RE: Can I get my progs to run on WIN95
Hi David,
thanks for keeping this can of worms propped open 
I agreee that this is
a real issue that must be dealt with, even if it is hard.
>I understand the difficulty of supporting multiple versions of
>Windows, but
>how much of the library are people locked out of if they don't
>have the most
>current release of Windows?
I assume you don't want this answered in percentage terms And I guess
that what you're really saying is that people should not be locked out of
parts of the library if they choose to run an old version of Windows.
> For example, I still use Windows 95 at
>home and
>work, and won't beupgrading any time in the near future.
>
I don't run Win95 or Win98 any more. I can test on WinMe and Win2000. Are we
expected to test on all possible platforms before releases? What is an
acceptable limit here? If anybody wants to be true alpha/beta testers, now
is a good time to speak up.
>Even more importantly, will people *know* that they are locking
>out users if
>they make certain calls?
Are we talking Win32Lib routines or API routines? I believe that for
Win32lib routines, we should provide that information. And for API routines,
they should consult the appropriate external sources such as Microsoft or
the many 3rd party books on the subject.
All the API calls are documented in the Microsoft website
(http://msdn.microsoft.com) and so are "freely" available, in theory. But is
it the win32lib custodians' responsibility to duplicate some of the
Microsoft documentation, in the win32lib documentation, about which platform
the various API calls are supported on. I will accept that we should add
this sort of information to the win32lib routines, but the API calls too?
And I could add to the docs which API calls (if any) are being used in each
Win32Lib routine.
> For example, I'm sure that Judith would
>prefer that
>the IDE run on as many versions as possible - not just the latest and
>greatest version of Windows.
>
What about 16-bit Windows. Should we try to support these (v3.1, 3.11) as
well? Again, what is the acceptable limit?
>If something is supportable, it seems to me that some effort
>should be made
>to have a fallback routine available. For example, when I first started
>coding the library, a number of bitmap related routines were not available
>under versions of NT. As a result, I made a couple fallback routines that
>would automatically kick in if the native API calls were not available.
>
That's wonderful, and g'donya. I believe you are also coming from the angle
that Win32lib is meant to be a platform independent library, not one that is
wedded to Microsoft Windows. I didn't realise this when I started helping
you out, David (the library's name threw me). The current library is
definitely NOT platform independent and some significant rework would be
need to make it so. I trying to formulate a plan on how to achieve this with
minimal impact to current applications. But in the mean time, if anyone
wants to supply emulation of new Windows functionality for old Windows
versions, please feel free. I just haven't got the time. Its hard enough
dealing with the official, and current, Windows API.
>I'm not saying that advanced features *shouldn't* be available, just that
>people should know when they use them that their code won't run on certain
>platforms.
The term "advanced" is overloaded with undue meanings. Be they advanced or
not, if currently available API features are not available on older
platforms, including ones that have been discontinued by Microsoft, how do
the Win32lib library custodians make this known, or cater for these, without
excessive effort? I'm open to suggestions people.
---------
cheers
Derek Parnell
|
Not Categorized, Please Help
|
|
