TeX, LaTeX and Friends - 2015-07-09 (page 1 of 2)

« first day (1712 days earlier) ← previous day next day → last day (3222 days later) »

2:02 AM

@AdamLiter You suggestion on the three knitr questions has merit. I need to carefully reread each of the questions, the answers, and the comments to see if we would need to edit in some information to the main question and answer. (And I think it probably be the one with your answer)

@EnthusiasticStudent Yeah, I just went with article myself. Basic, I know what I'm working with, and I have the easy freedom to really go ham with it. It's worked out pretty well for me: seanallred.com/assets/cv.pdf

@JosephWright Believe it or not, this was one of the triggering stumbling-blocks behind my object system. Some elements of my CV just have too many qualities. (Yes, keys and props, but they didn't work at the time for whatever reason.)

Speaking of, I remember @WillRobertson wanted to review that. The documentation is complete and the source is on GitHub.

@EnthusiasticStudent The limit really is your imagination

And, well, reality. (At least until we get those moving photographs in print like they have in the Harry Potter universe!)

2 hours later…

4:01 AM

tex.stackexchange.com/a/254430/17423 – this would have been a lot easier if \clist_use: supported a 'wrapping' function (similar to \ReducedFraction#1)

@SeanAllred Hmm. Good point. I don't suppose that you could wrap the items first with \clist_map before passing them to \clist_use ?

@WillRobertson Can \clist_map alter its target clist?

I opened up an issue on GitHub a few minutes ago, by the way

(But alas, it's bedtime :))

Good night!

@SeanAllred No, unfortunately! You'd need something like:

\documentclass{article}
\usepackage{expl3}
\begin{document}
\ExplSyntaxOn
\cs_set:Nn \my_wrap:n { \exp_not:n {$#1$} , }
\clist_set:Nn \l_tmpa_clist {\a,\b,\c}
\clist_set:Nx \l_tmpa_clist { \clist_map_function:NN \l_tmpa_clist \my_wrap:n }
\clist_show:N \l_tmpa_clist
\end{document}

4 hours later…

Enthusiastic Student

8:05 AM

@SeanAllred I really like how your CV looks like, it's nice indeed. Can I have it's .tex file? Feel free to say no if you are not comfortable to share it...

@SeanAllred I think we already have those moving photographs in form of gif extensions! :))) and moving graphics in our electronic devices!

@EnthusiasticStudent Probably I should get my CV on my work website: good job you reminded me :-)

Enthusiastic Student

@JosephWright the pdf file of your CV is on your website (the link you provided for me) I think.

@EnthusiasticStudent Yes, but that's my own site not my work one

Enthusiastic Student

@JosephWright Aha. One question I had last night... Do you import your papers in to your CV with a bib file or type them one by one?

You have many papers, that's really appreciated

@EnthusiasticStudent By hand

@EnthusiasticStudent I used to use BibTeX, but for a CV you have to control the order exactly and need to fiddle with the formatting a bit

@EnthusiasticStudent Chemistry paper titles are tricky to format correctly (we don't normally print them in references)

@DavidCarlisle See GitHub checkins for ltluatex

Enthusiastic Student

8:13 AM

@JosephWright I also do it by hand, but always had the feeling that I am doing a wrong work. Thanks.

@EnthusiasticStudent BibTeX is great for many things, but my experience here was that at least in my case I did as much work reformatting stuff in the .bib file as I would need to do working by hand

@EnthusiasticStudent I needed a separate .bib file just for my CV, so there was no real saving

Enthusiastic Student

@JosephWright Could you please share your source tex file of your CV?

@JosephWright will check, seem not to be getting mail notification again (I noticed yesterday hat I'd missed one)

Christian Hupfer

8:28 AM

From a German book seller, a brand 'new' copy, for 3 Euro

user image

It's quite up-to-date ;-)

LaTeX2e <2015/01/01> patch level 2

Enthusiastic Student

@ChristianHupfer A new book for only 3 Euro?! It's cheap indeed...

Christian Hupfer

@EnthusiasticStudent: Yes, stock sale ... The TeXBook is not available there, unfortunately

Enthusiastic Student

@ChristianHupfer :-(

Christian Hupfer

And the famous see p 192 for the unknown command sequence error message -- it's really there :D

@EnthusiasticStudent: Well yes, it is there, but 56 Euro is ... expensive.

Enthusiastic Student

8:35 AM

@ChristianHupfer That's about as expensive as a new copy of TeXbook which is available on Amazon...

Christian Hupfer

@EnthusiasticStudent Yes, exactly

I was luckily enough to buy the millennium collection from the US amazon when the dollar was really cheap.

Christian Hupfer

@daleif: Well, I have to buy it now :-( But it is worth the prize, I think

Enthusiastic Student

The used copies are also available for the price of 3 to 4 dollar, but I am uncomfortable with buying used copies most of the times...

Christian Hupfer

@EnthusiasticStudent: Yes, I don't like used copies too

8:42 AM

@ChristianHupfer I have to admit, I haven't used it much. And haven't looked much in the other books in the collection

@ChristianHupfer Still in print

@EnthusiasticStudent It's all on my blog I think (will check)

Enthusiastic Student

@JosephWright thanks

Christian Hupfer

@JosephWright Yes, apparently. Well, anyway, it's quite cheap for 3 euros

@EnthusiasticStudent texdev.net/wp-content/uploads/2011/11/cv.zip

@EnthusiasticStudent Probably I should do an update

Enthusiastic Student

@JosephWright that's fine. I just wanted to see how you organize your CV tex document... you did great. thanks.

8:48 AM

@DavidCarlisle There is some sneaky rebasing in today's commits :-)

9:22 AM

@PauloCereda chat.stackexchange.com/search?q=duck

10:13 AM

@JosephWright no whatsits yet? also i think some write_nl should be texio_write_nl (after earlier rename?)

10:29 AM

A new blog entry from @clemens about templates. mychemistry.eu/2015/07/the-template-story

3

10:53 AM

@DavidCarlisle Whatsits are next on my list

@DavidCarlisle Will check rename stuff

@JosephWright I could send a pull request but as you're here....

$ diff latex2e.lua~ latex2e.lua
17c17
<   module[name] = info
---
>   modules[name] = info
24a25
>      leader=""
59c60
<     write_nl("term and log", i)
---
>     texio_write_nl("term and log", i)
61c62
<   write_nl("term and log", "\n")
---
>   texio_write_nl("term and log", "\n")

@DavidCarlisle Well you've got full access so you could just commit to the repo, but I'll do it :-)

@JosephWright trying to get this document to run (not quite there yet, but need to leave it for a bit)

\makeatletter
\@namedef{ver@luatexbase.sty}{}
\def\RequireLuaModule#1{\directlua{require("lt-#1.lua")}}
\RequirePackage{ltluatex,lualatexsupport}
\directlua{require("latex2e.lua")}


\documentclass{article}

\usepackage{fontspec}
\setmainfont{Linux Libertine O}
\setsansfont{Linux Biolinum O}

\begin{document}
\showoutput
abc \textsf{xyz}

\end{document}

@DavidCarlisle You'll need at least \directlua{luatexbase = latex or { }} or similar, plus some TeX fix-ups

@DavidCarlisle I've not done any of that cross-compatibility stuff yet

@JosephWright yes was going to say that, if I had a fuly working document perhaps but currently no point me adding half thought ideas as i go along I suspect

@JosephWright yes had that (but currently trying without, and modifying copies of luaotfload instead) it only references luatexbase in half a dozen places....

11:00 AM

@DavidCarlisle Indeed, was what I was also thinking

@JosephWright luaotfload uses the local callback stuff, haven't quite worked out if that's essential yet

@DavidCarlisle You mean the custom callbacks: I think that's for third-parties

@JosephWright yes and yes

@DavidCarlisle There's a difference in how I've set up provides_module from the luatexbase version: I think mine is more sensible! I can adjust so it also works with the luatexbase approach: thoughts?

@Johannes_B, @clemens Excellent blog post

@JosephWright I noticed:-) one reason I just decided to edit a modified luaotfload, that and I couldn't be bothered how to fake the array of attributes lookup over a function call to registername(). I suspect any code is going to need bits of adjustment anyway, so I don't think I'd work too hard to make the syntax completely compatible.

11:13 AM

@DavidCarlisle I was thinking much the same: agree on a new set of interfaces, get luatexbase and packages using it adjusted, only once that has propagated release a kernel update

@DavidCarlisle Would welcome your thoughts on provides_module: does my approach seem sensible?

