PureBasic Forums - English

Hi Folks,

It's here. This version was supposed to be a small update, but the linker change took more time and raised more errors than expected. All seems ok again, so even if the visible changes aren't that important, the compiler back-end is now ready to quickly move forward again. It includes many small improvements and bugs fixes which makes this update a must have. You can grab it on your purebasic account. If you have a slow connection, you can get the new SmartUpdate version on the purebasic account to allow a minimal update (1.4 Mb instead of 7 Mb when updating from the v3.91 beta 2). Read the history section for a full list of changes.. The linux will follow very quickly.

Once again, thanks to fr34k, Berikco and all others which made this release possible !

Have fun,

Fred.

Thanks Fred!

Thank s Fred, all the hard work is always appreciated.

Yup.

Thanks Fred and the crew !!!!

Changed: WindowMouseX() and WindowMouseY() now returns the coordinates in the window client area (DesktopMouseX() now replaces the old behaviour).

That's not true and I find it sort of a pity that there is no true replacement of the old behaviour. This renders a great part of my old sourcecode unusable.
Why not introduce ClientMouseX/Y and DesktopMouseX/Y, leaving WindowMouseX/Y untouched?

Thanks Fred & Team.

I'd just like to say that the smart updater tool is excellent.

freedimension wrote:

Changed: WindowMouseX() and WindowMouseY() now returns the coordinates in the window client area (DesktopMouseX() now replaces the old behaviour).

That's not true and I find it sort of a pity that there is no true replacement of the old behaviour. This renders a great part of my old sourcecode unusable.
Why not introduce ClientMouseX/Y and DesktopMouseX/Y, leaving WindowMouseX/Y untouched?

WindowMouseX/Y() like it was before was unuseable due to skin problem (title bar and borders aren't fixed) and will result to bad result on non standard configuration. You can easily emulate the old behaviour with DesktopMouseX() and WindowX() combo (I agree the statement isn't exactly true). What are the case where the old functions are needed (just curiosity) ?

Well, you are right Fred. The possible applications for the old behaviours are very ... few and far between.
But the situation now is that I have to look through thousands of lines of code to replace occurences of the MouseX/Y functions where I already bypassed the problem you mentioned. There are many complex calculations in that too, regarding that I've written a graphical demanding application. Those can be hard to "re-think".

But well, the child has already fallen into the well (as we would say in Germany)

Thanks Fred & Sons... you rock!

- np

Cheers Fred

Since noon I am using 3.91 without any bug found in my apps.

Great.

Thanks a lot Fred.

for writing libs:
what version of PellesC is recommended ?
will lcc still work ? (i seem to recolect a mention to use Ansi-C)?
any other compilers that are known to produce libs that work with PB 3.91 ?

the latest version (2.80) of PellesC wont allow stdcall, and i get a lot
of polink errors when trying to use a lib compiled with it.
how about any Pascal compilers?
any help is much appreciated.

Fred wrote:Hi Folks,
It's here.

Still no PureBasic_Update_3_91.zip. Will all the future updates only be available via Purebasic user account ?

Fred wrote: If you have a slow connection, you can get the new SmartUpdate version on the purebasic account to allow a minimal update

Why not packing the purebasic releases with RAR or 7-ZIP (please not ZIP, it's obsolete) ? This will save some hundreds kb for the ones of us with slow connections ...

Yes, thanks for the free updates.

+ i wanted to stress that current installers are 100% satisfying, i meant the full 7mb updates, keep those in

These installers perform very well, i run them from within the unzipper and i have no worries with manual unpacking first, i like that.

O btw, i made an error by dl'ing the 390 since this was the 'only' link in public.
You might want to remove that completely (imo).