@JosephWright To be honest, I don't really know enough about hyphenation patterns to answer any question about them. However, what do you want to know exactly? Is the question whether ŵ or ŷ occur in words which should be hyphenated? I guess they do. ( assume that owls need hyphenation, for example, on occasion.
@egreg I took the question to be what they should be. I realise T1 won't support them, but isn't this a question about the hyphenation patterns Welsh should have, regardless of whether a particular encoding implements them or not?
@cfr well in particular the file /usr/local/texlive/2015/texmf-dist/tex/generic/hyph-utf8/patterns/tex/hyph-cy.tex is UTF-8 hyphenation patterns for luatex and xetex but doesn't have them (or any other accented letters but the ones listed a couple of comments up)
@DavidCarlisle Ah. I didn't look at the comments. Just grepped.
@DavidCarlisle At least the list of letters doesn't obviously suggest they've copied them from Breton. So they're ahead of Babel, Polyglossia, datetime and dateitme2. (Actually, Polyglossia is now corrected. Maybe even Babel. But they did produce half-Breton for years.)
@DavidCarlisle Maybe somebody could ask for the hyphenation patterns that paper says the authors already have for Welsh? The paper explicitly mentions the hyphenation needs of languages such as Welsh with its ŵ, so presumably the patterns that they had/have includes patterns for words with the additional characters excluded from the T1-based patterns currently distributed. If the patterns were already prepared then, one of the authors should know what became of them, shouldn't they?
@cfr You can retract a close vote by clicking on the "close" button again. It then warns you that you will not be able to vote again. I don't know if there's a limited time window for you to be able to do it.
@AlanMunn Hmmm. When I've tried it, it has said I can't vote again because I've already voted. But it hasn't let me undo it. Even when I've realised almost immediately I've made a mistake and voted to close something I didn't mean to. Obviously your powers exceed mind ;).
@cfr :) I think then there may be a short time span in which the powers are active, and that's why you weren't able to do it. Shorter than you remember?
@egreg Yes, I realise that. I was observing that there are deficiencies in some of the current pattern files related to the fact that they are really 8-bit simply transcoded to UTF-8, rather than being real Unicode pattern files
@DavidCarlisle (u)pTeX automatically treats anything outside the 7-bit range as Japanese. upTeX does seem to have a way of turning that off for the 8-bit range (I've not tested fully), but in pTeX that's not the case. However, you can use ^^ notation so that is what the pattern files have.
@JosephWright ah sounds vaguely familiar thanks for filling me in, although even then the patterns loader could presumably do the 8bit to ^^ notation conversion in macros so as to use the same input files.
@cfr (@JosephWright) I could mail Yannis, but they may be in the old omega files that should be around somewhere)
@DavidCarlisle I assume that this is ruled out: my guess is that the macro level never sees chars from the upper half of the 8-bit range but does seem anything put in using ^^ notation
@JosephWright but certainly the welsh ones could have been back converted to unicode rather than just dropping the characters in control symbol slots:(
@DavidCarlisle Re. your programming answer, I wonder you should include the fact that the 2e kernel is pretty stable (certainly if you don't want to work with engines that are themselves changing)
Google translate from the serbian wikibook gives me: "You should use a4wide package page with A4 size document with smaller margins. This is old fashioned and crazy. Use the "geometry" package instead,"
@DavidCarlisle A user said, that he loaded the package for tables (tabularx), i asked where he read it. L for complete novices. I greped the source and searched the pdf, nothing. I said, no, that document does not contain tabularx. Answer: Of course it does, do you think i am making that up?
@JosephWright yes although possibly safer than trying to auto-edit could do a "final check" test in the ctan build that checked the nominal date appears in readme and and changes and ltvers.dtx and anywhere else it is supposed to be
@JosephWright No version mentioned, i figured there wouldn't be a much difference because changes haven't been that big. But i don't know how big the diff between hard- and softcover is. Cn't be that big, since the toc is done by hand.
@DavidCarlisle The same page says that you can use \; in text mode without mentioning amsmath. And did you know, that xpsace can only be used in math mode?
@Johannes_B The interword space for cmr10 is indeed 1/3 of a quad, stretchable to 1/2 and shrinkable to 2/9 of a quad. But it depends on fonts, of course.
@egreg I noticed that today when a perfectly fine poster document failed with a missing $ error. The example was very very minimal when i noticed that the \name macro that failed was \def\name{Software name\;}.
@Johannes_B A correct statement would be that a\ b gives a normal interword space unaffected by space factor computations. So a.\ b has less space than a. b (assuming \nonfrenchspacing).
property for ideals in Krull domains to generalized
Krull domains, in the same spirit of a work on
generalized Dedekind domains by Gabelli and
Popescu [8]. \par
\ind A generalized Krull domain (GK-domain for
short) is a PVMD such that $P\not=(P^2)_t$, for each
$t$-prime ideal $P$, and each nonzero principal ideal
has only finitely many minimal ($t$)-primes
(cf. [5, Theorem 3.9]). GK-domains of $t$-dimension
one coincide with the class of Krull domains. For
more details see [5].\par
@JosephWright tex.stackexchange.com/questions/290401/… The user posted almost the same question a few hours ago, long comment thread, at least one answer by Bernard.
@cfr -- nominate yourself anyhow, please. there are only four names there now, and i think at least one has said s/he is already a member, so declines. and the deadline is tomorrow. there's no shame in being re-upped.
@Johannes_B -- dek has been very careful to avoid changing page breaks, and the softcover and hardcover are the same in that respect. doesn't depend on edition or "printing", and that essentially goes back to the first edition. changes are almost universally within a page, and usually new/rephrased paragraphs are able to be reset from the errata lists and even pasted in without obscuring anything.
@Johannes_B -- i'm sure i've said it before ... the wikibook is full of misinformation, and badly needs editing. maybe i'll attack it when my current crises are under control.
@barbarabeeton If you have some crippled rotting building structure, would you invest effort to minutely separate the good pieces from the bad and add new helpful pieces ... or would you just start from scratch? With a new blue print.
@Johannes_B -- yes, i agree. but the wikibook is much used, and i'm sure there are plenty of links into it. i don't know how to disable the wikibook (i haven't even been successful getting into it to make some easy edits, and i don't know how to get "approved" for that function either) without killing them. i don't like dead links, and to have this all at once disappear would not be good p.r. for tex.
@barbarabeeton Agreed, deleting it not sensible. But links will go dead if the wikibook gets some attention, since many parts are strangely put together.
@Johannes_B -- agreed to that too, but it's possible to insert a forward link in a page that's superseded. with cooperation from whoever oversees the site, that can even be made automatic; that's done at ams, where pages are moved around all the time.
Clemens and i have been doing some modifications to the german wiki, but i somehow think that documentation needs one defined downstream which all other languages (or other institutions) take info from. Of course, as mentioned in the mail thread with james allen, this is complicated.
and in no means something that should be open for modification by everyone.
@Johannes_B -- while i agree that unrestricted modification is a bad thing, i do somehow think i should be permitted to be a potential checker/modifier of this wikibook. if you know how to get that to happen, please send mail with instructions.
@barbarabeeton When i encounter something strange on wikbooks, i don't know where to start. It often would be deleting one third of the page, because is is faulty, moving one third to another section to find out that it already is there (no duping, so rm) and completely rewriting the remains. That is called a drastic edit and should discussed on the discussion pages. Of course, you can just complain about the quality and somebody with some kind of human sense says: Go ahead, apply fixes.
@Johannes_B -- okay, i will start there. regarding updating, i usually start by printing out what i intend to work on, read and mark up off-line, then go back to the electronic version and update. that way i know what i'm doing (well, as much as i ever do), so can plan ahead better. the last time i tried to make even simple updates to a wiki entry, i got so frustrated after a couple of hours of failure that i just gave up.
@JosephWright oh would definitely mean a new latex format before texlive 2016 if he's planning on pushing that out. I suppose we should make an extended allocator available just as a temporary package so people can test?
@JosephWright of course if he's active in xetex internals again now might be a good time to ask if hyphenation could be more like luatex and loaded in a document...
@JosephWright oh ok, that's what I thought. I really should look at context again one of these days, I looked quite hard at it when it first came out but I doubt any of that remains
@barbarabeeton I also do the changes off-line. The width of my editor is limited to 65 chars (sometimes 70, sometimes 80) and hard line wrapping. This makes the wiki-diffs completely useless. I don't know if you can force that built-in editor to wrap after a defined width and if so, if a consense would be found regarding a common width.
@JosephWright when i was pushing for fixed handling of utf16 surrogates last year I failed totally to build the sources that I fetched for the standalone xetex from sourceforge, but the texlive svn built almost straight away, just needed to add a couple of libraries. So I'm not completely confident but if I put it in the texlive tree, it might work....
Does anyone know whether BaKoMa TeX distributes its own distribution or can be used with TeXLive or MikTeX? If the former, do we know how up-to-date the distribution is? The site documentation doesn't seem to say anything about the distribution.
@JosephWright yes hard to answer without knowing how the catalogue works, can they have a latex-graphics catalogue entry without changing the ctan or tds location?
@AlanMunn it certainly used to be a separate tex implementation, not looked for years
@PauloCereda As part of the intro course I supervise the students have to read a text from which a plot of their vowels will be produced. The name of the text: "A bad day for ducks" books.google.com/…
@MarekKaski I think you might be better asking a question on the main site. But you should maybe explain what you are trying to do exactly, because it might be that there are other ways of doing it. And people will probably want a sample document to play with.
@JosephWright on branching I've only made the branches when bumping the version number so the 2015/10/01 branch is already there and I don't think we need a 2016/02/01 branch until we add stuff for the next dated release, otherwise we can't do patch level updates on trunk)
@JosephWright Would you think to a function, say \prg_leap:n, that takes as argument an integer and returns 1 if the integer corresponds to a leap year (after 1982) and 0 otherwise? See tex.stackexchange.com/questions/290418/…