@JosephWright I'll go back and look, really I just mechanically avoided the syntax error and moved on:-) I was hoping to get that document to load, then look at what I'd had to change and make some notes, but out of time for now.

@DavidCarlisle OK

@DavidCarlisle No rush: no-one else seems interested :-(

@JosephWright well if you can add whatsits you can go back to latex-l and point out that modules and they have been added as suggested, good to keep the discussion loop going in public even if it's just you me and Elie, should at least assure Karl that we are not planning to break every document that ever used luatexbase....

@DavidCarlisle OK, whatsits by the end of the day :-)

@DavidCarlisle Will also do some testing myself a bit later on

11:29 AM

@JosephWright thanks

@clemens Works :-) mychemistry.eu/2015/07/the-template-story/#comment-868

@Johannes_B nice :)

11:50 AM

@Johannes_B Not enough duck hits. :)

12:01 PM

@PauloCereda I did not know that existed:

Rubber duck debugging

Rubber duck debugging is an informal term used in software engineering for a method of debugging code. The name is a reference to a story in the book The Pragmatic Programmer in which a programmer would carry around a rubber duck and debug their code by forcing themselves to explain it, line-by-line, to the duck. Many other terms exist for this technique, often involving different inanimate objects. Many programmers have had the experience of explaining a programming problem to someone else, possibly even to someone who knows nothing about programming, and then hitting upon the solution in the...

3

@Johannes_B ooh

@EnthusiasticStudent It's up on GitHub github.com/vermiculus/resume-cv

Keep in mind that it's slowly going a refactoring to use l3coffins instead of all those blasted minipages

12:26 PM

My songbook arrived from the printing plant. It's insanely impressive, TeX is awsome!

6

@PauloCereda Yaay! Congratulations :)

@SeanAllred Thank you! <3

@PauloCereda prize for whoever finds first typo?

@PauloCereda oooh! Do you have pictures?

@DavidCarlisle I never maek mistakes. :)

@clemens Sure, I'll take photos, gimme some minutes. :)

12:42 PM

0

Q: How can I apply a regex (l3regex) on items in a sequence? [LaTeX3]?

I am trying to apply a regular expression to a string item from a sequence in LaTeX3. Example Compiles with error, but only by pressing return many times. ! Missing control sequence inserted. <inserted text> \inaccessible l.64 \listprint[separator={ \par }] Documents.txt Inp...

^^^ Why still separating on \par, compare with tex.stackexchange.com/questions/254247/…, from yesterday

@egreg, @DavidCarlisle, @cfr, @JosephWright, @Johannes, @clemens, @barbarabeeton, @SeanAllred and all:

user image

user image

user image

user image

user image

user image

user image

Powered by memoir! :) @daleif ^^ :)

@PauloCereda :) cool

@PauloCereda Nice, what are you using for the columns?

@daleif Frank's multicols. :)

@PauloCereda thought so

12:53 PM

@daleif standing on the shoulders of giants <3

Enthusiastic Student

1:17 PM

@PauloCereda nice typesetting. what is the title of the book? are you the author?

@SeanAllred thanks

@PauloCereda -- very nice job! everyone who uses it will be proud. question ... on the index page, the background on the "banner" for "D" gets gradually paler toward the right (it's the only one i see). is this intentional, and if so, why just one letter? (mainly curious.)

Enthusiastic Student

@barbarabeeton It may be a printing issue or because of the light focus on that part, I think...

@barbarabeeton As I can see, the light on the title page on the left part of the page is more...

@EnthusiasticStudent -- good point. on the title page, there's clearly a bright spot, and that's just about the same place that it would fall with the book open. @PauloCereda -- i retract my question, unless you have some other explanation.

1:41 PM

Yep, that's definitely just a lighting thing

@barbarabeeton Oh it's the light. :) My office is very dark and there was a lamp over the book for me to take the photos. :)

@EnthusiasticStudent It's a songbook to be used during masses. :) And yes, I am the author of the book (actually more like an editor). :)

2:07 PM

@PauloCereda does it include the song "happy birthday to you" ?

Enthusiastic Student

@PauloCereda I hope you sell many copies of your book. :-) You did great.

@DavidCarlisle s/you/Jesus/g :)

@EnthusiasticStudent That's the beauty of it: we will offer it for free for people to use it during masses. :)

@PauloCereda not so useful then.

@DavidCarlisle Unless Disney tracks me down to pay them royalties. :)

Enthusiastic Student

@SeanAllred many class and style files... I think it's too complicated for me to understand your files. But your CV looks really nice. I wish mine was as beautiful as yours. First I saw your CV, it never came to my mind it may have about six or seven styles and tex files...

2:19 PM

@PauloCereda Great job!

2:36 PM

@egreg Thanks! You all helped me. :)

@PauloCereda You're not showing the acknowledgments page. ;-)

@egreg I couldn't even add one because the priest said people would be jealous. :) So no one knows who did it.

@egreg: I even added longtable in the preamble! :)

2:55 PM

@EnthusiasticStudent It's just an organizational tactic :)

@PauloCereda Best kind of work :)

@PauloCereda Bendigedig!

3:10 PM

@cfr Quack! <3

XeTeXinterchartoken i cannot help here: latex-community.org/forum/…

@Johannes_B It's not \XeTeXinterchartoken, it's the fact they've put in a kern (try one by hand)

    \documentclass{article}
    \usepackage{ucharclasses}

    \newXeTeXintercharclass \QuotationMarkClass
    \XeTeXcharclass `` = \QuotationMarkClass
    \XeTeXcharclass `' = \QuotationMarkClass

    \XeTeXinterchartoks \QuotationMarkClass \BasicLatinClass {\nobreak\hskip 1pt}
    \XeTeXinterchartoks \BasicLatinClass \QuotationMarkClass {\nobreak\hskip 1pt}

    \hyphenation{hair-space}

    \begin{document}

    \XeTeXinterchartokenstate=0

    Between quotation marks and the quoted text there must needs be a `hairspace'.

Perhaps

@JosephWright I have no idea, and to be honest, i am not in the mood to look closer at it. I don't even know where to look for the doc.

@Johannes_B OK, I'll post an answer in a bit then :-)

3:27 PM

@JosephWright Thanks :-)

Is there any way to post a question but to not publish it?

I can't create a proper example which is sort of in the direction of what I want to do... :/

I hate to post something which can't be distinguished from showing no effort at all to answer the question (even though I have been trying for weeks to get something done...)

@1010011010 Try rubber duck debugging (see starred message on the right) ;-)

@Johannes_B Thanks, though my question is conceptual, not so much debugging.

@1010011010 it will work just as well, and i think it applies perfectly to your situation.

Stefan Kottwitz

@JosephWright I registered for the TUG meeting in Darmstadt :-)

4

3:36 PM

@StefanKottwitz Cool!

@StefanKottwitz I'm still pondering exactly when I'll get there

@Johannes_B Is it sort of clear what I want to do? tex.stackexchange.com/questions/254524/… I'll settle with deleting and undeleting the question as I add more examples.

Stefan Kottwitz

@JosephWright Will you arrive at some time on Sunday?

@StefanKottwitz Currently the plan

Stefan Kottwitz

@1010011010 I guess I take a flight on Sunday, it's hard to get one Monday morning

@StefanKottwitz Certainly

3:41 PM

@1010011010 to be honest, i couldn't even finish reading the question. I need coffee.

@clemens Can you remove the test-comment, or at least tell me where/how i can delete it?

@1010011010 you just need to write the values to the aux file, sorting is easy enough or I guess expl3 has some sort routines, but do you really have to do this? what is the actual use case? You ought to know where anything is going without asking pdfsavepos where it went:-)

@StefanKottwitz Question for me is whether to travel on Saturday to meet up with @DavidCarlisle and Frank Mittelbach

@DavidCarlisle Bruno wrote a tree-based sort :-)

@JosephWright May I ask you to have a look at this question? Maybe you have a hint how to get siunitx working with \mathversion{sans} provided by MyriadPro.sty.

@JosephWright seems like Kaveh's setting the agenda for sunday morning:-)

@JosephWright yes but for lists of length 4, setup costs might be a bit high:-)

@ArashEsbati For the more esoteric units you have to code the math mode into the symbol definition, so for complex font set ups you need to redefine the unit using \DeclareSIUnit

@DavidCarlisle At present: if Frank gets his way (see agenda for team meeting) we should be chasm-jumping soon

@DavidCarlisle You are going to have to finish xor ;-)

3:46 PM

@JosephWright I'm going to write it all in lua

@PauloCereda Cwac Cwac!!

