Arch Linux User Repository

@dmrz My situation is actually opposite, I cannot build without those modifications. Would you mind to provide your vulkan-headers version? you can run pacman -Si vulkan-headers to find it out. It is 1:1.4.303-1 on my computer.

This package would not build for me. In order to get it to build, I had to make the following change:

diff --git a/PKGBUILD b/PKGBUILD
index 9ad54db..e89cb55 100644
--- a/PKGBUILD
+++ b/PKGBUILD
@@ -112,11 +112,6 @@ prepare() {
     patch -Np1 -i ../001-change-binary-name.diff
     patch -Np1 -i ../002-install-and-load-localdocs-model-more-standardly.diff
     sed -i "s|https://gpt4all.io/models/gguf|file://$srcdir|" gpt4all-chat/CMakeLists.txt
-    sed -i "s/vk::DynamicLoader/vk::detail::DynamicLoader/g" \
-        gpt4all-backend/deps/llama.cpp-mainline/ggml/src/kompute/src/include/kompute/Manager.hpp \
-        gpt4all-backend/deps/llama.cpp-mainline/ggml/src/kompute/src/Manager.cpp
-    sed -i "s/vk::DispatchLoaderDynamic/vk::detail::DispatchLoaderDynamic/" \
-        gpt4all-backend/deps/llama.cpp-mainline/ggml/src/kompute/src/include/kompute/Manager.hpp
 }
 build() {
     CFLAGS+=" -DNDEBUG"

Yes, there seems to have been functions removed from Qt.

Is this busted for anyone else after updating to Qt 6.8? When I try to run it:

gpt4all-chat: symbol lookup error: gpt4all-chat: undefined symbol: _ZN21QHttpServerRouterRuleC1ERK7QString6QFlagsIN18QHttpServerRequest6MethodEESt8functionIFvRK23QRegularExpressionMatchRKS4_O20QHttpServerResponderEE, version Qt_6

When I try to recompile it, it fails at:

[ 55%] Building CXX object CMakeFiles/chat.dir/src/server.cpp.o /home/***/gpt4all-chat/src/gpt4all-3.4.1/gpt4all-chat/src/server.cpp: In member function ‘void Server::start()’: /home/***/gpt4all-chat/src/gpt4all-3.4.1/gpt4all-chat/src/server.cpp:460:20: error: ‘class QHttpServer’ has no member named ‘listen’ 460 | if (!m_server->listen(QHostAddress::LocalHost, MySettings::globalInstance()->networkPort())) { | ^~~~~~ /home/***/gpt4all-chat/src/gpt4all-3.4.1/gpt4all-chat/src/server.cpp:603:15: error: ‘class QHttpServer’ has no member named ‘afterRequest’ 603 | m_server->afterRequest([] (QHttpServerResponse &&resp) { | ^~~~~~~~~~~~ /home/***/gpt4all-chat/src/gpt4all-3.4.1/gpt4all-chat/src/server.cpp: In lambda function: /home/***/gpt4all-chat/src/gpt4all-3.4.1/gpt4all-chat/src/server.cpp:604:14: error: ‘class QHttpServerResponse’ has no member named ‘addHeader’ 604 | resp.addHeader("Access-Control-Allow-Origin", "*"); | ^~~~~~~~~ make[2]: *** [CMakeFiles/chat.dir/build.make:519: CMakeFiles/chat.dir/src/server.cpp.o] Error 1 make[1]: *** [CMakeFiles/Makefile2:359: CMakeFiles/chat.dir/all] Error 2 make: *** [Makefile:156: all] Error 2 ==> ERROR: A failure occurred in build(). Aborting...

Thanks, I figured out what caused this: My makepkg.conf was out of date, I updated and now I have additional parameters.

To be exact, the cause is this part: CXXFLAGS="$CFLAGS -Wp,-D_GLIBCXX_ASSERTIONS, you need the -Wp,-D_GLIBCXX_ASSERTIONS otherwise compilation fails. No idea why though...

@DarkShadow44 I cannot reproduce this problem on my computer, so I think I can do nothing now. Maybe you can setup a chroot environment and build this package?

Fails for me with /usr/bin/ld: ../deps/fmt/libfmt.a(format.cc.o): relocation R_X86_64_PC32 against symbol `stderr@@GLIBC_2.2.5' can not be used when making a shared object; recompile with -fPIC

Previous versions still work, but since you bumped to 3.3.0 on 2024-09-24 it fails.

If I enable this, LocalDoc model will be installed into /usr/resources/nomic-embed-text-v1.5.f16.gguf, this is not a standard path to place package content.

You can see here for more info:

https://github.com/nomic-ai/gpt4all/blob/main/gpt4all-chat/embllm.cpp#L81

Maybe I have to modify its code to use a more standard place like /usr/share/gpt4all-chat/resources/nomic-embed-text-v1.5.f16.gguf. But archlinux do not like applying too many patches.

TL;DR: a model file included in the package is not found (not copied in the expected directory, it seems). as a consequence the LocalDoc feature doesn't work, the embedding stops at 0%

from console:

embllm WARNING: Local embedding model not found
WARNING: Could not load model for embeddings

CONTEXT: the recent versions of gpt4all-chat (from 3.0 I guess) have a LocalDoc functionality that make use of a local embedding model. According to devs on Nomic AI linux-help discord channel, this model is included in the package, the file name should be nomic-embed-text-v1.5.f16.guff In the flatpak directory it is in /var/lib/flatpak/app/io.gpt4all.gpt4all/...../files/resources

@AndyRTR I build this package with pkgctl build and it does have the problem you say. I think I may do more research on how to make pkgctl happy because I use makechrootpkg and it can build this PKGBUILD without problem.

But I must say this is not a zsh-only declare option, -r is supported by bash's declare command: https://www.gnu.org/software/bash/manual/html_node/Bash-Builtins.html

Edit: After I did some research, I found that the PKGBUILD is sourced when code is executed on the line 334 of /usr/share/devtools/lib/build/build.sh, and sourced again on the line 37 of /usr/share/devtools/lib/util/pacman.sh, which results that bash tries override a readonly variable and throws an error. Maybe this is pkgctl's problem?