Re: experimental manual source for advanced translation handling

> i've done some experiments on one of the topics > we had discussed here before the list went down > > all the files in the function directory have been split up > in a general part and a subdirectory containing the actual > refentries, one function refentry per file This sounds like a good idea to me, as you already argue. That the xml processing is slowed down is a pity, but it is IMO not important. It's the output and the ease of modifying /translating that counts, not the number of cycles wasted on it. > i'm not sure yet wether we should follow the way presented > here, which would break CVS history a bit (which is not as > important for documentation as it is for actual code) > or whether it would be better to leave the xml files as > they are and have them split up and recombined in the > background during the actual built, but that is a different > story ... The latter option is IMO best: you do not break CVS, you do not have to change all files, and you can still edit the files in a module-basis, rather than a single file for every minor function. For translation it is less useful, you'll need to get single functions from the english sources when you want to translate a new function, rather than simply copying a file. Jeroen