@Johannes_B I already deleted it. The first real reaction is there BTW

@clemens I didn't refresh the page :-) Right now it seems you are running some updatess Briefly unavailable for scheduled maintenance. Check back in a minute. :-)

@JosephWright testing write access...

@Johannes_B update done :)

3:55 PM

@clemens papeeria. We have the in the community ads list, right?

@clemens it was literally one minute.

@Johannes_B I don't know. I wasn't familiar with Papeeria until today

24

A: Community Promotion Ads - 2015

And look, how nice. Templates

@Johannes_B I can't say that I like their position very much:

@LaTeX_Chemistry right, but if \bf makes text bold and it is exactly what I want to see on paper then it is okay

Tweeted by Papeeria on July 9, 2015 at 1:59 PM in reply to LaTeX_Chemistry

@cfr I don't understand Welsh.

@clemens No

@clemens The 'if they do their job' business is quite depressing: that's how you end up having to spend ages fixing stuff later, whether in LaTeX or anything else

4:06 PM

@JosephWright exactly

@clemens For a one-page one-off job you can justify hacking stuff up, but not for a thesis

@JosephWright got past luaotfload, now dying in fontspec, making progress:-)

@DavidCarlisle Great

@JosephWright of course until I get to the end no idea if it's doing the right thing but having the right symbols and things defined before use seems to be good.

@clemens Seems like Gundar arose.

@JosephWright Will you join the discussion with barbara?

4:12 PM

@Johannes_B lol – I had to think about this a few seconds before I understood what you just said

@Johannes_B As was mentioned yesterday, we can talk about it but unless the template-providers are engaged we can only get so far

@Johannes_B I hope my CV stuff on my blog is a reasonable way to go

@clemens As long as this view is shared by only some, the whole mission is on the edge.

@JosephWright Did not notice that. Can you give a pointer?

Output written on doctest.pdf (1 page, 4922 bytes).

@JosephWright ^^^^^^^^^^^^^^^^^^^^^^^^^^

@DavidCarlisle Yay

@DavidCarlisle As well as whatsits (kernel level) I think we need to sort Lua allocation for attributes: will do that tonight

@JosephWright So I made various changes in random bits of other peoples code, so I need to run diff and find out what I did then get something usable.

4:16 PM

@DavidCarlisle :-)

@Johannes_B what to do then? Become template providers ourselves and make sure to become the first hit in web searches?

@JosephWright for allocation I did this

function luatexbase.new_attribute (a)
tex.count[258]=tex.count[258]+1
return tex.count[258]
end

@JosephWright It could have some error checking added:-)

@DavidCarlisle Yes, but probably should include the error checking and define the TeX token

@JosephWright as I was writing:-)

@DavidCarlisle Ah, you spotted that part :-)

@DavidCarlisle We are at the advantage we know which count to use :-)

4:18 PM

I thought about that so many times, but ... really, do a bit of reading and you don't rely on templates. Lord love a duck, you don't watch a ten minute Youtube tutorial on how to run a nucular (yes nucular) power plant and go for it.

@JosephWright but I couldn't see how to have a virtual array so that attributes["foo"] returns registernumber("foo") so I just edited the call to do that.

@JosephWright -- what i hope to discuss is not what should ultimately be done, but how to proceed. a couple of talking points: establishing a mailing list for further, detailed, discussion; identifying the current sources of templates, the maintainers of such collections, and whether and how to invite them to an e-mail discussion; perhaps some principles for determining standards and "advertising" them after they're compiled. suggestions welcome.

@JosephWright people should read documentation

@barbarabeeton Sounds like a good plan

@clemens -- well, there are some templates on ctan. maybe that's a place to start looking and evaluating. (if you're willing to critique a few, i wouldn't mind -- in fact would appreciate -- having a few that i'm responsible for looked at.)

4:21 PM

@DavidCarlisle The use case is to force the correct caption numbering (?)

@Johannes_B no you watch the Simpsons.

@1010011010 so I guessed, but the correct thing to do is not have the incorrect numbers, rather than patch them up afterwards:-)

@clemens Hm. \bf is not a faulty command. It doesn't do what the average user expect from a bolding command and isn't defined in the kernel so that you can't rely on its existence with every class, but is is nevertheless a valid choice for a font command like font commands defined with \newfontface from fontspec.

@JosephWright Thnx, I will play around with your suggestion.

@DavidCarlisle That won't be possible. :-)

@barbarabeeton This is quite what clemens said in mychemistry.eu/2015/07/the-template-story

4:28 PM

@UlrikeFischer it is not \bf per se but the supposedly official position of Papeeria (not a private person) regarding templates which I don't like much. Quote »Many templates are quick and dirty hacks, but very often they don’t need to be polished, because they are just fine for their job.«

@UlrikeFischer \bf is really only in article, etc. to support 2.09 documents

@barbarabeeton I'll gladly help if I can

What annoys me is the blind redistribution of code. Vel found some code, it got published to the template site and was copied to the online compilers. But i don't think that any follow-tracks are set up. If vel updates a template, there is no way the online compilers know about it. And even if they did, a template update does not have to be backwards compatible. As we said, a template should to some extend be up to date and keep track of best-practice (whoever judges about that).

2

> a template should, to some extent, be up to date and keep track of best-practice (whoever judges about that).

@Johannes_B I wish I could star that a thousand times over.

@SeanAllred I think clemens mentioned that earlier ;-)

4:40 PM

@Johannes_B @clemens sometimes, you only have to leave the room for five minutes to miss the excitement :)

@SeanAllred true :)

@JosephWright sent doc by mail. Home time, offline for a bit.

@clemens It's an ongoing idea/project -- feel free to contribute :) github.com/vermiculus/texample

I'd like texample to become an actual, lively source of model examples for all manners of TeX stuff (and to advertise arara, of course :))

@SeanAllred so you already startet this:

30 mins ago, by clemens

@Johannes_B what to do then? Become template providers ourselves and make sure to become the first hit in web searches?

@clemens Yep :)

@clemens I feel like it has a lot of potential to be a de-facto resource if everyone chips in

/reason why it's on github

4:48 PM

@clemens Well I understand the frustration but you won't convince people if get too dogmatic and try to forbid commands which are quite valid. And I do confess that I maintain documents and classes which use fancyhdr together with a KOMA class.

@UlrikeFischer I agree that one of the advantages of TeX is the 'if it works, it will work' idea, but templates shouldn't 'just work' -- they should be models of the best way to go about something

since they're templates, they're going to copied over and again. intuitively, if those templates are good, that will benefit the overall quality of code out there. on the other hand, if they're bad...

@UlrikeFischer that's certainly true. Dogmatism never really is helpful

@UlrikeFischer You actually know what you are doing ;-)

Simple plan: I get the team to finish LaTeX3 then we have a proper basis for new templates :-)

5

@JosephWright Sound like a great plan. Everybody star this! ;-)

4:52 PM

I'm only partly joking: github.com/latex3/svn-mirror/wiki/TUG2015-discussion-points

@Johannes_B @UlrikeFischer just a guess but you probably don't openly advertize those templates next to a paid service?

@clemens that's a (*ahem* really) fine point. in the end, sub-par code isn't professional.

@SeanAllred Words i like: examples, showcases, minimalistic, special functionalities. Give them the blocks they need and can choose from to build a personalized house; Instead of providing a full setup house, with leaky plumbing and a lot of building sites that need fixing.

@StefanKottwitz You should add an answer to

29

Q: Who plans to go to the TUG 2015 conference in Germany?

TUG 2015, this years meeting of the TeX Users Group will be in Darmstadt, Germany, from July 20 to 22. Who would like to go there? Does anybody plan to make a presentation? April 10 - bursary application deadline. May 1 - deadline for abstracts for presentation proposals. May 15 - dea...

discussion tug conferences

@SeanAllred There exists something between a bad, only just working template and a perfect template. I don't think that a perfect template is possible, so in general I go for "rather good".

5:00 PM

@UlrikeFischer I don't think I'd put deprecated commands into the 'rather good' bucket, though.

@clemens -- on ctan, there are three templates, pretty minimal, for the three ams document classes, in macros/latex/required/amslatex/amscls/doc/<classname>-template.tex. we actually have more "informative" ones, with recommendations (such as \usepackage{graphicx}) and relevant comments for actual ams authors, but again, pretty minimal. is this a reasonable approach? feel free to make suggestions or criciticize (with explanations, please).

@Johannes_B Exactly. As @UlrikeFischer points out, it's very hard to create a 'perfect' template for a complex task. If you break the task down, though, you can achieve something you can call 'perfectly fine'.

@clemens @Johannes_B: It is not easy currently to use only KOMA tools if the class should also work on older systems. fancyhdr may destroy some KOMA features but it is easier to handle.

On the TUG2015 meeting: turnout is looking excellent

@barbarabeeton I'll have a look at them some time this weekend and then I'll get back to you with any thoughts I might have

5:02 PM

Must remember to take my TeX-sx moderator business cards :-)

@UlrikeFischer this is a very reasonable approach :) I like that

@JosephWright Do you really have something like that?

@Johannes_B because that would be awesome

@Johannes_B Yes

@Johannes_B When the site left beta they sent us all a couple of boxes of them

@UlrikeFischer Agreed, the KOMA dev went insane the last few months.

5:04 PM

@Johannes_B Nowadays you only get the standard StackOverflow design, but at the time they were doing custom versions so I have those

@JosephWright Can you show a photo?

@UlrikeFischer true.

@Johannes_B They are at home so not just at the moment: perhaps @StefanKottwitz can

@Johannes_B They've got the old logo on, before the kerning got fixed :-)

@JosephWright Oh, ok. No worries, i was just interested.

@Johannes_B I'll upload one when I can

@Johannes_B I've still got most of mine

5:06 PM

@JosephWright I just wonder how to give them away. In a bar, talking to a pretty girl? And by the way, i am a moderator in some internet thing. Oh really? A moderator? What do you do? Ok, I am the janitor.

@Johannes_B I've given out a few at various training events, plus I think some when I talked at GUTenberg and probably some to other people in UK-TUG

@JosephWright Probably a good thing, that it says TeX and not LaTeX. Could be confusing ;-)

@JosephWright -- i've just edited that question to add a link to the particpants list. (but @StefanKottwitz's name isn't on the list yet. if it's not there tomorrow, i'll nag.)

@egreg Cwac cwacio cwac @PauloCereda - Cwac!!

5:22 PM

@Johannes_B A bit harsh. But I doubt e.g. that they are many people (beside Elke) who understand scrlayer-scrpage (scrpage-scrlayer?) .

@UlrikeFischer The other stuff as well, many new things and updates. If dev is going soo fast, who should keep a template up to date? And even if, that would mean that the user has to update over and over and over again. So, if there would be standards, one would need to set a time span for what counts as recent or up-to-date.

@UlrikeFischer Some of the solutions for KOMA on site are deprecated due to the new stuff.

@UlrikeFischer I do quite like the idea behind it. Especially how you can clone and combine layers and pagestyles for different purposes…

Stefan Kottwitz

@Johannes_B Also I'm still at work where I don't have the TeX.SX cards.

@Johannes_B it can't be the goal to keep them up to date with bleeding edge development. Reasonably new and stable and (besides bug fixes of course) maybe once a year an update would be more than good enough!

Stefan Kottwitz

@JosephWright I added an answer.

5:33 PM

@clemens I like the idea behind too. But I don't understand how to use them in a sensible way. Say you have a standard page style with some elements. And you sometimes needs this pagestyle minus one element (say without a logo in the left foot). How do you setup the system with layers and clones?

@StefanKottwitz Ah, but when I say at home I don't mean my flat here, I mean at my mum's house :-)

@UlrikeFischer first idea: the logo should be on a layer of its own so that the logo-layer can be added to to a pagestyle (or to a clone of it) and removed if necessary. If this concerns more elements I suppose one layer per element is needed

@clemens Sure that the easy part, the theory ;-) Now show the practice: How do define, use and disable the layers.

