BarryG wrote: ↑Wed Dec 15, 2021 5:42 am
I agree with you on this. I'd be happy if everything but Windows was ditched and Fred concentrated only on making a Windows-only version. I know it won't happen, though. But it would be way less work and stress, and free up so much more time for him. I know from experience. There's a line you have to draw between being all things for everyone while making sacrifices for each, or just making one cool product that does everything well with no sacrifices.
If you are looking for a perfect solution for Windows switch to Visual Studio. I think the good thing about PB is that it is cross-platform and thus the experiences are also shared across platforms. For the few individual cases where something does not work right away, there is also a solution.
Fred has written a very good software and united a very good community I hope he keeps this way because I appreciate it to be able to work across platforms.
I'm okay with the decision to be cross-platform; I just wish it wasn't so. Half my code is API stuff to compensate, and that's fine. It gets the job done.
I use Windows, Linux, MacOS and from the weekend (when it arrives) also Raspberry.
I don't want to miss the multiplatform in any case. For special cases there is also the Compiler Directive PB_Compiler_OS. I have already written applications that run on all OS.
mk-soft wrote: ↑Wed Dec 15, 2021 6:33 pmI don't want to miss the multiplatform in any case
My comment wasn't a request for it to happen. Just a pipe-dream personal wish that I know isn't going to happen. I've removed my lamentation because it's making people defend cross-platform when that wasn't my intention.
- Added /OPTIMIZER flag to the commandline, and a 'Use code optimizer' checkbox in IDE 'Compiler Options' to turn on C optimizations
I noticed that compilation with optimization enabled takes longer even if using the debugger.
Shouldn't optimization be automatically disabled (level -O0) with the debugger enabled?
Not really. Because there can be differences in the result and thus one can check it.
For example, Procedure.b foo() as a boolean return value does not work with optimisation. Boolean as return value should always be a whole register! (Integer)
The files "PureBasic_Linux_beta_ARM32_6.00_(xxxxxx).tgz" and "PureBasic_Linux2_beta_X64_6.00_(xxxxxx).tgz" are identical to those of November's Beta 1. Is it supposed to be like that?
Cezary wrote: ↑Wed Jan 12, 2022 4:47 pm
The files "PureBasic_Linux_beta_ARM32_6.00_(xxxxxx).tgz" and "PureBasic_Linux2_beta_X64_6.00_(xxxxxx).tgz" are identical to those of November's Beta 1. Is it supposed to be like that?
Confirmed: all files in archive date 2021 Nov 25. ARM version seem still Beta 1
Testing: