TeX, LaTeX and Friends

@egreg I've added \siunitx_if_number_token:N(TF) to the v3 experiments (github.com/josephwright/siunitx/commit/…): do you think \siunitx_if_number:N(TF) would also be useful (parse but don't flag issues)?

Danu

10:41 AM

o/

I've got a TikZ problem: I'm interested in drawing the set of points on $S^2$ with distance less or equal to some (small) $\epsilon$, so a disk in the $S^2$ metric

Does any of you recognize this question/has anyone thought about this before? I didn't find any relevant links in my own quick search. I could of course ask a new question on the site if this hasn't been asked before.

yo'

@Danu TikZ is not a math software, it's a drawing software. What you want seems to me more a job for math software.

Danu

@yo' But what if I want to draw my math? I managed to draw some decently mathematical stuff with it in the past :( In any case, it doesn't have to be exact. A good approximation would be fine too.

egreg

11:24 AM

@JosephWright If it's not much more work, it could be useful. It could help in defining macros where the final typesetting might need parse-numbers=false.

Joseph Wright

@egreg Indeed: the v3 refactor should make it reasonably easy

@egreg I'm working on the tables side of things at the moment: once that is working the code will probably be usable at least by me for testing :)

yo'

12:00 PM

@Danu well, you can want a lot of things, the question is whether the chosen tool is the right one :-)

CarLaTeX

@PauloCereda My first year on TeX.SE! Yearling badge earned: tex.stackexchange.com/users/101651/carlatex?tab=badges

Paulo Cereda

@CarLaTeX Yay! Congrats! Quack! <3

CarLaTeX

@PauloCereda :):):)

Danu

@yo' I think I realized that my problem is considerably simpler than I thought---the disks are probably just ellipses!

yo'

12:16 PM

@Danu much easier -- explicit plots are always easier than implicit ones

Danu

:-)

yo'

(that's basically the different between a drawing software and a math drawing software :-) )

barbara beeton

1:06 PM

@CarLaTeX -- and you've made a good contribution. stick around!

CarLaTeX

1:26 PM

@barbarabeeton Thank you, Barbara, surely! I'm also thinking to write a Duckcolumn for the TUGboat hahaha!

Paulo Cereda

1:45 PM

OpenSSL is relicensing their code under Apache 2.0!

Joseph Wright

3:00 PM

@PauloCereda I suspect this is quite an 'interesting' exercise!

Paulo Cereda

@JosephWright It is. They are sending a term of acceptance to more than 400 contributors.

> After nearly 20 years and 31,000 commits, OpenSSL wants to change to Apache License v2.0. They're now tracking down all 400 contributors to sign new license agreements, a process expected to take several months. Slashdot reader rich_salz shares links to OpenSSL's official announcement (and their agreement-collecting web site).

Joseph Wright

@PauloCereda Yes, I saw that. There are lots of very obvious questions, not only on the list of contributors but also on the issue of any license change at all leading to some 'interesting' views on different possibilities!

Paulo Cereda

@JosephWright Indeed. Most licenses are quite similar, but there are subtle differences that might impact greatly on the future of the project.

@JosephWright: There's also an 'interesting' discussion on the Drupal list. Have you seen?

Joseph Wright

@PauloCereda I hope they know what they are doing!

@PauloCereda No: link?

Paulo Cereda

@JosephWright Me too. I do not know enough of the Apache license to comment, but I believe it's a less viral one, at least compared to GPL. But I do not have any quarrels with GPL...

@JosephWright It's... well... a rather unusual context: techcrunch.com/2017/03/26/sex-and-gor-and-open-source

Joseph Wright

3:12 PM

@PauloCereda Very strange

Paulo Cereda

@JosephWright you can say that again.

@Joseph: in other news, how's the Steam wishlist? :)

Joseph Wright

@PauloCereda See email

Paulo Cereda

@JosephWright on it.

Joseph Wright

@PauloCereda Not going far: I'm trying to finish The Witcher or rather to complete the middle game before doing TW3 a second time. But there is one really hard part I'm stuck with ...

@PauloCereda And of course I have siunitx v3 to sort ...

@PauloCereda Speaking of licenses, I see GitHub now gives a short 'summary' of any LICENSE file: see e.g. github.com/josephwright/siunitx/blob/master/LICENSE

Paulo Cereda

@JosephWright ooh

David Carlisle

3:30 PM

@JosephWright their description of LPPL?

Joseph Wright

@DavidCarlisle Yes, or rather a set of extracted ideas

@DavidCarlisle Quite reasonable set of bullet points I think

David Carlisle

@JosephWright yes I'm sort of surprised it was high enough up their radar for them to do that:-) unless they just listed all the licences listed as free by debian or whatever

Paulo Cereda

@DavidCarlisle: github.com/supertunaman/cdl/blob/master/COPYING

Joseph Wright

@DavidCarlisle I believe GitHub are using some third-party library to do the work. Presumably someone somewhere has listed in a simplified form the basics of a large number of OS licenses, so it's just a question of picking the right one and grabbing the statements.

@DavidCarlisle There is a reason the commit message suggests I was testing if it would work at all ...

@DavidCarlisle, @PauloCereda I like the last part in github.com/benbalter/licensee/blob/master/docs/…: 'What about checking every single file for a copyright header?': 'Because that's silly in the context of how software is developed today.' Erm, that's what we do ...

Paulo Cereda

3:48 PM

@JosephWright I have subprojects with different licenses...

egreg

@JosephWright “You wouldn't put a copyright notice on each page of a book.” Well, a publisher I worked with wanted the copyright notice and the ISBN number on each page!

Joseph Wright

@DavidCarlisle The full list is I think choosealicense.com/appendix, so we make the 'top' ~30

@egreg :)

@egreg Each chapter is common nowadays as they tend to be available as separate PDFs

Paulo Cereda

> Optional: Add LPPL-1.3c to your project's package description, if applicable (e.g., Node.js, Ruby, and Rust). This will ensure the license is displayed in package directories.

Yes, LPPL in Ruby. :)

egreg

@JosephWright He wanted to scare people trying to xerox the books. I don't know how successfully. I tell my students that photocopying a book is illegal and that, at the least, they shouldn't pull out the photocopied book when they come to my office.

Joseph Wright

@PauloCereda ?

Paulo Cereda

3:53 PM

@JosephWright choosealicense.com/licenses/lppl-1.3c Top right.

Joseph Wright

@PauloCereda Ah, I see

@PauloCereda Like I said, interesting that it makes the list when you consider that it's really only used by LaTeX (kernel + packages)

Paulo Cereda

@JosephWright indeed

yo'

@PauloCereda well, interesting is that the two limitations they list are both known of and -- to an extent -- also desired.

Paulo Cereda

@yo' indeed!

Joseph Wright

4:24 PM

@egreg github.com/josephwright/siunitx/commit/…

egreg

5:23 PM

@JosephWright :)

Johannes_B

5:35 PM

@TorbjørnT. You are a member of LaTeX-community, right? You may know how to do this. latex.org/forum/…

Ulrike Fischer

@Johannes_B A simple \renewcommand\sf{\normalfont\sffamily} or \DeclareOldFontCommand{\sf}{\normalfont\sffamily}{\mathsf} should work too.

Johannes_B

@UlrikeFischer Yes. But that might be a bit heavy for a LyX starter.

Ulrike Fischer

@Johannes_B Adding some tex code in the preamble shouldn't be difficult for them.

Torbjørn T.

@Johannes_B Answered.

Johannes_B

@UlrikeFischer I added it. Thanks to you and @TorbjørnT.

StrongBad

5:58 PM

Any other Linux users having problems with Windows users opening LaTeX made PDFs?

Paulo Cereda

@StrongBad Not that I can think of. What happened, pal? :)

StrongBad

I have made a couple of PDFs (they have some common elements) emailed them from my government email to other people's gmail accounts and Acrobat on Windows says they are corrupt, but other viewers work.

Not sure where in the chain the problem is.

David Carlisle

6:19 PM

@JosephWright new luatex (still 1.04) running test suite again just in case

Joseph Wright

@DavidCarlisle :)

David Carlisle

7:05 PM

@JosephWright test suite still running (laptop shut down while I was eating:-) but results so far....

wildcard *diff
-rw-r--r-- 1 davidc Domain Users 671 Mar 27 19:19 tl2e2.luatex.diff
-rw-r--r-- 1 davidc Domain Users 1150 Mar 27 19:20 tl2e7.luatex.diff
-rw-r--r-- 1 davidc Domain Users 1205 Mar 27 19:25 tlb0380.luatex.diff
-rw-r--r-- 1 davidc Domain Users 2180 Mar 27 19:26 tlb0438.luatex.diff
-rw-r--r-- 1 davidc Domain Users 720 Mar 27 19:30 tlb1451.luatex.diff
-rw-r--r-- 1 davidc Domain Users 1295 Mar 27 20:03 tlb2727.luatex.diff
-rw-r--r-- 1 davidc Domain Users 360 Mar 27 19:20 tlb-callbacks-001.luatex.diff

Joseph Wright

@DavidCarlisle Oh good

@DavidCarlisle Some of those look 'new'

David Carlisle

@JosephWright yes I'll let it get to the end before looking at the tl*diff files but three new hyphenation primitives, so i would guess hyphenation changed again:

+ automatichyphenpenalty
+ explicithyphenpenalty
+ hyphenpenaltymode

Joseph Wright

@DavidCarlisle That would do it, yes

@DavidCarlisle Do we know what they do?

@DavidCarlisle I guess I should add to expl3

David Carlisle

No I'll see if they are in the new manual in a bit

Because we tend to make hard codes behaviour configurable a few new primitives have been
added:

\hyphenpenaltymode
\automatichyphenpenalty
\explicithyphenpenalty

The first parameter has the following consequences for automatic discs (the ones resulting from
an \exhyphenchar:

mode automatic disc -    explicit disc \-

0 \exhyphenpenalty       \exhyphenpenalty

1 \hyphenpenalty         \hyphenpenalty

2 \exhyphenpenalty       \hyphenpenalty

3 \hyphenpenalty         \exhyphenpenalty

4 \automatichyphenpenalty \explicithyphenpenalty

@JosephWright ^^

Joseph Wright

@DavidCarlisle Ah

@DavidCarlisle This is good?

David Carlisle

7:15 PM

@JosephWright well it seems to imply that mode 0 (which is what I would expect we have) should work as before, but something has broke the other tests, let me look at a diff...

oh, not what I expected...

*** ../build/test/tl2e2.luatex.tlg	2017-03-27 19:19:43.692486200 +0100
--- ../build/test/tl2e2.luatex.log	2017-03-27 19:19:46.422656300 +0100
***************
*** 256,262 ****
  .......\hbox(6.94444+1.94444)x10.00002, direction TLT
  ........\glue(\tabskip) 0.0
  ........\hbox(6.94444+1.94444)x10.00002, direction TLT
! .........\OMS/cmsy/m/n/10
  ........\glue(\tabskip) 0.0
  ....\mathoff
  ....\penalty 10000
--- 256,262 ----
  .......\hbox(6.94444+1.94444)x10.00002, direction TLT
  ........\glue(\tabskip) 0.0

Joseph Wright

@DavidCarlisle Erm, that is odd

@DavidCarlisle We have some normalisation for blank lines but that's because LuaTeX doesn't (or didn't) print ^^M ...

David Carlisle

@JosephWright the etex one has .........\OMS/cmsy/m/n/10 ^^M but the existing luatex.tlg was showing nothing presumably it had a literal newline there so I guess it's all good

Joseph Wright

@DavidCarlisle That would make sense but perhaps check with Hans that it's deliberate!

Paulo Cereda

@JosephWright uh-oh

David Carlisle

@JosephWright he'lll tell me not to look at the logs:-)

Joseph Wright

7:20 PM

@DavidCarlisle Probably

Paulo Cereda

@DavidCarlisle What has John Napier ever done for us? :)

David Carlisle

@JosephWright I think something is confusing the l3build normalisation with control characters:

*** ../build/test/tlb1451.luatex.tlg	2017-03-27 19:30:10.788400300 +0100
--- ../build/test/tlb1451.luatex.log	2017-03-27 19:30:10.797401300 +0100
***************
*** 356,363 ****
  ...\hbox(6.59525+1.94444)x20.35776
  ....\OML/cmm/m/it/10 i
  ....\glue(\medmuskip) 2.22217 plus 1.11108 minus 2.22217
! ....\OMS/cmsy/m/n/10 ^^@
! ....\glue(\medmuskip) 2.22217 plus 1.11108 minus 2.22217
  ....\OML/cmm/m/it/10 j
  ....\kern0.57243
  ...\mathoff
--- 356,362 ----
  ...\hbox(6.59525+1.94444)x20.35776
  ....\OML/cmm/m/it/10 i

Joseph Wright

@DavidCarlisle What do the raw logs show? Working out where to wrap lines is not always easy

David Carlisle

@JosephWright test suite still running I'll collect up in a bit

Joseph Wright

@DavidCarlisle Cool

David Carlisle

7:25 PM

@JosephWright meanwhile the new callback is + - glyph_stream_provider (not in the manual)

Joseph Wright

@DavidCarlisle Oh goody, something not in the manual!

Paulo Cereda

@JosephWright undocumented features!

David Carlisle

@JosephWright shocking (although to be fair I did just randomly build off whatever is at the head of their experimental branch)

Paulo Cereda

@DavidCarlisle: I played the UK version of Monopoly in my Xbox. I bought a couple of places I have no idea of what they are. :)

David Carlisle

@PauloCereda is it the classic London one with Mayfair being the most expensive one?

Paulo Cereda

7:35 PM

@DavidCarlisle Yes, I was about to mention that! Mayfair made me go bankrupt. :)

David Carlisle

@PauloCereda that is the idea of the game

@JosephWright test finished full list is

  Check failed with difference files
  - ../build/test/tl2e2.luatex.diff
  - ../build/test/tl2e7.luatex.diff
  - ../build/test/tlb-callbacks-001.luatex.diff
  - ../build/test/tlb-extraprimitives-001.luatex.diff
  - ../build/test/tlb0380.luatex.diff
  - ../build/test/tlb0438.luatex.diff
  - ../build/test/tlb1451.luatex.diff
  - ../build/test/tlb2727.luatex.diff
  - ../build/test/tltc001.luatex.diff
  - ../build/test/tltx001.luatex.diff
  - ../build/test/tnfss7.luatex.diff
  - ../build/test/vlatex02.luatex.diff

Joseph Wright

@PauloCereda Do you usually play the Atlanta one?

@DavidCarlisle More than I'd expect

David Carlisle

@JosephWright odd. diffs for etex and xetex in there as well...

Paulo Cereda

@DavidCarlisle I deserved to win, I had two railways, including King's Cross. I thought I should get a lot of money because of Harry Potter fans...

Joseph Wright

@DavidCarlisle That is odd

Paulo Cereda

7:38 PM

@JosephWright No, is it good?

Joseph Wright

@PauloCereda I've never played it, I'm afraid, but it is the original ...

Paulo Cereda

@JosephWright cool, I will take a closer look!

David Carlisle

@PauloCereda railway stations (even if you have all 4) are pretty useless if your opponent has a hotel on Mayfair

Paulo Cereda

@DavidCarlisle Ouch. I think the CPU had Park Lane as well, which completes the colour set.

David Carlisle

er what did we do with hyphens in tt the other day....

*** ../build/test/vlatex06.tlg	2017-03-27 20:31:52.880909200 +0100
--- ../build/test/vlatex06.etex.log	2017-03-27 20:31:58.056226900 +0100
***************
*** 945,950 ****
--- 945,962 ----
  *** Bug 82 ***
  *** Bug 85 ***
  \hyphenbox=\skip43
+ Underfull \hbox (badness 10000) in paragraph at lines 167--167
+ []\OT1/cmtt/m/n/10 gnomon
+ \hbox(4.30554+2.22223)x36.74968
+ .\hbox(0.0+0.0)x0.0
+ .\OT1/cmtt/m/n/10 g
+ .\OT1/cmtt/m/n/10 n
+ .\OT1/cmtt/m/n/10 o
+ .\OT1/cmtt/m/n/10 m
+ .\OT1/cmtt/m/n/10 o

David Carlisle

7:57 PM

@JosephWright guess what \let\-\@dischyph does in \@parboxrestore answers on a postcard to .....

David Carlisle

8:14 PM

@JosephWright running test suite again:-)

Joseph Wright

@DavidCarlisle Good stuff

David Carlisle

@JosephWright if I'd used etex protected I wouldn't have broken it, it's just that I had to move the definition until after \DeclareRobustCommand was defined:-)

ilFuria

8:58 PM

Hello

is anyone here? :)

Joseph Wright

@ilFuria Yes

ilFuria

I have kind of a question about answering a… question. Is this the right place to ask?

