[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [MiNT] wind_get/set mode



Hello,

Sorry for my late answer. The discussion was about the constant:

#define WF_FIRSTAREAXYWH 13

we introduced in XaAES and MyAES.

The point is that some applications used that value (when it was undefined) : the application checks the error code returned by the AES when invoking wind_get(13), and then the application deduces from the error code returned if the AES supports "modern" wind_set/get extensions.

It seems that for now, some applications use that feature to test the availability of WF_OWNER.

On Tue, 1 Feb 2005 21:46:00 +0100, Gerhard Stoll <Gerhard_Stoll@b.maus.de> wrote:

AB> then this old application may believe that WF_OWNER is not supported
AB> by this AES.

The application can believe that no new wind_get/set mode is present.

Please remember that the only legal way to check if new wind_get/set() mode is present is by calling appl_getinfo().

All modern applications should use appl_getinfo() instead of the wind_get(13) hack.

Now, if the AES doesn't support appl_getinfo()... then wind_get(13) may be used to check the availability of "new" modes for wind_set/get, and that's not a problem because all the AES that support wind_set(WF_FIRSTAREAXYWH) also support appl_getinfo().

Furthermore, if we look at the consequences: applications that check the availability of new wind_set/get modes only by using wind_get(13) may believe that WF_OWNER is not supported by MyAES/XaAES. That's no a so terrific consequence for old or bad programmed software.

Other point: that should only concern old softwares that are not aware of the existence of appl_getinfo()... so such applications know nothing about very new modes of wind_get/set() which appeared after appl_getinfo().

To sum up, i would vote to keep WF_FIRSTAREAXYWH as is (value already supported by MyAES and XaAES at the moment).

If someone think it's really a bad idea to keep WF_FIRSTAREAXYWH=13, then speak now. I'd like to release a new version of gemlib, with WF_FIRSTAREAXYWH defined inside... and then other pieces of software will use that value too. Then, it will be very hard to change WF_FIRSTAREAXYWH to another value.


best regards,
Arnaud.