I've got a short question: I would like to start a question on TeX.SE like "How are / should BibTeX fields be used?". The answers can explain any BibTeX field (e.g. "comment", "title", "review", "owner", "file", ...).
The question would be very broad, hence I hesitate to ask it like this. My real question is "Is 'comment' used in the created PDF output or is it just for me to organize my literature?". But I really would like to make the question broader as I came across this type of question a couple of times.
Do you think I should ask such a broad question? (I wouldn't care if we made the question / answers community wiki)
@PauloCereda Ooo Paulo, I've just seen you have a new profile image. A duck-bee. Reminds me of youtube.com/watch?v=txj6ROnIUIo
@moose only the ad hoc standard of what the standard bib styles do but in this case it's very loose as you can use any name at all as a field name and bibtex just silently ignores it so any field not used by the style you are using is a "comment" and conversely any such field might be used by another style
so moose-my-private-bibtex-data-dont-write-a-bst-that-uses-this="something" might work for example.
@JosephWright for the l3 case changing stuff do you have an uppercase variant that drops accents? (thinking about xgreek, mostly but also some French styles I think) I suppose you could extract the relevant data from the "decomp" field in unicodedata to get the base letter, unless there is specific uppercase data elsewhere in UCD?
@DavidCarlisle I've not done this yet, but I am aware of the Greek situation!
@DavidCarlisle I checked up on French: as I understand it, the 'dropping accents' thing there was really just to do with using typewriters and was never part of proper typographic use
@DavidCarlisle I do have a list of mapping for Greek, taken from xgreek but modified as some of those are wrong (for example, 'hacks' to align non-letter chars with cased letters)
@DavidCarlisle The input I got was that Unicode mappings are 'wrong' but as it's stable we are stuck with them. I've not checked the UCLDR
@DavidCarlisle True, but as I say the feedback I had was that this would be 'out and out wrong', not just 'a stylistic choice'
@DavidCarlisle I've probably got to revise things anyway: I'd like to use the new 'raw' Unicode data directly and not pre-compose the intermediate data file. I think that's doable as with a Unicode engine \lccode and \uccode should be regarded as 'fixed' for all code points.
@DavidCarlisle I'm thinking of using 'list of exceptions then \lccode/\uccode': we need about 500 special cases, which is not too bad as we need 300+ data entries with the 'packed' approach anyway.
@DavidCarlisle No, for French. Greek seems complex: the rules for modern Greek require accent removal, but for ancient Greek the UCD might well be correct!
@DavidCarlisle Yes, I'd picked all of that up. Makes life slightly tricky as really the standard mappings should I think be 'in common use today'.
@JosephWright Not tried running it (not really stable enough to try a cygwin compilation) but things seem to be happening
@JosephWright well that is loading a shared library so doesn't need any direct support from luatex core but may need support from texlive infrastructure to see if its possible to have a suitable shared library version of harfbuzz on all platforms.
@JosephWright yes the followon from the above would be that use would be the same but distribution would probably be easier if it was statically linked to the luatex executable, but that would mean forking luatex probably (or at least a variant compliation of same sources)
@JosephWright which is why I opened an issue the other day to guide them towards using the version of harfbuzz that's already in texlive for ease of compilation (and compatibility with xetex)
@JosephWright -- i'm no expert in either greek nor what you're doing to support unicode "natively", but i do know that the handling of diacritics for classical greek is much different from that for modern greek. from the couple of semesters of modern greek i've struggled through, i know for sure that modern greek preferably uses only a single fairly "neutral" accent mark, but for classical greek, if even one accent/diacritic is removed, the meaning could be changed drastically.
@barbarabeeton As I explained at TUG2015, we can have different case changing outcomes. What's awkward is that I'm told that the standard Unicode mappings are the 'wrong' ones: ancient Greek, probably done by people who don't actually use the language day-to-day
@barbarabeeton The issue arises with both monotonic and polytonic modern Greek (and indeed that split isn't our issue: it will be present in the input anyway)
@JosephWright -- i think i can find out the background of the unicode greek mappings. would you like me to do that? and i think i can also find someone who can be considered an expert in classical greek who'd be willing to check it. (for that, i would ask diana wright, who was pierre mackay's companion in his last years; she's a classical greek scholar herself.)
@barbarabeeton The UCD mappings are what they are, it's just a question of saying 'the standard mappings may be unsuitable for modern Greek, you should use ...'
