On Wed, 31 Jul 2002 22:51:37 +0200, [EMAIL PROTECTED] wrote:
> Not really. I'm talking about something else in fact. You can create
>a unit containing array of messages, and possibly a function
>returning individual messages. That unit could be compiled into a
>DLL, and such a DLL could be linked in dynamically then.
Oh, no. Such method good and easy for developer, but not for user.
> OK, I understand how .msg files are being created. It's just a
>different thinking on my side (a complete solution encompassing
>everything - including a possibility to have everything in Pascal
>source file - e.g. in resourcestrings).
Hm... It's useful, but can't be done good yet.
>> Also, msgcompiler can be included to LanguageManager of program, so no
>> any additional external unils will be required. I consider this is
>> best solution.
> OK. Graphical programming is not really my domain. Do it as you
>think.
This is not graphical depened thing. Problem is: Any Sibyl's program will use SPCC.
So, we must select such nls method which can be easely used for any program. If
SPCC uses one NLS method and program uses another then too many problems with
localication. (This is like WarpIN. Strings in *.tmf file and in resources. And *.wpi
files. And during install showed messages as from one as from another places)
>> BTW, I prefer to use cross-compiler tools for cross-compiler
>> libs/programs, but not for native. Cross-compilation limits in
>> developing and, in most cases, cross-platform things slowly. This is
> I'm talking about cross-compilation here, i.e. possibility to
>compile native OS/2 executables e.g. on Linux, etc.
Oh... This is impossible for good-coded PM/SOM applications. For pure PM it is
possible. But not for SOM/WPS.
>> like gnu development tools: I must use stupid sh, perl, awk, etc.
>> instead of usual rexx, alway execute one, two, three and more tools,
>> instead of one.
> Portability (this is what you were talking about most probably)
>makes it possible to have more OS/2 software (software written by
>people who don't really use OS/2 themselves).
I don't see any problems with ported software. I only see problems with
stupidly-ported software. Currently we have various tools, but too hard make working
environment.
This is because various porters ports in various ways. May be UnixOS/2 solves such
type of problems, but I think no. They can't compile such easy thing as GNU Gettext
with Andy's iconv (5Kb of source!).
> This is something I consider quite important on a platform lacking much software
>(I've
>tried to find an OS/2 DVD video player to no avail now for example).
Try WarpVision/WarpVisionGUI (BTW, this is good example of mix native code and
cross-platform code)
>"Portable" doesn't mean "inefficient" if you design it properly.
Of couse. But in most cases no any "design". (Lazarus is one example).
-----------
To unsubscribe yourself from this list, send the following message
to [EMAIL PROTECTED]
unsubscribe sibyl
end