Martin Scharrer

5:58 PM

Anyone tried to use Virtual Box? I'm looking for a prebuilt Linux + LaTeX image. Too lazy to do it myself

@UlrikeFischer I'll try:

\documentclass{scrartcl}
\usepackage[automark]{scrlayer-scrpage}

\DeclareNewLayer[
  background,
  align=br,
  area={\dimexpr\paperwidth\relax}{2.5cm}{2.5cm}{2.5cm},
  contents={\rule{2cm}{2cm}}
]{logo}

\usepackage{lipsum}

\AddLayersToPageStyle{empty}{logo}

\pagestyle{empty}

\DeclareNewPageStyleByLayers{scrheadings+logo}{}
\ForEachLayerOfPageStyle{scrheadings}{%
  \AddLayersToPageStyle{scrheadings+logo}{#1}%
}
\AddLayersToPageStyle{scrheadings+logo}{logo}

\begin{document}

\lipsum[1-5]

\clearpage

1 hour later…

7:14 PM

@UlrikeFischer @clemens Rather complicated (or easy?). As it seems, that is still beta. komascript.de/node/1929

7:37 PM

@DavidCarlisle I see the etex update has hit CTAN

7:51 PM

@JosephWright @WillRobertson Out of curiosity, what exactly doesn't work when you have nested props?

\input expl3-generic \relax \ExplSyntaxOn

\prop_put:Nnn \l_tmpa_prop {key}{value}
\prop_show:N \l_tmpa_prop
\tl_show:N \l_tmpa_prop
\tl_set_eq:NN \l_tmpa_tl \l_tmpa_prop
\tl_show:N \l_tmpa_tl

\prop_put:NnV \l_tmpb_prop {key} \l_tmpa_tl
\tl_show:N \l_tmpb_prop

\exp_args:Nf \tl_show:n { \prop_item:Nn \l_tmpb_prop {key} }

\bye % works as expected

@Johannes_B that's how I interpret the manual, anyway: »Die Entwicklung dieses Pakets ist noch nicht abgeschlossen. Teile des Pakets sind auch noch als experimentell einzustufen. Daher können sich in Zukunft insbesondere die internen Funktionalitäten und Funktionsweisen noch ändern.«

@clemens Nice try, but absolute values like 2.5cm are not ok. The logo must be placed relative to the standard header. Oh and yes and the customer just decided that left to the logo there should be the firm name, but if there is no logo then the name should wander to the left ... and he will probably need also a layout in A3 landscape ...

@SeanAllred Interface design: you are relying on the fact that a prop is a macro and so can be stored in one go

@JosephWright I'm not sure I follow?

(I get the idea of black-box design, but I'm not sure where I'm violating it)

@SeanAllred You are relying on the implementation

7:56 PM

@UlrikeFischer I don't think that relative positioning of elements on different layers is possible. In my eyes (but I have to read scrguide closely again to be sure) elements of different layers shouldn't be aware of each other

@JosephWright Frankly I looked at the implementation and I have no idea why it's so complicated -- I don't understand it.

@SeanAllred In the past a prop was stored as a toks, which would have failed here. We've talked about moving to a set up where we use one macro per key, etc.

I'm not sure how I could be relying on something I don't understand

@SeanAllred What's complicated?

@JosephWright yes and i've contacted (or at least sent mail to) most package authors that were loading etex suggesting conditional loading, several have already replied saying they'll do it at next update

7:57 PM

\__prop_pair:wn, I think it's called

@SeanAllred You are assuming that \l_tmpa_prop contains all of the data of the property list rather than being a pointer

@JosephWright Ahhhh

@SeanAllred We've been through several implementations to get the best performance we can

@JosephWright I suppose I've never considered any alternative implementations

@SeanAllred That's the point: we don't want people relying on internal detail, only on what we document

7:58 PM

@JosephWright Having the data be all together in one macro doesn't seem to have any downsides (from what I understand of TeX's internals)

@SeanAllred Try a prop with say 100000 items and see how it slows down: we are worried about that

@JosephWright Good lord, why?!

@SeanAllred Adding items one at a time gets slower as the list gets longer

@SeanAllred Think something like pgfplots reading large data files

@SeanAllred I can happily have 10k data points in a spectrum

« first day (1712 days earlier) ← previous day next day → last day (3222 days later) »

Transcript for

Jul '159

$TeX - LaTeX$ TeX, LaTeX and Friends

\begin{chat}

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

$TeX - LaTeX$

site design / logo © 2024 Stack Exchange Inc; legal

mobile