An updated version of Spice3f5 is available, which now supports SPICE2-like POLY E,F,G,H sources and also a working PZ (pole/zero) analysis mode. The former are translated internally to arbitrary nonlinear sources, with no limit on degree. The latter was in the original 3f5 code, but suffered a number of problems, often failing to achieve a…
Here’s a handy way of loading a techfile (or in fact running any python script) on startup. Glade reads a .glade.py file, if it exists in the user’s home directory, when it starts. On WIndows, the env var HOME defines the user home dir. So you can load a techfile e.g. # .glade.py – load…
Continue reading Automatically loading a techfile at startup
The Qt docs are pretty hopeless when it comes to describing how to build Qt6 from source if you need the documentation (QDoc) to compile. This is because it needs LLVM, and that is not present for a WIndows distribution. So here’s how to do it. We are going to build LLVM 20.1.0, not 18.1.8…
There have been a few build issues with the 64bit Windows builds getting the 32 bit Qt libraries, which caused them to fail. These have been fixed and tested now. but please do report any issues with the builds if they occur again.
There was a bug introduced in version 5.0.64 / 4.7.103 that caused reading a library to either crash or fail to correctly set up techfile layers, Please update to version 5.0.66 / 4.7.105 asap.
The Mac M1 build had a few issues. The Python library path was incorrectly set, also a couple of the libraries and Gemini were built as x86_64 code, not arm64. This has now been fixed.
The Mac versions of Glade have now been signed, so Gatekeeper should not use randomised path execution of the app (and hence fail to correctly initialise Python or read the basic library). Note you may still get a popup saying ‘Glade can’t be opened because Apple cannot check it for malicious software’. If you do,…
Ubuntu 22.04 and Mint 21.1 (Cinnamon) have been added to the build list as native builds. Enjoy!
The long awaited port of Glade to native M1 code on the Mac has finally happened. Preliminary results show it’s about 50% faster than the equivalent x86 Mac or emulation under Rosetta. There may be a few build dependency issues, if so please report them.
An incorrect rpath to the Python framework on the most recent version caused a crash on Mac OSX. This has now been fixed and a new build uploaded. Thanks to those that reported it.