FYI guiding-resume bug

[Jasem Mutlaq]

Hello Hy,

Good catch. I'm not sure what's the best way to avoid this.. I mean, the
state machine change is correct, but FLIP_PLANNED is called before focus is
done and guiding is resumed, correct?

That variable is set to make sure we are indeed already guiding. What if
guiding was never utilized when capture was started? I recall that before
we set this immediately after you start a sequence, so we know whether
guiding is used or not. But later on, it was moved to check for these
conditions later on.

--
Best Regards,
Jasem Mutlaq



On Wed, Jun 3, 2020 at 10:06 PM Hy Murveit <murveit at gmail.com> wrote:

> I just saw a bug where,
>
> meridian flip is about to got to FLIP_PLANNED
>
> capture starts a focus on filter change, so guiding goes from guiding to
> suspended
>
> capture gets the FLIP_PLANNED, but because guiding is suspended, it sets
> resumeGuidingAfterFlip to false.
>   capture.cpp line 3636 resumeGuidingAfterFlip = isGuidingActive();
>
> then focus finishes, guiding is resumed, then eventually flip happens, but
> because the variable is set that way, guiding doesn't get resumed.
>
> Do you think it's important to not resume guiding on a suspended? I would
> think, pretty much any time we're in suspended then flip, we'd want to
> resume guiding after the flip.
>
> Hy
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/kstars-devel/attachments/20200604/bd63855d/attachment.htm>