CONTROL SET/GET USER statements

[Jordi Vallès]

In my programs, if ownerdrawn controls are present, I use extensively the sentences CONTROL SET/GET USER due that allows some kind of communication between normal program and ownerdraw routines during normal execution.

PB documentation says:

CONTROL SET USER ...    or    CONTROL GET USER ...
. . .
index& is the index number of the user data value to set/retrieve, 
in the range 1 to 8 inclusive.

Due to type errors in my last program I used 11 to 18 instead of 1 to 8 and the program is compiled without errors and works as expected. I realized some tests and seems that this sentences accepts up to 23 user data areas per control.

My question is this behavior not documented is a problem or maybe is an expansion for future versions and I can use more that 8 possible user data areas ? Would be fantastic for me because I need up to 10 user areas in one of my current projects.

Jordi

[Bob Zale]

Jordi --

Please do not use more than 8 32-bit data items. If you use more, you are overwriting other data which will eventually cause serious, but unpredictable results.

Bob Zale
PowerBASIC Inc.

[Jordi Vallès]

Ok Bob. Is what I supposed. But it's nice to imagine more of eight user areas.
Thanks

[Michael Mattias]

>But it's nice to imagine more of eight user areas.

TYPE UserAreaType
   UserArea (1 TO  many many many) AS LONG 
END TYPE

  LOCAL pUA AS UserAreaType PTR,  hMem AS LONG 
  GLOBALMEM ALLOC SIZEOF(@pUA) TO hMem
  GLOBALMEM LOCK hMem to pUA 
  FOR Z = 1 To Many many many 
      @pUA.UserArea(Z) = my value 
  NEXT
  GLOBALMEM UNLOCK  hMem
  CONTROL SET USER   CBHNDL,%ID_CONTROL, 1, hMem
....

(I had submitted a NFS for ONE (1) such setting. I got eight. I was happy).

MCM

[Jordi Vallès]

Sounds a good alternative, and probably faster. Thanks Michael.

A question, what NFS means ? The FreeDictionary on Google has 43 different acronyms for this, from Network File System to Not for Sale, or perhaps means
Not Fully Specified ? I'm not very fond of acronyms.

Jordi

[Michael Mattias]

Sorry about using an acronym; NFS in this context means "New Feature Suggestion."

You send those to the PBPTB.

MCM

[Jim McLachlan]

Jordi --

Please do not use more than 8 32-bit data items. If you use more, you are overwriting other data which will eventually cause serious, but unpredictable results.

Bob Zale
PowerBASIC Inc.

I am puzzled why, if PB only wants you to use a maximum of eight 32-bit data items, it would allow you to use more. I've never written a compiler before but I would hope it could trap any number over 8 that is sent to a CONTROL/DIALOG GET/SET USER command.

[Bob Zale]

PowerBASIC Philosophy on Run-Time exceptions...

http://www.powerbasic.com/support/help/pbwin/index.htm

[Stuart McLachlan]

http://www.powerbasic.com/support/help/pbwin/index.htm

Where on that page would I find "PowerBASIC Philosophy on Run-Time exceptions..." ?

[Michael Mattias]

In summary the philosophy is implemented as , "If we set the system ERR variable in response to an invalid parameter being passed to an intrinsic function, it's news to us, too. For the most part, such actions will get you undefined results with equally undefineable consequences... either some of the time or most of the time."

In case you had not guessed, I would happily sacrifice a little runtime speed in exchange for better error handling in these cases.

But I was not in the room when that decision was made.

[Jim McLachlan]

http://www.powerbasic.com/support/help/pbwin/index.htm

Hi Bob, thanks for checking in and posting.

However, your post was a little cryptic, and perhaps also a bit sarcastic

It contained a link to PB9. Was your intent to indicate that PB9 traps this condition whereas PB8 did not?

[Gösta H. Lovgren-2]

In case you had not guessed, I would happily sacrifice a little runtime speed in exchange for better error handling in these cases.

A little speed here, a little speed there, .... Pretty soon we'll have #Mommy On all the time.

=================================================
"A newspaper is a collection of half-injustices"
=================================================

[Marc Giao]

Since we're talking errors, I assume Mr. Zale is refering to the following:

Error Overview
Unlike the DOS versions of PowerBASIC, Windows versions of PowerBASIC employ a completely different philosophy: to generate the smallest and fastest possible code. Consequently, error handling is placed firmly in the hands of the programmer. PowerBASIC does not stop your program when a run-time error occurs. It is responsibility of the programmer to check for any conceivable errors that may occur after executing a statement...

See "Error and Error Trapping/Error Overview" on the link provided for more details.

[Michael Mattias]

little speed here, a little speed there, .... Pretty soon we'll have #Mommy On all the time.

The solution to that is pretty obvious... after debugging, you just set #MOMMY OFF and recompile!

[Stuart McLachlan]

See "Error and Error Trapping/Error Overview" on the link provided for more details.

I don't see that on the link provided - all I see are the following links,
%-W buttons and:
Introducing PowerBASIC For Windows
New Statements and Functions
Command Summary
Running PB/Win
Data Types
Built-in numeric equates
Built-in string equates
Glossary
Register
Technical Support
Downloads
Peer Support Forums

[Marc Giao]

Hi Stuart,

Wrong list

Look to the left, the "Contents" tab, twelfth item down.

[Stuart McLachlan]

Look to the left, the "Contents" tab, twelfth item down.

You mean the "Flash inside another Flash" left hand panel that my Firefox "FlashBlock" add-on just shows as a blank - not even an icon to show that there would be more content there if I activated the top Flash menu?

Grrrrr!

[Marc Giao]

Should look as follows: