New Dispatcher Issues

[d.healey]

@Christoph-Hart Seems to be working here on Linux. I'm having some other issues but I think they are related to a merge I did as using vanilla new_dispatcher branch has no problems.

[ulrik]

@Christoph-Hart said in New Dispatcher Issues:

Hmm, compilation works fine here. Can you compile the demo project from the HISE repo?

Which demo project?

[d.healey]

@ulrik https://github.com/christophhart/HISE/tree/new_dispatcher/extras/demo_project

[ulrik]

@d-healey Thanks!

[ulrik]

@Christoph-Hart
Comiling the "Demo Project" result:
Xcode 14.3.1

[d.healey]

@ulrik Are you using the new_dispatcher branch? I was getting that error in my fork where I merged the new_dispatcher in, but with Christoph's new_dispatcher it worked.

@Christoph-Hart What do these do?

NOMINMAX=1 
WIN32_LEAN_AND_MEAN=1

[ulrik]

@d-healey said in New Dispatcher Issues:

@ulrik Are you using the new_dispatcher branch? I was getting that error in my fork where I merged the new_dispatcher in, but with Christoph's new_dispatcher it worked.

Yes

[ulrik]

@ulrik I compiled again and now it worked...

[ulrik]

@ulrik @Christoph-Hart but if I compile one of my projects, that easily compiled with the "develop code", it fails with "No global header file was found" using the NewDispatcher code, how come?

[Christoph Hart]

1. Try deleting the entire Builds folder
2. Try a new empty project

[ulrik]

@Christoph-Hart Ok, I found the issue, the project I was compiling had some files in the "AdditionalSoureCode" folder which doesn't belong to the project.
I deleted it and now it's compiling!
Thank you Christoph!

[d.healey]

Building HISE on Linux is currently segfaulting. (Using latest commits)

[Thread debugging using libthread_db enabled]Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".Programreceived signal SIGSEGV, Segmentation fault.juce::CharPointer_UTF8::isEmpty (this=0x2f7e258 <juce::juce_xmltextContentAttributeName>) at ../../../../JUCE/modules/juce_core/text/juce_CharPointer_UTF8.h:7373         bool isEmpty() const noexcept
              { return *data == 0; }

[d.healey]

@d-healey I just reenabled the Linux Font Handler (I usually have it disabled because it slows down HISE startup). And I get a bit further before the segfault

[Christoph Hart]

jupp, that's on me. Weird that it crashes though. Have you changed the HISE_USE_NEW_CODE_EDITOR preprocessor? That would explain it, otherwise...

[d.healey]

@Christoph-Hart It seems to be the fix autocomplete commit where the issue started, the commit before that works.

[d.healey]

Develop branch still segfaulting on Linux. Anything I can do to help narrow it down?

[d.healey]

@Christoph-Hart said in New Dispatcher Issues:

Have you changed the HISE_USE_NEW_CODE_EDITOR preprocessor?

Oh somehow I totally missed your question. I haven't touched that. I'm just building the unmodified develop branch straight from your repo.

[d.healey]

@Christoph-Hart Still segfaulting on Linux with the latest develop branch

[d.healey]

Maybe this is useful...

It's returning a nullptr from juce::Component::findParentComponentOfClass<hise::ComponentWithBackendConnection>()

So is there some component not being created when it should be?

Segfault occurs on line 95

[d.healey]

I've narrowed it down to this commit - https://github.com/christophhart/HISE/commit/b2471b5bbe54fe177ee43eb57a4a7839bcd0aa48

Commenting out line 118 in ScriptingPanelTypes.cpp gets rid of the segfault, not sure why though...