@barbarabeeton I'm pretty sure the request I've had is right, I just haven't implemented it :-) (The same has come up for things like web browsers, and they seem to have been convinced that a dedicated mapping was required)
@DavidCarlisle I'm not sure the issue is the LPPL per se, my guess is they want something with a very permissive license and the LPPL is simply an example of one that's not permissive enough
@DavidCarlisle I am in fact quite curious if someone could get it working under Windows, I mean at least compiling Luaharfbuzz library, I managed to get Luarocks working on Windows, but compilation of C modules doesn't seem to work
@DavidCarlisle -- yes. ugh. my understanding is that the lppl was fought over long and hard with debian, and only grudgingly did they accept the logic that the renaming requirement was necessary for stability. there were probably other aspects too, but i eschew legal jargon as fervently as i can. if you don't know the details, then probably frank is the only person who does, and i'm sure he won't be happy to go into this again.
@JosephWright -- well, requested then. i think tex is fortunate in having most developers in agreement with the reason for the request, and honoring it.
@JosephWright -- well, mojca has just posted a cogent and enlightening response to the google request. if google is willing to foot the legal expenses, then perhaps it's possible to resolve this mess. i agree with the authors who want recognition for their work, or at least a "guarantee" that no one else will be able to claim it, hence the insistence that their name remain with the work. poor mojca.
@barbarabeeton These things are very personal, of course: I've had more-or-less the opposite discussion with Frank (I don't want my name on anything done for the team, TUG, ...)
@barbarabeeton CC0 would deal with the re-licensing business (as it's intended to be almost equivalent to public domain) but as you say people can then find their names removed from stuff.
@JosephWright -- yes, but you assign credit to the team, and i think you'd be unhappy if some plagiarist in lower slobbovia undeservedly claimed ownership.
@barbarabeeton Yes but that would be both possible and actionable irrespective of the license (even CC0). Taking a public domain work and putting your name on it is possible but doesn't mean it's got any legal standing.
@michal.h21 I use the cygwin version of texlive so more like linux than windows, except when it isn't like linux at all:-) the complete texlive build works OK as all the cygwin dependencies have been sorted out, so it just needs make but I'm hesitant to try code that is still in development:-)
(/usr/local/texlive/2015/texmf-dist/tex/generic/pgf/systemlayer/pgfsys-common-p
df.def)
! unexpected use of \pdfvariable.
<to be read again>
l
l.418 \pgfutil@setuppdfresources
I'm using TikZ' graph layout engine, and I'm very confused as to why
\graph {1, 2, 3, 2--3};
draws the desired graph, while
\graph {\directlua{dofile("foo.lua")}};
does not when foo.lua just contains
tex.print("1, 2, 3, 2--3")
I'm obviously very confused with regards to \directlua…
Edit...
@JosephWright first thing was pdflastobj not defined as I had missed out a block using \pdffeedback from luatex85.sty (just pushed that) but it still dies in the pgf code
@JosephWright although I had wondered before about that area. In the change to allow 256 math \fam, I restricted the symbol font declarations to 16 and just allowing math alphabets to use the higher ones but it occurred to me the other day that if \DeclareMathSymbol used the U form when defined then it wouldn't have bit packing restrictions from \mathchardef
So... I kind of have completed my version of the Advent Calendar challenge, except for one thing I feel too newbie about, but... Should I paste 350 lines of code in an answer or use an external link? lol
@JosephWright yes I had a browse around as well. Hopefully I'll get a reply to my question about what he meant.... I'm wondering if I should try to set up a build from the luatex svn rather than texlive as it feels like it lags behind at times like this...
@JosephWright ah you mean I'd have to build the right thing. . I was a bit put off as I failed building the standalone xetex but once I fetched the whole texlive tree basically just typing make worked and all the cygwin dependencies got resolved somewhere.
I would like to plot a diagram in latex. I found that xfig can do this for me. I am using windows 7, I installed cygwin. In addition, I have installed xfig during the cygwin installation. My question is that, I can't open xfig inside cygwin. is there any specific software that I should install?
I'm using TikZ' graph layout engine, and I'm very confused as to why
\graph {1, 2, 3, 2--3};
draws the desired graph, while
\graph {\directlua{dofile("foo.lua")}};
does not when foo.lua just contains
tex.print("1, 2, 3, 2--3")
I'm obviously very confused with regards to \directlua…
Edit...
You did ask for a Christmas tree? (I updated the tree to straighten it out)
\font
\t=cmtt10 at 1.5pt
\t
\baselineskip
1...
