- From: Robin Berjon <robin@w3.org>
- Date: Tue, 26 Feb 2013 16:46:27 +0100
- To: Michael Cooper <cooper@w3.org>
- CC: Karl Dubost <karl@la-grange.net>, "spec-prod@w3.org" <spec-prod@w3.org>
On 26/02/2013 15:35 , Michael Cooper wrote: > Robin Berjon wrote: >> For (1) there's Michael's objection that he likes to post-edit in an >> XML editor. I'd love to hear about what kind of post-editing we're >> talking about because the point is sort of that there shouldn't be any ;) > There can be any number of reasons to need to post-edit the result of > automatic generation. A couple that I've had to deal with recently on a > repeated basis: > > 1. Respec outputs some weird attributes like 0="" which cause the > document not to validate and that weren't present in the source > document, so I have to remove those. That's not something I've seen before and it's quite surprising (validity bugs tend to be immediately reported). Can you send me a spec that shows that error? If it's there it's a bug. > 2. Respec puts a paragraph with public comment instructions in the > Status section that is not compatible with our process. I always > have to remove that; the correct instructions are in the custom > paragraphs I add to the status. Can you send me examples with what you want? That seems like a feature others would use. > 3. There are times I prefer to tweak aspects of the output like the > heading number format, or sometimes the TOC should go to a deeper > level for some sections than others (in particular, appendices > usually don't need the depth in the TOC that mainline sections do). > I don't expect it will be possible to predict and provide > configuration options for every scenario like this. We should just > always expect there there might be a need to edit the output. Not everything can be supported, but there's already an option for ToC depth — I can look at making that controlled on a per-section basis. -- Robin Berjon - http://berjon.com/ - @robinberjon
Received on Tuesday, 26 February 2013 15:46:36 UTC