@JosephWright -- eek!

@wilx (and @MarcelKrüger) I was just told that the issue you are seeing with that document is indeed a bug, but in the luatexja patch to microtype (lltjp-microtype.sty). Robert has already reported it here: de.osdn.net/projects/luatex-ja/ticket/39710 (in case you want to keep track of the progress).

@PhelypeOleinik -- I know very little about how microtype actually works, so this is only a conjecture. Since one of the goals of microtype is to make text appear maximally uniform, i.e., keep the spaces a consistent width, the mechanism it uses can be to subtly adjust the shape of individual characters. That would require "knowledge" about the algorithms for drawing the characters in order to be able to modify them. My guess is that when each new font is introduced, it is examined in detail. (cont'd)

(cont'd) Such examination could account for the additional time when each new font is introduced. It may also be the case that a second introduction of the same font is not checked to see if that font has already been examined; if I were debugging (which I'm really not competent to do in this case), I'd look for that possibility.

@barbarabeeton Thanks for your comment! I can say I know as much of microtype as you do, but my initial guess was more or less what you said. But as Robert said in the bug report I linked above, microtype should examine the font only once, but the bug introduced by luatexja makes it do so every time a “font is loaded” (which I suppose means a font change, since there's no font loading mid-document), thus the slowdown.

@barbarabeeton In fact, if I comment out the luatexja as suggested in the bug report, the overhead caused by microtype is noticeable only in the preamble, where the fonts are actually loaded, and the rest of the document takes roughly the same time to compile.

@PhelypeOleinik -- Aha! Well, that seemed a rather obvious possibility worth checking. Thanks for confirmation.

@PhelypeOleinik Nice. Thank you.

@DavidCarlisle smbc-comics.com/comic/the-talk-3

@PauloCereda ↑↑↑

hi, I want to change font of my name CV, but i dont know how :((

@HaraldHanche-Olsen NICE!

@PauloCereda ... and it's getting even better ... :)

You should come here really :)

@EnthusiasticEngineer no one can help you given that amount of information

@DavidCarlisle I have written my CV in PDF latex, and not used the usual packages for CVs. I have designed what I love. The font of my name in the header is the same as that in the CV's text. I want it to be something else. A bolder font which differs from the text. I have tried AllCaps feature but I do not like it so much...

nothing you have said says anything about the coding of the cv or names in it, so your question is just "how do I set some text in a different font in latex?" to which the answer is pretty much "there are lots of ways, depending on what you are doing, and which font you want to select."

exactly, my question is that...

@DavidCarlisle wow. you gave me the best guide... thank you...

I searched your typed question and found this:

@yo' ooh :)

\newcommand*{\myfont}{\fontfamily{<familyname>}\selectfont}

should I write the name of font after \selectfont?

it worked!

@EnthusiasticEngineer is that a different font to other sans serif in your document?

I wanted Avant Garde but it seems to be not different

in permeable: \newcommand*{\myfont}{\fontfamily{pag}\selectfont}

in main: \textcolor{black}{\fontsize{32}{16} \sc{ \textbf{\myfont{enthusiastic engineer}}}}\[3mm]

@EnthusiasticEngineer what I mean is, almost always the overall font choice is set up at the document level so to typeset that it should just be \textsf{EnthusiasticStudent}

@EnthusiasticEngineer that looks wrong, \sc shouldn't be in any document written after 1993, and the braces after it do nothing other than { adds a space. similarly the braces after \myfont do nothing.

@DavidCarlisle do we have other fonts like that avantguard?

@DavidCarlisle I should use \textsc instead of \sc? like \textbf?

@EnthusiasticEngineer or \scshape

\textcolor{black}{\scshape \fontsize{30}{16} { \textbf{\myfont aaa bbb}}}\[3mm]

@EnthusiasticEngineer what do you man by "like" there are thousands of fonts that differ in various ways. what font do you get for \textsf{enthusiastic engineer} ?

@DavidCarlisle serif series?

