Front page | perl.perl5.porters |
Postings from March 2016
On 03/26/2016 06:12 AM, demerphq wrote: > I was wondering why we .gitignore files like pod/perlapi.pod and > pod/perlintern.pod. I know they are autogenerated, but I don't > understand why they are different from the files that are > autogenerated by make regen. > > I guess I could frame this more generally. Why do we do some build > steps "on demand", via make-regen, and why do we do some "in-line", > via make all? > > Can we speed up the build process by checking in more of our files, > and having their regeneration step be "on demand". That way we would > see the diffs, and would have less stuff to build on an average run. > > I noticed this issue when I added some documentation, but did not > notice any diffs after I built, only after manually checking the files > was I able to see that my changes made a difference. I was expecting > it to work much like the files covered by make regen, where i would > see the results of the change. > > At the very least I really would like to add these two docs files to > git, and remove their .gitignore files, but I also think there is a > bigger question here, should we VC more files so we have better > visibility of the changes in our code? I can understand ignoring files > that might differ from install to install, but docfiles? Unicode > tables? > > Yves This might solve the problem that the generated files do not get put up properly online. I can't remember the details right now, but a link that ends up calling a search function will fail to find them.Thread Previous | Thread Next