home bbs files messages ]

Forums before death by AOL, social media and spammers... "We can't have nice things"

   linux.debian.bugs.dist      Ohh some weird Debian bug report thing      28,835 messages   

[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]

   Message 28,022 of 28,835   
   Faidon Liambotis to Boyuan Yang   
   Bug#1128252: ITP: uharfbuzz -- Streamlin   
   17 Feb 26 19:50:01   
   
   From: paravoid@debian.org   
      
   On Tue, Feb 17, 2026 at 12:55:16PM -0500, Boyuan Yang wrote:   
   > No problem; it was my (bad) habit to do 100% of packaging work first,   
   > file ITP, and then immediately upload the prepared package to the NEW   
   > queue with the ITP number. I wish I had filed the ITP earlier this week.   
      
   Yep, seems like both of us have the same bad habit :)   
      
   > I remember the (Python) Team had the conclusion that pypi was not the   
   > preferred point for retrieving source code, and the actual upstream source   
   > code repository (GitHub in this case) should be used whenever possible.   
   > This is exactly what I did in https://salsa.debian.org/fonts-team/uharfbuzz/   
   .   
      
   It's a minor point, but I was not aware of that -- do you happen to have   
   a reference? In any case...   
      
   > > 2) I also stripped the tarball from the pregenerated Cython code   
   > > (src/uharfbuzz/_harfbuzz.cpp, src/uharfbuzz/_harfbuzz_test.cpp) through   
   > > Files-Excluded, as well as debian/clean.   
   >   
   > Repacking will not be necessary if we directly take the upstream source   
   > code from GitHub.   
      
   ...that's a pretty good reason to fetch from GitHub instead of PyPI,   
   indeed!   
      
   > All fonts in uharfbuzz are test data, and are not shipped to the end   
   > user anyway. In this case, I don't think we should bother pursuing   
   > the preferred source of modification with patches since they do not   
   > affect our delivered .deb file in any way. On the other hand, preserving   
   > upstream unit tests ensures quality test in post-build tests and   
   > autopkgtest. As a result, I did not repack to strip the font files,   
   > but opted to document the full license of test data. Let's see what   
   > the DFSG Team may say about it.   
      
   I understand the reasoning (you even identified a real failure through   
   those tests, that I didn't). I also understand that it comes down on how   
   pedantic you want to be with the DFSG.   
      
   Note however https://github.com/harfbuzz/uharfbuzz/issues/234 -- I don't   
   think we have the full license for all fonts, sadly.   
      
   > > 4) I think you need python3-all and python3-all-dev, not   
   > > python3/python3-dev.   
   >   
   > Oh, this is not the case. If you read into the build system, you will   
   > find out that uharfbuzz enabled the CPython limited API [1], which (largely)   
   > ensures ABI stability across different supported python3 versions. Building   
   > with python3-all-dev will thus have no effect, and even raises a warning in   
   > the build log. You should also observe through the generated dependency   
   > that dh_python3 did not have a python3 dependency version range due to   
   > the use of PY_LIMITED_API.   
      
   Oh I didn't notice that. That's great :)   
      
   > > 5) I think my extended description is a bit nicer :) I did not know the   
   > > source extended description trick though! TIL.   
   >   
   > You are welcome to integrate this part into future packaging once   
   > uharfbuzz cleared the NEW queue.   
      
   Will do.   
      
   Faidon   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)   

[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]


(c) 1994,  bbs@darkrealms.ca