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: Mon, 5 Mar 90 16:36:22 CET Reply-To: LaTeX-L Mailing list Sender: LaTeX-L Mailing list From: PZF5HZ@RUIPC1E.bitnet To: Rainer Schoepf Status: R X-Status: X-Keywords: X-UID: 51 In a letter to Rainer, Nico raised a few important questions which I try to answer here, for all of you who didn't meet me at Karlsruhe or Stanford where I essentially said the same. The questions were: 1) When do we decide about the new style interface or new features of LaTeX 2.10/3.0? 2) Who is going to decide these questions? 3) Who is going to implement the new code and when? My experience with the developement of a huge piece of software like the redesign of LaTeX is similar to the statements made by Don Knuth in his keynote at the Stanford meeting: A consistent piece of software has to be designed by one person, that is the final decisions have to be done by one person or at most by a very small group. The designer (author) has to be someone who is going to use this system extensively. He should also guide or better do the implementation because this will give many new insights about how original specifications could be improved. It is also necessary to be prepared to throw away the first (or later) versions of the system and start again from scratch using only the valuable information (about good and bad concepts) while using the old system. Guided by these believes and my experience in various parts of document processing Rainer and I outlined a concept for a re-implementation of LaTeX. This concept was presented and discussed in Stanford both in public as well as in private between Leslie and me. As a result of this, our paper appearing in TUGboat 10#4 got an update section stating that Leslie is in agreement with the principles contained in our paper. In Stanford we discussed how this redesign should take place and decided on a two step plan, namely to first redesign the style file interface and afterwards tackle the other topics raised in our paper. One of the main results of the Stanford meeting was that our proposal wasn't any longer a proposal beyond the official LaTeX but that I was in charge of the redesign of LaTeX. Otherwise such a project would have been bound to fail and I think I would not have invested any time in a product to be in competition with the official LaTeX. (** footnote: don't get me wrong, I don't want to say that the project needs me to succeed, but there has to be a leader, NOT a committee making the final decisions, and it has to be LaTeX, not a NEW product. **) As a next step both Rainer and me gave a talk about this project in Karlsruhe, this time in greater detail, discussing this topic afterwards with various people throughout the meeting. Then we went into closure coming up again two days before christmas with a completely new interface for font selection. To see that we really believe in throwing away prototypes compare it with our first attempt which ran for four months successfully on several test installations. At the same time we re-wrote nearly everything we got >from the AMS what they called amstex.sty. We had long debates about the syntax for the math features and the final product which you will see in short time has, in my opinion, a good chance to survive nearly unchanged as peripheral part of LaTeX 3.0. It will probably be released this month. Coming back to Nicos questions, the final decisions will be made by Leslie and me, and nobody else. That is, I hope that we end up in agreement about the currently open questions but there might be a point where I decide what happens and then it is `take it or leave it'. The implementation will be supervised by me and probably a great part of it will be done by Rainer and me. But I would be very glad if some of you would also help in this phase. LaTeX is sufficiently modular so that it is possible to divide the implementation when the interfaces are properly defined. But the goal at the moment is the design of the internal interfaces to document styles. This does not mean that there should not be some short test installations like the one I did last week to test Nicos and my ideas about \@startsection. Such prototypes are very helpful to find the flaws in the syntax but they will probably not resemble the logic of the final implementation. As for the timetable: I don't share Don's enthusiasm and would think that it would be fine if the style interface is defined at the time of Texas meeting but I don't believe that a prototype implementation will be finished by then. As a consequence of our decision to concentrate first on the style interface suggestions like the one by Malcolm are filed away but are otherways ignored by me at the moment. To make this point clear: I acknowledge this as a very important topic and it is already on the list. When 2.10 is finished (as a test version) I will go back to all these problems and solve them for version 3.0. By the way: indexing and bibliography problems are already on the list in our paper. My timetable for this is around the Cork meeting but we will see if even this is too enthusiastic. (** Please don't stop sending those requests and/or suggestions. What I want is not to be carried away by too many topics at a time. **) To summarize, I don't want to discuss math at the moment and I don't want to discuss font selection. But it might be a good idea to send around a syntax description for these two topics and take up the discussion next month or so when some other things are settled and you have had time to study our solution. So, this was a long announcement, in some sense it could have have been the opening statement for this list but I'm still not through because of an article in Nouvelles de GUT enberg 1/3/90. I've thought long about it, whether to discuss it in private with Bernard Gaulle or to discuss it in public (through the french list). i finally decided to use this forum for telling him (he is also on this list) how I feel about his article because I want you to know my reasons for being angry about it, in the hope that such things won't happen again. In his message he announced the existence of this list installed at the listserv at HD, asking for french LaTeX gurus to join because he felt that the french positions are not properly repesented in an `adhoc list' consisting of some `allemands, hollandais, anglais et americains'. It seems that he missed our nordic participants but I suppose this knowledge wouldn't help much. But actually his understanding of the situation is faulty: This list will not discuss things where a `french voice' or a `xxxx voice' is important. Actually, I don't see any point in our proposal where language specific topics are raised except the hardwired captions where we already took the position to remove them. Second, this list is not an adhoc list where all people with interest in the future of LaTeX are gathered. Instead it is a list of people which were personally selected and asked by me to join this list because I knew of their knowledge about the topic and because of my personal experience that I can work with them. There were a few exceptions, Mr. Gaulle was added to the list because I wanted to keep the french user group informed and one other was added because of recommendation by someone else. This group is not a group of LaTeX experts (even if some of you are) and is was never ment to be a `round table' of all parties involved. As I said above (and that's the reason why I made this part so long) I don't believe in committees, to make this a working group I had to make decisions and many people which are good in this field where not asked for several reasons. You may blame me for these decisions when the results don't please the french community but they are mine and you have to accept them face value. These decisions weren't easy as one can imagine and one of the reasons why this is A CONFIDENTIAL LIST is that I didn't want to turn down many offers of help. In this regard Mr. Gaulle did a great job for the LaTeX project by announcing the existence and location of this list in public, and also giving a wrong impression about its purpose. May I ask why he didn't tell the members of this list about his feelings and impressions first, knowing that this is a confidential list? May I ask why he never approached me during the meeting in Stanford where more than one meeting was held with Leslie Lamport, where my talk was given etc.? May I ask where he or another representative of the french TeX user group was during the European TeX meeting at Karlsruhe this year? At all these places and even afterwards it would have be possible to talk to me about this but for doing this behind my back, I can only say again: thank you. There is already enough damage done so I don't think I like to respond in public and I hope that Mr. Gaulle will stop talking about this list in public, too. As for the LaTeX guru he is searching, unlike him I know many of them personally (and some of them do speak french) and had no intention to add them to the list. I don't like to do this at the current state of affair with somebody I don't know at all. Frank Mittelbach