[fpc-devel] FPC AVR embedded

[Florian Klämpfl]

Am 19.04.2019 um 10:37 schrieb Christo Crause:
> On Thu, Apr 18, 2019 at 4:40 PM Florian Klämpfl <florian at freepascal.org <mailto:florian at freepascal.org>> wrote:
> 
>     Am 18. April 2019 12:07:33 schrieb "Dimitrios Chr. Ioannidis via fpc-devel" <fpc-devel at lists.freepascal.org
>     <mailto:fpc-devel at lists.freepascal.org>>:
> 
>     > Hi,
>     >
>     >   I'll appreciate if you can tell me if the AVR target will continue to
>     > be supported. Is there a maintainer for this target ?
> 
>     Define maintainer. I did most of the port,  I fix bugs from time to time and apply patches.
> 
>  
> Florian, there is no doubt that you are active in many (probably all) aspects of FPC, including AVR & embedded. My
> concern is that there is no other regular maintainer/developer for this part of the compiler currently.  Jeppe was quite
> active in contributing to AVR a couple of years ago, but his focus seems to have shifted to developing FPC support for
> other architectures. While you are often quick to react to bugs and evaluating patches, there are a couple of other
> instances where reported issues have been left hanging in an unresolved state.  Again, I mean no criticism here, I just
> want to point out that despite your prolific support some issues fall off the radar.

The best approach is in this case imo asking at the mailing list, if an issue is really important.

> 
> It is my perception that a platform maintainer would focus on tracking, fixing, patching or re-assigning platform
> specific issues as required to ensure that issues don't get forgotten.  The alternative practice of bumping issues to
> keep them on the radar is not my preference.

Yes, this is exactly what I wanted to point out. In the strict sense, no target has a maintainer. The same you mentioned
applies even to the "first tier" targets of FPC. There is not really an official win32/win64 or linux maintainer. There
are people typically building the releases for a certain target, but that's it. One reason is also that the scope of the
FPC sources is so broad that not everybody can cover everything. While I can fix compiler stuff and basic rtl stuff for
almost all targets probably comparable easily, fixing e.g. data base related stuff requires much more effort for me
because I am not familiar with it. So even if I was called the win64 maintainer, I would not touch database bugs on win64.

At the end of the day for a community driven project OSS like FPC the situation is: you have the current state available
and the sources. What happens in the future, is unpredictable.

Still, I am interested in the AVR target because I think keep FPC working on a little platform (resource-wise: memory, 8
bit, RISC) helps FPC in general.

> 
> Anyway, thank you for the progress of AVR support so far. I think there is still scope for improvement and feature
> growth both for AVR specifically and embedded in general.