NO

Author Topic: small bug and 2 suggestions 2.90 beta7  (Read 4292 times)

nitex

  • Guest
small bug and 2 suggestions 2.90 beta7
« on: October 26, 2004, 07:45:18 PM »
hi pelle,

i haven't tested this with the beta 8 yet because i don't have internet at home. i'm using win xp prof sp2.

bug:

if you have the ide open and a project loaded and then create a new project from file->new->project... the addins get loaded twice (dgb/rls switch).

suggestions:

it would be nice if one could define either #UNICODE or #_UNICODE to compile for unicode and not both. currently one needs to define #UNICODE for windows headers and #_UNICODE for tchar header to compile successfully.


when enabling the optimize for size option one could optimize

#define APP_CLASS _T("MyAppClass")

to something like

LPTSTR lpszAppClass = _T("MyAppClass");

so the string would be only once in the final executable.


thanks for making such a nice developement environment.

alex

Offline Pelle

  • Administrator
  • Member
  • *****
  • Posts: 2266
    • http://www.smorgasbordet.com
Re: small bug and 2 suggestions 2.90 beta7
« Reply #1 on: October 26, 2004, 08:00:44 PM »
Hello,

Quote from: "nitex"
if you have the ide open and a project loaded and then create a new project from file->new->project... the addins get loaded twice (dgb/rls switch).

This should be fixed in beta 8.

Quote from: "nitex"

it would be nice if one could define either #UNICODE or #_UNICODE to compile for unicode and not both. currently one needs to define #UNICODE for windows headers and #_UNICODE for tchar header to compile successfully.

I havn't checked this lately, but at least before, you needed to do this for the Microsoft compiler. I have done it like this for 'compatibility' reasons. But I agree, It's a bit annoying...

Quote from: "nitex"

when enabling the optimize for size option one could optimize

#define APP_CLASS _T("MyAppClass")

to something like

LPTSTR lpszAppClass = _T("MyAppClass");

so the string would be only once in the final executable.

Yes, 'string pooling' doesn't work for wide character (Unicode) strings, because they are handled in a different way than 'normal' strings. I will look at it. Not for v2.90 though.

Pelle
/Pelle

nitex

  • Guest
small bug and 2 suggestions 2.90 beta7
« Reply #2 on: October 28, 2004, 06:41:55 PM »
thanks for your answer...


alex