(no subject)
So I re-examined OpenOCD today, the latest version as well as a older version that reportedly works for somebody else. The older version kind of works, but is still super-flaky, and the latest version dumps core for every command. A quick examination of the latter reveals brand new code that's transparently, and profoundly, broken.
At this point, I have absolutely no confidence in the OpenOCD developers working on support for STM32 processors. The likely reality is that this project will not have robust support for my board for the foreseeable future. Egads.
So, frustrations aside, I've been hacking on HPDF, a PDF generation library for Haskell. The code is not the prettiest, but the author is competent and the code works. I'm highly partial to these qualities. So far, I've mostly been working on making things prettier. Thanks to my changes, the code is now a bit smaller, a bit faster, and the PDF documents it generates a bit shorter.
At this point, I have absolutely no confidence in the OpenOCD developers working on support for STM32 processors. The likely reality is that this project will not have robust support for my board for the foreseeable future. Egads.
So, frustrations aside, I've been hacking on HPDF, a PDF generation library for Haskell. The code is not the prettiest, but the author is competent and the code works. I'm highly partial to these qualities. So far, I've mostly been working on making things prettier. Thanks to my changes, the code is now a bit smaller, a bit faster, and the PDF documents it generates a bit shorter.