Message-Id: Date: Tue, 10 Nov 98 16:24:46 MET From: RJ vd Boon To: Eli Zaretskii cc: djgpp AT delorie DOT com Subject: Re: ANNOUNCE: TeX/Web2c v7.2b ported and uploaded Reply-To: djgpp AT delorie DOT com On Tue, 10 Nov 1998, Eli wrote: > On Tue, 10 Nov 1998, RJ vd Boon wrote: > > I've read it completely (twice), I only didn't notice any word like: > > "by default the TEXMF and TEXMF/fonts are read-only" It might be > > usefull to cut and paste the little explenation to TeX.README. > As far as I see, the only thing that the README doesn't say is that these > directories already come write-protected; every other aspect of this > feature is described in detail. Well, OK. (I don't know the readme by heart...) You also said in your previous reply that I should add my new tree to the TEXMF variable (which I just deleted from djgpp.env, according to the readme) or add it to TEXMFLOCAL. That's what I wanted in the README, but as I now think, it also is/should be described in the docs, so the readme isn't exactly the place to put it. (`it' and `that' refer to the variables) > Is the comment on the default setup what you lack? If so, I didn't Yes, that's what i lacked. > describe that on purpose: some (most?) unzippers ignore the read-only bit > and unzip the directories without it (which unzip program did you use, > btw?). Info-Zip's unzip, version 5.32 (and about to upgrade to 5.33j(beta)) So if I understand right, it doesn't help setting the RO-bit for some unzippers and for others it does. This creates (IMHO) an inconsistent situation between different users, and even for a user[1] who sometimes unzips in a DOS-box (with infozip), sometimes with wincmd, and sometimes from nortoncommander, which give different situations, which I think was never your intention. [1] like I ;-) > So I didn't want to lie about the default being magically > arranged; instead I told that *if* you want that setup, you need to > actually set the attribute manually. Which I understood as: "The normal behaviour is a tree which is not write-protected". (i.e. I read *if* as: if and only if) Also a quote from TeX.README: The generated fonts will by default be installed inside the ^^^^^^^^^^ ^^^^^^ share/texmf tree, where they will be available for future use by the package programs. However, you might want the generated fonts ^^^^^^^^^^ to go to a different hierarchy, for example if you'd like to purge those files from time to time (to preserve disk space). If the share/texmf tree is on write-protected media (e.g. CD-ROM), mktex* programs will do that automatically; if not, you will need to set the read-only bit of share/texmf and share/texmf/fonts, like this: This led me to believe that the read-only attribute by default is NOT set. A note about the default _in_ the .zip and the behaviour of some unzippers would be something to consider. Or changing the default;-) > I figured that setting it twice would be a lesser evil than telling > it is already set when it might not be so. I also thought that those > who add additional fonts usually know better what they are doing > (like you did) and will be able to figure it out on their own. That's completely true. > I'm open to suggestions about how to fix all this mess in a better way. Hmm, this shows that some messy things are just messy. bye now, Robert PS I also noticed that you have ported dviljk-2.6 and not 2.8. I also noticed that 2.8 is mainly 2.6 + a little bugfix in tfm.c + DJGPP support-files. Wouldn't it be wise to use 2.8 with DJGPP? Or should I just compile and try it myself (and upload if stable)? -- rjvdboon AT cs DOT vu DOT nl | "En dat is niet waar!" sprak ex-Staatsecretaris- www.cs.vu.nl/~rjvdboon | van-Onderwijs Netelenbos fel.