[Cuis-dev] Morphic glitches out/hangs when adding faulty code to drawOn method on morphs that step

[Alan Dao]

Hi Juan,

Thanks a bunch for looking into this! I've tested on Sample09Clock and the
error handling works properly now. Unfortunately, I tried testing on my
60Hz step morph and still see issues. The image doesn't hang and I
eventually see a debugger, but the Morphic rendering of the whole world
breaks for the rest of the image runtime. I tested this in Sample09Clock
and can reproduce it there. To reproduce, put the previous faulty code into
the drawOn method add this method to Sample09Clock instance:

```
stepTime
    ^ 17
```

Then create the clock.

Looks like the fix is prone to race conditions :(

Alan

On Thu, Dec 15, 2022 at 2:54 PM Juan Vuletich <juan at cuis.st> wrote:

> Hi All,
>
> I finally could reproduce and understand the problem. Fix is now at GitHub.
>
> Thanks for reporting!
>
> Cheers,
>
> On 12/15/2022 9:50 AM, Juan Vuletich via Cuis-dev wrote:
>
> Hi Folks,
>
> I guess I need more precise steps to reproduce.
>
> I just fired a fresh Cuis, from an updated repo. Just installed updates
> and opened a Sample09Clock. It works ok. Then I open a browser on it (from
> the halo), went to the drawOn: method and added your line Alan. I get a
> debugger "Instances of SmallFloat64 are not indexable". I close the
> debugger and see the clock displayed now as a red box with yellow borders
> and cross. Image is responsive. I go to the browser and correct the method.
> Then I do World / Debug / Start drawing all again. The clock is correctly
> drawn again.
>
> In short, I see the expected behavior. This means I need enough detail, so
> I can reproduce the problem.
>
> Thanks,
>
>
> On 12/15/2022 6:28 AM, Hilaire Fernandes via Cuis-dev wrote:
>
> Hi Alan,
>
> Thanks to report! I confirm the issue, even with non stepping morph. It
> should not behave like that.
>
> The expected behavior with faulty code in a #drawOn: method is a broken
> morph rendered as a red rectangle with yellow diagonals and stepping set in
> pause, to avoid recurring debugger to pop up I guess. By the way there is
> the WorldMenu>Debug  and Morph halo menus to restart stepping and rendering
> once the broken code is fixed.
>
> It seems the Morph is not detected as broken, and rendering is tried again
> and again.
>
> Juan recently introduced the HybridCanvas to speed up graphics rendering,
> may be it is related, and a fix should come. I can't help more, sorry.
>
> Hilaire
> Le 15/12/2022 à 07:43, Alan Dao via Cuis-dev a écrit :
>
> I know that the code I've put in is inherently flawed, but my development
> process involves experimenting a lot and not always knowing what I'm doing.
> Is there something I'm missing when it comes to not having mistakes break
> Morphic? Any advice would be appreciated.
>
> --
> GNU Dr. Geohttp://drgeo.euhttp://blog.drgeo.eu
>
>
>
> --
> Juan Vuletichcuis.stgithub.com/jvuletichresearchgate.net/profile/Juan-Vuletichindependent.academia.edu/JuanVuletichpatents.justia.com/inventor/juan-manuel-vuletichlinkedin.com/in/juan-vuletich-75611b3twitter.com/JuanVuletich
>
>
>
> --
> Juan Vuletichcuis.stgithub.com/jvuletichresearchgate.net/profile/Juan-Vuletichindependent.academia.edu/JuanVuletichpatents.justia.com/inventor/juan-manuel-vuletichlinkedin.com/in/juan-vuletich-75611b3twitter.com/JuanVuletich
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.cuis.st/mailman/archives/cuis-dev/attachments/20221215/9e9da01e/attachment-0001.htm>