[Ohrrpgce] early_debuginfo

[Ralph Versteegen]

Oh, right, the 1.08.x problem was due to a change to the arguments to
SCREENINFO in gfx_fb. We can certainly work around that. From the changelog:

- Change overload to SCREENINFO( byref as long = 0, byref as long = 0,
byref as long = 0, byref as long = 0, byref as long = 0, byref as long = 0,
byref as string = "" )
- Add overload for SCREENINFO( byref as longint, byref as longint, byref as
longint = 0, byref as longint = 0, byref as longint = 0, byref as longint =
0, byref as string = "" )

I see that we're using size_t for those args, which is unsigned, whereas
long and longint are signed. I'll fix that now so you can use the latest FB.

On Sun, 5 Sept 2021 at 04:04, James Paige <Bob at hamsterrepublic.com> wrote:

> Sure! Which fbc version do we want to target?
>
> I know I am using fbc 1.07.3 on my new linux box because 1.08.x had some
> kind of breakage, I forget what
>
> But I would be happy to get all the nightly build boxes upgraded. I might
> even have time to do that this weekend!. is 1.07.3 an acceptable target for
> now?
>
> ---
> James
>
>
>
> On Sat, Sep 4, 2021 at 11:53 AM Ralph Versteegen <teeemcee at gmail.com>
> wrote:
>
>> The fix I was talking about was just to print a sensible error message
>> instead of a misleading one. I went off on a rant there...
>>
>> However, it would be nice (at some point) to upgrade all the nightly
>> build machines to the latest FB version just to get various bug fixes. In
>> particular, array handling is a bit better (FB now keeps track of whether
>> each array is dynamic or fixed length to print clean errors instead of
>> corrupt memory if you try to redim a fixed array), and you can enable array
>> bounds checking without expensive and pointless null ptr checking.
>>
>> On Sun, 5 Sept 2021 at 03:22, James Paige <Bob at hamsterrepublic.com>
>> wrote:
>>
>>> Haha!
>>>
>>> So maybe the best fix would be for me to upgrade the fbc version on the
>>> nightly build box, huh?
>>>
>>> ---
>>> James Paige
>>>
>>> On Sat, Sep 4, 2021 at 11:01 AM Ralph Versteegen <teeemcee at gmail.com>
>>> wrote:
>>>
>>>>
>>>> On Sat, 4 Sept 2021 at 20:34, Ralph Versteegen <teeemcee at gmail.com>
>>>> wrote:
>>>>
>>>>> Oh, yes, that "Illegal parameter specification" error is confusingly
>>>>> printed when there's a cdecl/stdcall calling convention difference between
>>>>> two places a function is declared/defined. I've been meaning to submit a
>>>>> patch for that.
>>>>>
>>>> Oh, someone already fixed that message in recent FB versions. Oh, turns
>>>> out it was me... back in 2017.
>>>>
>>>>
>>>>> Those errors only happen on Windows because stdcall is the default in
>>>>> FB (and Windows) but not C, while there's no difference on Unix.
>>>>>
>>>>> On Sat, 4 Sept 2021 at 14:33, James Paige <Bob at hamsterrepublic.com>
>>>>> wrote:
>>>>>
>>>>>> Oh, I see, it must have been  r12334
>>>>>>
>>>>>>   Fix ?_debug_archive.txt getting spammed with
>>>>>>> "setup_exception_handler" on Unix
>>>>>>>
>>>>>>>   It was always appended to the log regardless of whether there was
>>>>>>> a ?_debug.txt
>>>>>>>   file to archive.
>>>>>>>
>>>>>>
>>>>>> I can see that it changes the definition of early_debuginfo to be
>>>>>> extern C. I guess that is what broke on the Windows builds.
>>>>>>
>>>>>>
>>>>>> On Fri, Sep 3, 2021 at 8:12 PM James Paige <Bob at hamsterrepublic.com>
>>>>>> wrote:
>>>>>>
>>>>>>> I am puzzled. The Windows nightly builds show multiple failures on:
>>>>>>>
>>>>>>> *common_base.bas(28) error 40: Illegal parameter specification in
>>>>>>> 'SUB early_debuginfo (msg as const zstring ptr)'*
>>>>>>>
>>>>>>> But none of the lines that define or use early_debuginfo have been
>>>>>>> toiuched since like 2019. I wonder why it started failing recently?
>>>>>>>
>>>>>>> ---
>>>>>>> James
>>>>>>>
>>>>>> _______________________________________________
>>>>>> Ohrrpgce mailing list
>>>>>> ohrrpgce at lists.motherhamster.org
>>>>>> http://lists.motherhamster.org/listinfo.cgi/ohrrpgce-motherhamster.org
>>>>>>
>>>>> _______________________________________________
>>>> Ohrrpgce mailing list
>>>> ohrrpgce at lists.motherhamster.org
>>>> http://lists.motherhamster.org/listinfo.cgi/ohrrpgce-motherhamster.org
>>>>
>>> _______________________________________________
>>> Ohrrpgce mailing list
>>> ohrrpgce at lists.motherhamster.org
>>
>>
>>> http://lists.motherhamster.org/listinfo.cgi/ohrrpgce-motherhamster.org
>>>
>> _______________________________________________
>> Ohrrpgce mailing list
>> ohrrpgce at lists.motherhamster.org
>> http://lists.motherhamster.org/listinfo.cgi/ohrrpgce-motherhamster.org
>>
> _______________________________________________
> Ohrrpgce mailing list
> ohrrpgce at lists.motherhamster.org
> http://lists.motherhamster.org/listinfo.cgi/ohrrpgce-motherhamster.org
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.motherhamster.org/pipermail/ohrrpgce-motherhamster.org/attachments/20210905/13f1aefe/attachment.html>