X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["4809" "Tue" "2" "November" "93" "15:03:02" "-0700" "Michal Jaegermann" "michal@GORTEL.PHYS.UALBERTA.CA" nil "83" "I do not like 'nf' in style names" "^Date:" nil nil "11"]) Return-Path: Received: from sc.ZIB-Berlin.DE (mailserv) by dagobert.ZIB-Berlin.DE (4.1/SMI-4.0/24.6.93) id AA15429; Tue, 2 Nov 93 23:07:23 +0100 Received: from vm.urz.Uni-Heidelberg.de (vm.hd-net.uni-heidelberg.de) by sc.ZIB-Berlin.DE (4.1/SMI-4.0-sc/03.06.93) id AA22003; Tue, 2 Nov 93 23:07:19 +0100 Message-Id: <9311022207.AA22003@sc.ZIB-Berlin.DE> Received: from DHDURZ1 by vm.urz.Uni-Heidelberg.de (IBM VM SMTP V2R2) with BSMTP id 5715; Tue, 02 Nov 93 23:06:07 CET Received: from DHDURZ1 by DHDURZ1 (Mailer R2.08 R208004) with BSMTP id 3158; Tue, 02 Nov 93 23:06:05 CET Received: from DHDURZ1 by DHDURZ1 (Mailer R2.08 R208004) with BSMTP id 3156; Tue, 02 Nov 93 23:06:02 CET Reply-To: Mailing list for the LaTeX3 project Date: Tue, 2 Nov 93 15:03:02 -0700 From: Michal Jaegermann Sender: Mailing list for the LaTeX3 project To: Multiple Recipients of Subject: I do not like 'nf' in style names Status: R X-Status: X-Keywords: X-UID: 1064 The release of NFSS2 introduced "prefixed" names for various style files associated with fonts, e.g. nftimes.sty, nfpandor.sty, etc. I personally feel that this is a gross mistake and recently had some exchange of letters with Rainer Schoepf on this topic. I will try below shortly present both sides of the argument. If I misunderstood something in his stand, then undoubltely I will be corrected pretty fast. :-) We are both interested what are opinions of others on that matter. My most serious objection is about backward compatibility. I know that NFSS2 includes also files like times.sty, which loads nftimes.sty, but this makes a lot of noise, including threats that this particular times.sty is going to disappear in not so distant future. I do not think that forcing users to dabble with their existing documents is a good idea to introduce a new version of a software. Apparently trivial matter of slightly modifying your document to conform to "new" style names is not so trivial if you deal with an existing archive which holds a substantial number of such documents and which could be processed automatically. It also makes not so (La)TeX fluent users and system administrators much more reluctant to upgrade their systems. In "sysadmin" class I include here both owners of sigle-user machines and maintainer of multi user facilities. The later are less likely to know anything about TeX and are, understandably, much more cautious in any system changes, althouth the former also include many which follow "if ain't broke, don't fix it" principle and tend to regard any new warning as a sign that something went broke. As a result we have around surprisingly many installations running long obsolete versions of TeX and things like nftimes.sty vs. times.sty only help to perpetuate this indefinitely. I will also have problems like this: for this my particular corespondent, which uses quite a bit of TeX - but only in a "black box" mode and is afraid to touch anything - and is located at the other end of a possibly quite long wire, should I use now 'nf...' variant or not? I know surprisingly many folks like that. Rainer thinks that there is too many variants of 'times.sty' floating around and if a wrong one will be used things will break and this is even worse. New names will help to prevent this from happening. My position is that of course there is many variants of 'times.sty', because they are used with different system setups. In my opinion this is a Good Thing(!!), since it allow to insulate your "naive" user from what is going under-the-hood and this is exactly what all this game is about. There are also many variants of 'latex.tex' and not exactly compatible in every single aspect and in each direction. If you will start to mix-and-match files from different releases things WILL break and sometimes they indeed do, as we all know perfectly well from questions in comp.text.tex. I do not think that there is a way to prevent that, although Rainer can be tired answering questions related to this issue. On the other hand, if you are starting to write more complicated style files, then you are expected to know at least basic rules of the game and you should be able to differentiate between various variation on the same theme. My next contention is that we all hope to see days when "New FSS" will become "Very Old and Tired FSS". If this will happen, then names like 'nftimes.sty' will sound somewhat stupid. Expecting that with times these names will aquire a tinge of a "Nieuw Amsterdam" is maybe a bit too optimistic. :-) Here Rainer tends to agree and he thinks that names like 'ltimes.sty', with 'l' for LaTeX can be used (if I did not mis-interpreted something). Tad better, but this does not address points raised above. Besides, '.sty' suffix is generally associated with LaTeX anyway - few assorted files from AMS-TeX notwithstanding. The last, but not the least point. Quite a few OS-es is limited to eight characters in a base part of file names (MS-DOS is one of these, but not the only one). Taking effectively away 25% of available letters from that meagre allocation does not seem to me to be a very good idea. Rainer thinks that this is not an issue, since names like 'nftimes' and 'nfncs' fit and 'nfpandor' is loosing only one letter. I contend that when more various fonts will be used these six character names will become very constraining. Even right now, can somebody tell me how I am supposed to fit into this scheme, for example, 'mathtime.sty'? 'nfmthtme.sty'?? Shudder!! You need an old Fortran mind to do that. :-) BTW - does anybody know of the corresponding file in question which works with NFSS2? So what do you think? Michal Jaegermann michal@gortel.phys.ualberta.ca ntomczak@vm.ucs.ualberta.ca