@EnthusiasticEngineer you still have spurious braces and spurious space tokens, and \fontsize{30}{16} makes no sense at all, how can you fit a 30pt font on baselines that are 16pt apart?

:((( if I were at your office, you would killed me now

@EnthusiasticEngineer I'd have to check, but I suspect that murder on company premises is against company policy.

@DavidCarlisle unless it's an emacs command

@PauloCereda emacs isn't mean: No apropos matches for ‘murder’

@DavidCarlisle ooh

@DavidCarlisle it probably means: emacs: enthusiastic murdered at carlisle's space :(((((

@PauloCereda do you think that we could arrange this? ^^^

@DavidCarlisle ooh yes

@DavidCarlisle sorry I do not know latex perfectly... reading this page is helping me understanding what you are trying to teach me...

en.wikibooks.org/wiki/LaTeX/Fonts

@EnthusiasticEngineer you are using low level commands that are designed for setting up document classes, and most people will use latex for years without needing them. What is wrong with \Huge \textsf{Enthusiastic} use other commands just if you need them,

@DavidCarlisle got it, thank you :)

@DavidCarlisle Even \scalebox in the worst scenario.

@JosephWright @DavidCarlisle the newest luatex works now with non-ascii file names:

luatex plain-test-grüße
This is LuaTeX, Version 1.11.1 (TeX Live 2019/W32TeX)
 restricted system commands enabled.
(./plain-test-grüße.tex (./grüße.tex) kpse FOUND        ./grüße.tex

@UlrikeFischer WOO

Das ist sehr gut!

@DavidCarlisle ^^ flawless German

@PauloCereda I could try with portuguese accents too ;-)

@UlrikeFischer ooh and spaces

@UlrikeFischer ^^

@PhelypeOleinik ^^ we are wacky :)

@PauloCereda works:

This is LuaTeX, Version 1.11.1 (TeX Live 2019/W32TeX)
 restricted system commands enabled.
("./O sabiá sábio encontrá-lo-á perto do órgão da igreja.tex"
("./O sabiá sábio encontrá-lo-á perto do órgão da igreja-input.tex") kpse FOUND         ./O sabiá sábio encontrá-lo-á perto do órgão da igreja.tex

@PauloCereda The one I know is: O sabiá não sabia que o sábio sabia que o sabiá não sabia assobiar

@PhelypeOleinik ooh :)

@UlrikeFischer YAY

@EnthusiasticEngineer still don't see why \myfont rather than \sffamily do you have two different sans serif fonts in the document?

@UlrikeFischer You grabbed latest version from W32TeX? Any set up needed beyond building the formats?

@JosephWright yes, but you really need the latest version, it took me a bit until I found id 7215 in the ring servers. It worked directly, but \file_full_name:n seems still not to work, which kpse command are you using here?

@DavidCarlisle \newcommand*{\myfont}{\fontfamily{pag}\selectfont} for AvantGarde font...

is AvantGarde a sans series font?

@UlrikeFischer local kpse_find = (resolvers and resolvers.findfile) or kpse.find_file

@UlrikeFischer Then local file = kpse_find(name, "tex", true)

@EnthusiasticEngineer yes, as I said, why? Why not \sffamily ? which would be the normal command. If you want all sans serif in the document to be pag then just \renewcommand\sfdefault{pag} It is really rather rare to be using \selectfont directly in a document. If you need two different sans serif fonts in the same document then yes declare a new family for pag and leave the default, but not many documents would want that.

Ah, it's on TLContrib: I'm trying from there

@DavidCarlisle then how should i use that avantgurd in a specific place...

@UlrikeFischer Seems to be fine here using TLContrib version ..

@EnthusiasticEngineer why do you want a different sans serif font there? where are you using \sffamily with the default sans serif (whatever that is in your document)

@JosephWright I may try compiling a cygwin one this evening

@JosephWright hm, it all works fine with direct kpse calls, but not with expl3:

[a b.tex]
[zzz-noext]
[space here/abc.tex]
[]
[]
XXXX    ./grüße2.tex
YYYY    ./space here/grüße.tex
ZZZZ    ./space here/grüße.tex

@UlrikeFischer Oh, the file business: I've not tested that yet

@DavidCarlisle in my header of cv... it is much nicer...

