>>>>> "Andre" == Andre Poenitz <[EMAIL PROTECTED]> writes:
Andre> On Thu, Sep 19, 2002 at 01:59:35PM +0200, Jean-Marc Lasgouttes Andre> wrote: >> For example, the use of layouts (a flat thing) and depth to nest >> environments is great. Andre> No, it is not.... It is very convenient to use 99% of the times. It just works. >> Of course, as it is now, it causes problems (like two consecutive >> theorems), Andre> ... because of that. Moreover, it allows only changes on a Andre> paragraph level: No more, no less. Not all things are one Andre> paragraph. What do you mean? >> but this is something that can be solved with enough thought. Andre> The only "solution" that can be implemented using the current Andre> approach is having four layouts Andre> "Single-paragraph theorem" "Multi-paragraph theorem, first Andre> par" "Multi-paragraph theorem, middle par" "Multi-paragraph Andre> theorem, last par" Andre> And that's certainly Not Nice(tm). Andre> Do you know any other? For example have a checkbox in the layout>paragraph menu that says: 'break environment after paragraph' (or whatever). This would go together with the linbreak, space and line options. Then break-paragraph would set it and break-paragraph-keep-layout would leve it unset. I'm not sure this would work, but something like that is possible. Andre> [This is btw the typical situation with the "conventional Andre> approach": You get almost everything working, with a quite bit Andre> of effort a few cases more hacked in somehow, but the remaining Andre> things are technically impossible] Whereas with the structured approach, you get almost everything annoying to do, with quite a bit of effort the complexity is hidden somehow, but the latex experts will be able to do their favourite tricks. >> The ``theoretical'' approach of "let everything be an inset that >> can be inserted anywhere" just sucks when you actually want to >> write a document. Andre> Why? Have you tried it? When one will have created enumerate insets or theorem insets and that it will not be possible to switch from theorem to lemma without creating a new inset and cut-and-pasting the old one, I will complain and you will say: "I don't care. Fix that as you want". For me the capability to mutate an object is important and I really fear that the inset-everywhere approach will kill that. Andre> We are more or less implementing it in small steps now anyway Andre> to solve "issues". ERT has gone inset, and this was Good. Andre> Minipage support has emerged - as insets. Tables are using Andre> "insets" nowadays etc. I'm not against insets in general. Of course minipage, footnote and tables should be insets. That's why insets exist. When I see your proposal for user-defined environments, for example, I am not sure anymore that we will not have to pay for that in terms of usability. Andre> -- I wonder why we are discussing these things right now Andre> anyway. I was not suggesting moving outer world font changes to Andre> insets recently and the math font stuff is in since 1.3cvs was Andre> opened. Because when we try to discuss them afterwards, all we get is ``I don't care''. JMarc