On Thu, 2015-08-06 at 12:43 -0600, Jonathan Corbet wrote: > On Thu, 16 Jul 2015 08:31:59 -0600 > Jonathan Corbet <[email protected]> wrote: > > > On Thu, 16 Jul 2015 15:28:56 +0100 > > Ben Hutchings <[email protected]> wrote: > > > > > On Thu, 2015-07-16 at 23:22 +0900, Masanari Iida wrote: > > > > make htmldocs on linux-next have some parser warnings. > > > > Following commit cause the warning. > > > > > > Yes it has warnings but they should be non-fatal. > > > > Even so, we don't really want to be adding warnings to the build. > > Hopefully it shouldn't be too hard to fix these...? > > So I've not heard any further on this.
Sorry about that. I didn't forget about it but didn't have a good idea how to fix it. > I think at this point I'm going > to revert the patch in question; adding warnings for everybody is just > not the right thing to do. The other patches in the series can stay. OK. > Having spent some time pondering on the issue, I've also concluded that > this isn't quite the right solution to the problem. Better, I think, to > have docproc filter out the duplicates directly, and to direct the output > to a different file type (.fxml, say). Otherwise the .xml files can have > two different things depending on which type of docs was built first, > leading possibly to other formats being built with missing stuff. > > I may hack together a solution along these lines, just because I do think > that the reproducible builds goal is an important one. But I've got a > *bunch* of other stuff that I have to get done, so I'm not quite sure > when I can get to it. I've now thought of another approach, which is to keep the duplicates during the build process and de-duplicate when installing. I've just sent a patch along those lines. Ben. -- Ben Hutchings Sturgeon's Law: Ninety percent of everything is crap.
signature.asc
Description: This is a digitally signed message part