@DavidCarlisle Cool: was thinking about running the test suite

@EnthusiasticEngineer I didn't ask where you were using this font I asked where you are using \sffamily

@JosephWright was too tired last night to do filenames (or finish off oberdiek tests) having to be at M's school at 3 in the morning didn't help:-)

My plan for oberdiek is to simply disable any test that fails because it has set the swathes of characters to catcode 15. It does nothing useful except force dozens of lines into the package guarding against the test. I had been trying to more or less carefully adjust but I think a hard line is needed, it's taking too long to get them to ass (@UlrikeFischer @JosephWright)

@DavidCarlisle you mean in the permeable?

@EnthusiasticEngineer I give up sorry, this thread is too long:-)

@DavidCarlisle fine with me, it is certainly sensible to remove odd tests.

@DavidCarlisle sorry I made you tired or angry... my knowledge is too low. sorry

@EnthusiasticEngineer not tired or angry, just at work....

@HaraldHanche-Olsen -- Something fun to lay on my engineer nephews. Thanks.

@yo' -- He'll need double-woolly long johns and a big fur hat and mittens.

@barbarabeeton no worries, they can be purchased here (probably more easily than in São Paulo)

@yo' yay

Trivia fact of the day: In fossil, blame and praise are synonymous.

Something to think of next time someone wants to blame @UlrikeFischer

@PauloCereda they should have used siunitx the spacing of the unit would be better (@JosephWright)

@HaraldHanche-Olsen same in svn

@DavidCarlisle ooh

@DavidCarlisle what do they know that we don't?

@HaraldHanche-Olsen one man's bug is other man's feature. :)

@egreg up to his usual tricks again

@DavidCarlisle % ooh

where did dante go? downforeveryoneorjustme.com/dante.de

@DavidCarlisle Do you spot the space?

\pgfutilsolvetwotwoleqfloat #1#2->\begingroup \pgfmathfloatcreate {1}{1.0}{-4}\
let \thresh =\pgfmathresult \edef \pgf@temp {#1}\expandafter \pgfutilsolvetwotw
oleqfloat@A \pgf@temp \edef \pgf@temp {#2}\expandafter \pgfutilsolvetwotwoleqfl
oat@r \pgf@temp \pgfutilsolvetwotwoleqfloat@ifislarger \maa \mba {\def \Pa {a}\
def \Pb {b}}{\def \Pa {b}\def \Pb {a}}\expandafter \pgfmathfloatabs@ \expandaft
er {\csname m\Pa a\endcsname }\let \tmpa =\pgfmathresult \pgfmathfloatlessthan@
 {\tmpa }{\thresh }\ifpgfmathfloatcomparison \let \pgfmathresult =\pgfutil@empt

@DavidCarlisle Usual journey to hell? They haven't Internet down there, AFAIK.

@egreg ooh

First circle is where keming is

DIVISIOOOOOOOOOOOOOON, CLEAR THE LOBBY

@egreg There are two spaces, but one is probably harmless (between the two arguments to \pgfmathfloatdivide@, assuming that command really consumes two arguments).

@HaraldHanche-Olsen Yes, I suggested in the issue report to remove also the other one; but it's indeed harmless.

> Starting just before 12:00 a.m. UTC on November 3, 2019, iPhone 5 will require an iOS update to maintain accurate GPS location and to continue to use functions that rely on correct date and time including App Store, iCloud, email, and web browsing. This is due to the GPS time rollover issue that began affecting GPS-enabled products from other manufacturers on April 6, 2019. Affected Apple devices are not impacted until just before 12:00 a.m. UTC on November 3, 2019.

Friends, if someone has an iPhone 5, please update.

Friends, speaking of phones, do you use any mail app?

@PauloCereda only standard gmail one

So far, I have Gmail and Aquamail.

@DavidCarlisle Ah cool.

