updated ?

We have downloaded the "new" version but it obviously does not as it thinks it is already installed.

We were assuming there would be some "versioning" contained with the solution which would enable the "revised" patch to be updated but this does not appear to be the case in our installation

ulrich.fuchs
2nd May 2002, 14:12
Frank,

which version are you on? Do you use session "import patch" to install solutions or are you using the PMC module?

Frank Rogers
2nd May 2002, 15:31
Hi Uli

We are on 4C4 SP8 using PMC

victor_cleto
2nd May 2002, 17:19
Frank,

Are you sure that you need to re-install this patch?!

I checked it and there's no mention that has been changed (don't take as change date the date that is shown at the top in the "last Modified" field !).

When Baan changed the Knowledge Base more or less two months ago, several patches got a "last modified" new date! When a solution content is changed, then the text message mentions it!
Anyway, you will be out of luck since this patch was installed within Service Pack 6 (you are on 8) so you cannot re-install it (without hacking PMC) - that brings me to the first question, why would you need to re-install it in the first place?!.

Frank Rogers
2nd May 2002, 17:52
Hi Victor !

Thank you for clarifying "date modified" with reference to comments on the case.

The reason for the request to "install" came from an internal source who also was "misled" by the "date mofified" and thought it would resolve an issue we had with restting a repayment status

victor_cleto
2nd May 2002, 19:56
Glad to clarify that.

Baan has been issuing new solutions since more than two years instead of updating the old solution (the change was made while I was still working at Baan so I remember that well). That's why I asked why the need to re-install...

A solution is made when the defect case (handled by PEG) is closed, so the correct date is the date shown as "creation date" - except for the text solutions (help guides, information of solutions that point to the latest for a particular object/portingset) that are updated regularly (more or less).

Now, sometimes old bugs re-ocurr, so it's possible that you may find more than one solution mentioning the same problem - this may be your case and, as I always suggest, have it checked with Baan.

Frank Rogers
8th May 2002, 21:07
Victor

Many thanks

On a similar question , are there any standards as to how long Baan will work a case before classing it as a defect or is it up to the users to "holler"

victor_cleto
9th May 2002, 12:25
Some sort of standard procedure:

1. You open a case where you state your problem.
2. If the bug is well known, Baan will send you the solution and this will solves your problem (case closed).
3. If it's a bug (not something that needs hack of tables), Baan usually upgrades your session(s) with the latest object(s).
4. I the problems persists, then it's tested internaly in a server (fully correctly updated regarding patches). If it's replicated, a defect is open to have it corrected.
5. You are informed a defect is created, they will provide a workaround if needed, when the solution is released you will be informed to install it.

Frank Rogers
9th May 2002, 12:39
Victor

Thanks for the clarification again

In our case they are unable to reproduce the problem in house
so is there then a "another sub process" that is followed ?

In fact the problem in question is the one I have raised on this board to do with negative values after an MRP run which 2 other postings have now stated they also have