[L2Ork-dev] Won’t finish starting up

Jeff Morris jeff at morrismusic.org
Mon Oct 21 13:16:51 EDT 2024


Thank, Albert and Ico,

This is enlightening information about the lower-level state of things.

So what do you guys use to get game controllers to work with your patches
if not [hid]?

Any idea what's up with [cyclone/mousestate]? It loads without error but
never produces any output, and I don't see anything in the DevTools
console. Clicking [poll( sometimes produces these messages in the Pd
console:

legacy tcl command at 473 of ../shared/hammer/gui.c: global
hammergui_ispolling
legacy tcl command at 474 of ../shared/hammer/gui.c: set
hammergui_ispolling 1
legacy tcl command at 475 of ../shared/hammer/gui.c: hammergui_poll

And yes, for Purr-Data 2.20, I tried removing the quarantine attribute
first and it wasn't until I used spctl that I got it to work.

Thanks,

𐌌

Jeff Morris
http://morrismusic.org
http://weblogmusic.org


On Mon, Oct 21, 2024 at 2:57 AM Albert Graef <aggraef at gmail.com> wrote:

> Hi Jeff,
>
> On Sun, Oct 20, 2024 at 7:15 PM Jeff Morris <jeff at morrismusic.org> wrote:
>
>> I was able to get Purr-Data working finally using sudo spctl
>> --master-disable and it mostly works okay, except that it crashed upon
>> quitting. The macOS console crash log shows the following. Looks like a
>> memory segmentation fault experienced by nwjs.
>>
>
> Completely disabling Gatekeeper shouldn't be necessary. At least with the
> latest 2.20.0 version, just removing the com.apple.quarantine attribute
> after installation should be enough (cf.
> https://github.com/agraef/purr-data/wiki/Installation):
> sudo xattr -r -d com.apple.quarantine /Applications/Purr-Data.app
>
> The nw.js crash-on-exit is a known bug:
> https://github.com/agraef/purr-data/issues/44. You can work around the
> annoying popup message from the OS as described in the ticket. Other than
> that, this bug is mostly harmless since the nw.js crash happens after
> purr-data already went through its entire finalization sequence. And you
> can also avoid the bug by closing all patch windows before exiting the
> program. At least that works for me on Big Sur (old Intel MB) and Sonoma
> (MB M1), your mileage may vary.
>
> As Ico already pointed out, there are various issues with the nw.js
> versions that we use. Probably the most serious regression is that NW2, the
> new browser window implementation being used as default in nw.js 0.42.4 and
> later, has some severe performance problems with dynamic code execution in
> the JavaScript interpreter. This makes rendering dynamic svg graphics on
> the canvas (such as array displays and pd-lua graphics) horribly slow,
> which is why both pd-l2ork and purr-data switch nw.js to the old NW1
> implementation. (Ico first discovered this and reported it upstream some
> two years ago, but we still need to produce a minimal example for the issue
> before it will be considered.)
>
> This in turn limits the nw.js versions that we can use, because anything
> past nw.js 0.67.1 has broken window management in NW1 mode. The
> crash-on-exit bug on macOS is also related to the use of NW1 AFAICT.
> Moreover, on Linux, nw.js 0.56.0 and later have another serious bug
> (nwworkingdir not working, which means that file dialogs will open in the
> wrong places) which AFAICT has never been fixed. Which is why I decided to
> go with nw.js 0.55.0 in purr-data 2.20.0, which is pretty much the last
> nw.js version that works without any serious regressions on all major
> platforms.
>
> But note that purr-data also includes a little shell script which lets you
> experiment with different nw.js versions and switch between NW1 and NW2, to
> find out what works best for you. See
> https://github.com/agraef/purr-data/wiki/Installation#patching-purr-datas-nwjs.
> (I doubt that the script will work with pd-l2ork as-is, but you can
> probably adjust it if needed, if you know a little shell programming.)
>
> HTH,
> Albert
>
> --
> Dr. Albert Gr"af
> Computer Music Research Group, JGU Mainz, Germany
> Email: aggraef at gmail.com, web: https://agraef.github.io/
> _______________________________________________
> L2Ork-dev mailing list
> L2Ork-dev at disis.music.vt.edu
> https://disis.music.vt.edu/listinfo/l2ork-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://disis.music.vt.edu/pipermail/l2ork-dev/attachments/20241021/8e6d8aba/attachment.htm>


More information about the L2Ork-dev mailing list