SC: Faster speed of subs (by loading from memory) [Granted]

[FluxTorpedoe]

✔ v24.30.0116 is working out of the box (and commenting out "_Terminate" in v24.30.0115 prevented the Error 9 indeed)!
No other apparent issues on my side — and the subs are still muuuch faster!

 — Note 1: After investigating, the slow-down with the dynamic script didn’t come from itself, but from the visibility toggle of System or Hidden files!

 — Note 2: I’m also still trying to find out why I have a ~80ms difference when toggling between sessions, between a fresh XY install and my XY settings (with the same Session Manager settings).
Some internal XYplorer settings must come into play… I’ll post the results in Session Manager forum when I find out.

In the meantime, I think I can mark this wish as granted!
Thanks a lot

   Flux

[Filehero]

Hmm, many of my scripts appear to be broken (24.030.0116). It seems to be related to loading dynamically assembled scripts (pattern load($mscript, , "s"); )

Unfortunately, right now I don't have any time for looking into it deeper - will have to roll back.

[klownboy]

Hmm, many of my scripts appear to be broken (24.030.0116)

Hi Filehero. Yes, definitely refer to viewtopic.php?t=26131 and viewtopic.php?t=26132

[admin]

Hmm, many of my scripts appear to be broken (24.030.0116). It seems to be related to loading dynamically assembled scripts (pattern load($mscript, , "s"); )

Unfortunately, right now I don't have any time for looking into it deeper - will have to roll back.

Might be fixed in v24.30.0118.

[Filehero]

Might be fixed in v24.30.0118.

Nope.

script_load_broken-1.PNG (31.6 KiB) Viewed 489 times

script_load_broken-2.PNG (26.83 KiB) Viewed 489 times

"Every" (haven't checked all) script generated and run from within a script gets borked. It's hard to decipher these interpreter messages into an overall pattern but it's somewhat indicating the script scopes/boundaries get "lost" and/or the AST is scrambled.


@Ken: thanks.

[admin]

Shall I OCR your shots or will you provide an actual script for testing?

[Filehero]

Unfortunately, none of those scripts will run on your side ootb.

Don't know when I will find the time to strip them down to the relevant "bones".

[admin]

OK, I think see it now. No time today anymore for the fix. Tomorrow...

[Filehero]

Heredocs?

I started hunting, but my scripts are heavily nested (includes) so it's a real time-swalling pita (commenting/uncommenting by trial&error) to even touch a potential region of troubles.

Currently my (a?) candidate is an "included function" defining a heredoc which in turn defines a script header with further include statements.

[admin]

Recursion is a bitch. Tomorrow...

[admin]

Is it fixed now (v24.30.0119)?

[Filehero]

At work on MacOS. Maybe later.


Edit: with v24.30.0121 the script errors reported above are gone.