(by the way, Aquamail Pro)

I am about to replace my phone in the upcoming weeks, so I was wondering which app people are using. :)

I am considering ditching Gmail. :)

@PauloCereda just the app or as an email provider as well?

@DavidCarlisle just the app. :)

If you write doc.new in your browser address bar, it will create a new Google Docs document.

Google is now releasing their .new domain to these sorts of things.

@PauloCereda but will it start if off with tex syntax, and open it in emacs?

@DavidCarlisle er... no...

> The European Comittee agreed to postpone Brexit until 31st of February.

@yo' -- hmmm. What year?

@barbarabeeton Well, whatever year...

> One 17-inch pizza is actually more pizza than two 12-inch pizzas.

@PauloCereda -- Easy arithmetic, but the difference is marginal, assuming that both pizzas are circular. However, it's likely that the difference in cost per square inch is more noticeable.

@barbarabeeton indeed. :) It also made me hungry. :)

@JosephWright's cameo in frame #3. :)

ooh dr. Who reference

Hi mr. moderator person

@PauloCereda lol, n.6 for @DavidCarlisle

@CarLaTeX ooh that's mean

We should not tell @DavidCarlisle

@PauloCereda I am mean

@CarLaTeX OOH

@CarLaTeX isn't that standard software engineering practice?

@DavidCarlisle it is. :) Version 1.0, usually. :)

@DavidCarlisle Yes, it is! By the way, you're still an EU citizen :)

@CarLaTeX good

non è colpa mia

@DavidCarlisle I'm astonished by your perfect Italian!

@CarLaTeX Dovresti sapere che parlo molte lingue.

@DavidCarlisle Yes, I know :):):) perfect Italian again :)

@CarLaTeX Posso avere l'ananas sulla mia pizza?

@CarLaTeX this ^^ is what every real Italian would say, I know.

@DavidCarlisle This is wrong, you can't put pineapple and pizza in the same sentence, it's an Italian grammar rule :)

@CarLaTeX -- Grammatical quibble here: The sentence "you can't put pineapple and pizza in the same sentence" has both words in the same sentence. (Paraphrase of my sister's eternal comment "you can't put good and morning in the same sentence.") Not that I disagree with the sentiment -- either of them.

@barbarabeeton but it's an English sentence, but I think the implied meaning of @CarLaTeX's rule was that these can not go in an Italian one

@DavidCarlisle -- I'm willing to accept its application to English.

@DavidCarlisle Yes, I specified Italian rule :):):) @barbarabeeton

@MarcelKrüger I imported the context files and travis passed (with all tests, as I named the branch master-XXX) but locally I have a failure for math in harf: there is a again this curious line break. Could you try l3build check -cconfig-harf math in this branch?

@UlrikeFischer The test still passes for me.

@UlrikeFischer linebreak in a log or in the output?

@DavidCarlisle in the output, I get different glue and \abovedisplayshortskip instead of \abovedisplayskip.

@UlrikeFischer choose the 5th option above

@UlrikeFischer meanwhile I'm edging towards an oberdiek release....

@DavidCarlisle I'm edging towards a luaotfload release with "first harf implementation". But I just saw that Hans corrected the tounicode values for cm-unicode for old style numbers only in context and not in the generic fontloader <- @MarcelKrüger: the changes are in font-age and font-agl and this not imported by us.

@MarcelKrüger it passes for travis too, so you are in a majority ;-(

@DavidCarlisle ;-). For some reason we have different demerits and so a different breaking: chat.stackexchange.com/transcript/message/52262409#52262409. Is there a way to output a list of all defaults of demerits?

Hi. Has anyone ever used the langsci skeletons?

@UlrikeFischer exactly 1000 different

It works fine until I try to use arabic language- whatever I try fails.

@UlrikeFischer I don't really like that change so I will set up a overload based version. The ConTeXt fix seems to change the AGL but I think that this might risk breaking stuff if someone tries to reencode legacy fonts. (I never tested that code so that might be broken anyway)

@UlrikeFischer \showthe\adjdemerits ? (there are not that many demerit primitives)

@MarcelKrüger ok, I will try to add bit documentation tomorrow.