X-VM-v5-Data: ([nil nil nil nil nil nil nil t nil] [nil nil nil nil nil nil nil nil nil nil nil nil nil nil nil nil]) Date: Wed, 18 Apr 90 16:00:49 CET Reply-To: LaTeX-L Mailing list From: PZF5HZ@RUIPC1E.bitnet To: Rainer Schoepf Status: R X-Status: X-Keywords: X-UID: 80 I like to respond to Michaels description about the enunciation environment he is planning for amsbook.sty and amsart.sty. First a few general remarks: I think it is very important that we find a syntax for things like theorem envs or enunciation or whatever you call them which is sufficent general to serve in a wide range of documents. This is one of the reasons for this list. It would be really bad if Michael is going to propose some syntax for use with the amsart.sty and we, later on, propose a different one. One of my main goals is to specify document classes, I called them meta styles in previous messages, and any incarnation of such a class should define exactly the same commands with the same syntax. This is the only way to achieve portability between document styles in the same class. I therefore hope that Michael is not going to implement things of this sort as long as there is no agreement. If the enunciation env. is implemented as a style option, i.e. usable with, let's say the siam document style, too, it would be okay, but his intention is, if I judge this correctly, to make it part of the document style, and therefore implicit part of the document class. This is not a judgment of his ideas, I only want to avoid producing standards through the back-door (as it often happens) by simply throwing it to the user and later on saying: sorry can't be changed is used too often. ******** Since I think this is a very important topic, I would urge you, if you could spare a few minutes, to write your opinion about it. ******** Now for some detailed answers to his last message. ... They are based on the term `enunciation' rather than `theorem' (\newenunciation instead of \newtheorem, and so on) because the idea of `theorem' doesn't seem general enough to encompass the whole class of structures including examples, remarks, notes, claims, etc. for which the macros are designed. In German we would say names are `Schall und Rauch'. I accept that the name \newtheorem is probably missleading but I never saw any LaTeX user who had realy any problems with it. I don't intent to change the name for LaTeX version 2.10 because this would produce unnecessary compatibilty problems. But maybe you can convince me otherwise? For LaTeX 3.0 a rename is more apropiate. I define an enunciation as a piece of text at least one paragraph long that has a label (or `header') and ... Michaels definition correponds exactly to my understanding of theorem like environments and my implementation of an extended theorem env in TUB is very similar to his opinion how things should be handled. To make things clear I try to sumerize the differences between his and my syntax. 1) enunciation styles: We both support the idea of having different classes of enunciations with different typographical treatement. 1a) FMi uses three seperate macros: \theoremstyle \theoremheaderfont and \theorembodyfont during definition to denote the treatement. MDo uses one optional argument. Advantage FMi: single users can define their wanted layout in a simple but easy changable way. The splitting of style and fonts keeps the number of predefined style low. Advantage MDo: his idea supports the document class idea better because the decision that a `remark' has to be formatted using ....... is part of the style design and not a problem of the user. Personal remark: I don't like to fix everything in a style file because this means that one has to design a new style file for every change one like to have. The golden middle way seems to be better. It is no problem to say that files submitted four specific journals have to use style suchandsuch for remarks but this would allow the user to do differently at home (still using amsart) 2) Special variants: FMi provides nothing, MDo provides star and colon form. Advantage MDo: the possibility for changing or removing the number is important. Personal remark: I don't think the : form is necessary. I had enough experience the mathematician using LaTeX there was no confusion about the meaning of a star form. In LaTeX star means only this is a special variant, think of \\* figure* tabular* etc. I would also propose a different syntax for \newtheorem, namely using the starform to denote that this enunciation has also a star form, i.e. a variant where the number can be *left out* or changed. 3) Extra enunciation env: Not handled by FMi. Again a very good and important idea. Summary: The main difference between both implementation is that I used descriptive names like `marginbreak' while Michael would like to say `remark class'. His class is the combination of one style of mine with a font. So I would propose one of the following syntax variants: \theoremstyle{