slightly irritating 9 thing with F1 help

[Chris Holbrook]

Maybe this is just me being fussy, but it is a change from PBWin8 and PBCC4:

Suppose you have a message identifier on the screen in the editor, with the % prefix. move the cursor to it and press F1. In V9 you get the PB online help index page. There was a time when it was the Win32 Programmers reference for the symbol which appeared, which was more "helpful".

Or maybe there is a configuration option?

[Richard Angell]

If the % is separately selected or the cursor is in a user defined equate Win32Help will pop up. The caveat is you must have configured the path to your winhelp file in the Options.

However if the % is part of an system defined equate and you do not highlight just the %, PBWin Help wil load.

For example. in the following code line:
#IF NOT %DEF(%WINAPI)

Highlight the % bolded below,Press F1 and you should get the Win32 help:
#IF NOT %DEF(%WINAPI)

But if your cursor is immediately before, in or immediately after %DEF then PBWin Help will load.

Now try %WINAPI with the cursor immediately before, in or immediately after the equate. Win32Help should load.

If it doesn't check the options and make sure Win32 help is where you think it is.

[Chris Holbrook]

If the % is separately selected or the cursor is in a user defined equate Win32Help will pop up.

Rick, I did not intend to indicate that there was a fault in it, the behaviour which you describe is - as far as I can be bothered to investigate - what happens.

I dare say PB had a good reason to change from the way it used to work and I'm sure that in time for V10 I will get used to it.

[Kev Peel]

Actually, I believe this behaviour is unintentional. You can no longer quickly reference the Win32 help for a particular message as the IDE now displays the PB help file when F1 is pressed on any highlighted keyword -- which these messages are now classed by the IDE. Obviously, F1 now brings you to a dead end as not all (most?) of these constants are documented in the PB help file. Another gripe (for me) is the forced lower case - I prefer the case to be left alone

[Richard Angell]

Actually, I believe this behaviour is unintentional.

Do you rather mean "intentional"?

Another gripe (for me) is the forced lower case - I prefer the case to be left alone

Not sure what you are referring to here as being forced to lower case, Kev? Couldn't be keywords could it ... they are set in the Options for case.

[Kev Peel]

Rick,

No I meant "unintentional" as it's a step backwards in functionality since version 8 and all previous versions. Start the IDE, type "%WM_CREATE" and press F1 over any part of it. You're not going to get the Win32 help file up any more.

Also, there does not appear to be any setting to change or disable the auto lower-case for these constants (it is forced). I did mention this previously.

[Richard Angell]

I do not see any auto-lowercasing here, unless Window\Options\Editor tab for Keyword Case is set to Lower Case. Perhaps what you want is the IDE to distinguish between compiler supported system equates and language keywords?

The IDE would be improved if it could distinguish between PB unique equates and Win32 equates to deliver the best of both worlds. Alternatively, PB's help file might summarize the Win32 equates in PB BASIC terms and provide a link to load the Win32 help. 9.01?

[Kev Peel]

Rick,

What IDE version do you have? - I''m using 9.00.0085.

I have keywords set to mixed case, and it forces constants to lower-case. It would be nice if there was an option to not change the case of constants, or even configure this separately (I don't consider constants as "keywords")

It would be better if the IDE did not display the PB help file (and use Win32.hlp) if it did not support a keyword. At the moment, it displays PB's help for all highlighted keywords and constants (and also in white space), regardless whether that keyword is supported or not by the help file.

[Richard Angell]

What IDE version do you have? - I''m using 9.00.0085.

The same version

I have keywords set to mixed case, and it forces constants to lower-case.

Mine is set to all upper case, which accords with how tthe help file topics are also presented, at least in their topic line. Mixed case would only consider the first character, which in this case would be "%".

It would be nice if there was an option to not change the case of constants, or even configure this separately (I don't consider constants as "keywords")

Most would agree they are not keywords. Those highlighted are however PB directly supported in the compiler equates. In some cases this means one does not need to include COMMCTRL.INC as well as COMDLG32.COM and perhaps some others. That said I do agree the flexibility to differentiate system supported equates from keywords as an option would appeal to some users.

It would be better if the IDE did not display the PB help file (and use Win32.hlp) if it did not support a keyword. At the moment, it displays PB's help for all highlighted keywords and constants (and also in white space), regardless whether that keyword is supported or not by the help file.

I agreed in part previously, but with the caveat that if PB decided to support the equates with a short one to three line explanation, plus a link to load the equate equivalent page in the Win32 help. then I tthink that would be even better. Some of MS's wording could be improved on for users with more VB and similar BASIC backgrounds. That being as opposed to C style explanations.

[Michael Mattias]

Those highlighted are however PB directly supported in the compiler equates. In some cases this means one does not need to include COMMCTRL.INC as well as COMDLG32.COM and perhaps some others

Browsing the online help I find it interesting that the notification and style constants...
TCN_*, TCS_*
EN_*, ES_*
TVN_*, TVS_*

.. are now predefned, but the messages used (TCM_*, EM_* and TVM_*) with the associated controls are not.

I guess we are expected to use the intrinsic DDT support, which is fine for DDT users.

But since equates can be defined twice or more as long as they are the same (and predefined equates RE-defined) I guess we SDK users can #INCLUDE the header files anyway.

Generically speaking, I am of the opinion that any predefined equate should be completely documented. I could not find said documentation.

MCM