(Hi @JosephWright)

Joseph Wright

@ilFuria Don't ask if you can ask, just ask ;)

ilFuria

Right. So this question tex.stackexchange.com/questions/360694/… is very similar to this one tex.stackexchange.com/questions/331840/…. I tested its answer that, with minor tweaking, works. What it's best to do? Simply point to the aforementioned answer or to post a new specific one?

@JosephWright Ok, the tweaking exceeded the "minor" range, I think that a proper answer is in order. Thanks for the support and the rubberducking!

egreg

@ilFuria You're welcome!

Alan Munn

9:17 PM

@egreg I hadn't realized how much of an expl3 evangelist you are. :)

egreg

@AlanMunn :)

Alan Munn

@egreg Perfectly good \csname ... \endcsname solutions get alternative answers now from you!

egreg

@AlanMunn More modern, not like old fashioned fellows like @DavidCarlisle

David Carlisle

@AlanMunn you only just noticed?

Alan Munn

@DavidCarlisle I guess I'd not had the personal experience of purely gratuitous modernisation. :)

Joseph Wright

9:26 PM

@DavidCarlisle I see you went for \protected ;)

Alan Munn

@DavidCarlisle So far though the answer hasn't resulted in a tick theft.

David Carlisle

@JosephWright well it was that or work out where memoir was expanding stuff it shouldn't have been expanding (\protect didn't work:-)

Joseph Wright

@DavidCarlisle In other news, I'm being reminded how awkward working around colortbl is ...

@DavidCarlisle :)

@DavidCarlisle Easier to just go 'oh well, use e-TeX'

David Carlisle

@JosephWright well it's a fairly delicate balancing act as it is so not really designed to be worked around

Joseph Wright

@DavidCarlisle :)

@DavidCarlisle I'm working on siunitx v3 and it has to deal with spacing in columns: I'm just thinking again about the best set up for colortbl correction

David Carlisle

9:43 PM

@JosephWright tlb3855.lvt still fails in etex (it is testing hyphenation in tt explictly) not sure why yet

@JosephWright oh that's OK it's supposed to fail, it shows an intentional change, phew (I'll add a version with latexrelease to check the 2016 behaviour isn't changed

barbara beeton

@JosephWright -- not atlanta ... atlantic city. (and yes, that is the original one. i've never seen the london version.)

BFH

10:13 PM

Is there any huge reason to use the memoir class as the base class for my school's thesis class which I am designing? It seems to be the fashionable thing, but my book based class works just fine.

David Carlisle

@BFH yes or no, depending on who you ask:-) lots of thesis classes are based on book so that can work fine, but if you find yourself re-inventing 7 of memoir's eight wheels using that as a base can also work.....

BFH

@DavidCarlisle Is there a compilation speed benefit, or is it just for convenience?

David Carlisle

@BFH latex isn't a compiler in that sense so nothing extra is compiled, if you use memoir it's a big class so you need less packages so that is possibly a bit quicker than loading the classes separately, but if you don't actually need all the functionality just loading book and not loading the classes might be quicker, but I'd be astonished if the loading difference was within orders of magnitude to say the time it takes to load tikz

Alan Munn

@BFH As someone who uses memoir for the class, it speeded up development and maintenance a lot for me. It also allows users who need last minute modifications more easily if necessary.

Paul Stanley

10:45 PM

Is there any way to close a question as an exact duplicate and merge the answers? Someone has managed to ask two (bad!) identical questions, and each has been answered once.

Obviously one must go! But maybe simpler just to repost my answer in the open one once the other's closed. If I hadn't been bored, I wouldn't have bothered anyway ... tex.stackexchange.com/questions/360538/… vs tex.stackexchange.com/questions/360669/… (which needs some more close votes too!)

David Carlisle

@PaulStanley you can flag it for moderator attention, they have powers to do such things (although they do not use them that often)

Paul Stanley

@DavidCarlisle: I think this one has already wasted enough time already. Better just close it and if I feel moved to copy my answer over to the other thread I will do that.

OK. Deleted my answer on the duplicate second thread and moved it to the other one. Who needs moderators!

Transcript for

$TeX - LaTeX$ TeX, LaTeX and Friends