@egreg: there's even an easter egg! :)

@AndreyVihrov Oh my! :-|

@JosephWright: In the file l3styleguide is a small typo at line 180 and 182.

\begin{verbatim}
\tl_set:No \l_some_tl \l_some_other_tl
\end{verbatim}

This gobbles the first backslash

If you use the star version it will work.

@MarcoDaniel Sorted

@JosephWright or you indent ;-)

@MarcoDaniel Well yes - one day I'll sort out native verbatim support, and we will have something akin to listings setting gobble = <number> to allow indentation without messing things up.

@JosephWright I know my comments (yesterday and this one) aren't really helpful. Maybe anytime I can help more efficient

@MarcoDaniel Oh, they are very useful. We don't have the opportunity to check over all of the text (it's in a constant state of flux), so it's very helpful indeed if others do.

I provided an answer to improve my LaTeX3 skills. So I am opening the discuss

@MarcoDaniel Not very useful, but the 10pt option is the default.

@AndreyVihrov Please explain your statement "not very useful". This comment doesn't show me any mistakes. What do you mean with 10pt. I didn't use this dimension.

@MarcoDaniel "The comment I'm going to make is not very useful to you, but the 10pt document class option is already default in the article class".

@AndreyVihrov 10pt is a copy paste error ;-) -- I will fix this. Every comment improving this is useful. First of all I wanted to use a sequence, but clist is more comfortable.

I'm coming back here and never leaving. I just answered a question on english.sx and the question was closed by a mod with an irrelevant question as the proposed duplicate. When I commented on this, he changed it to another duplicate (which it wasn't) at which point he edited the question so that now it's a duplicate. And he doesn't seem to be responding to the comments. Wow.

@AlanMunn Welcome home, Alan. :)

Hic sunt dracones out there. :)

@PauloCereda Absolutely.

@AlanMunn Sadly those communities have minds of their own. E.g, for them, upvote means "upvote the one you like and downvote the others you don't". :(

@AlanMunn Now I can see the difference between the three questions, though it was not obvious to me immediately.

@AndreyVihrov Well especially now that the mod has edited the new question to get rid of the evidence, so to speak.

The lessons to be learned from this little skirmish are: (i) our mods are great (ii) our policy of having mods not close questions as duplicates until there's been enough voting (and comments) by others is definitely the way to go.

@AlanMunn We aim to please

@AlanMunn Moreover, linking to a question with a hen in it is obviously abusive. :)

@JosephWright It's true. We don't know how lucky we are over here. I think I'm done with that site altogether. I should have kept away in the first place.

@egreg Actually now the OP has jumped on and rolled back the edit (agreeing with me). I'm going to have a little chat and then basta.

For how long do we see [migrated] questions still linked in meta? :)

@PauloCereda It will always stay listed as [migrated] when viewing meta questions. However, if you view the top/active questions, it will remain there until other active questions push it out of sight.

Top (active) questions I think default to 50.

So, we would need 47 new active ones.

And "new" here refers to brand spanking new, or old but revised (edited or answered).

When does area51.stackexchange.com/proposals/16617/academia go public?

@AlanMunn @JosephWright I second that; we probably have the best moderators of all the stackexchange sites. I personally prefer a little bit of duplication, as a means to elicit new answers from persons that recently joined like Frank or David to provide a different angle to a problem or for someone like Herbert to add his 4 liner. It also gives a chance for LaTeX3 code to be added to questions. It is also good from a pedagogical point of view as it provides some form of spaced repetition.

That xkcd cartoon has now ruined my life. I couldn't concentrate in church this morning because of the awful kerning in the hymns. The worst had a proportional font but the apostrophe had the same spacing as a regular letter. Aaaagh!

Oh, and Tom Bombadil has gone critical. There are 15 new "answers" on the "Favourite TikZ/PGF" meta question.

Does anyone have (on hand) a quick reference on how to make a .dtx, .ins and .pdf bundle for CTAN? At the moment, I have a working .tex example file with all the definitions. I guess my main pondering is how to get something in the .dtx format...

@DavidCarlisle: Beat me to it by one second, with a better answer. >:-{

@Werner You can use sty2dtx to get your dtx-file

@MarcoDaniel So I would include the documentation in the .sty file then? I guess I could read the sty2dtx documentation, right?

@Werner Sorry wrong answer. Do the following: take your sty-file and run:

sty2dtx -B mypackage.sty

After this you can copy your tex-file into the dtx-file.

@Werner Not a quick reference, but several nice posts by Joseph: texdev.net/tag/dtx one of them, texdev.net/2009/10/11/working-with-dtx-files , deals with releasing to CTAN

The related part is detectable ;-)

It's important that every related text of your documentation (tex-file) is commented.

% \maketitle
%
% \section{Introduction}
%
% Put text here.
%
% \section{Usage}
%
% Put text here.

@StefanKottwitz Thanks, I'll check into this...

Finally I recommend ctanify tex.stackexchange.com/questions/25116/…

@MarcoDaniel So you're saying the .tex file (the documentation) should be entirely commented, and after that you put your sty2dtx output...?

Don't worry if the answer to too long-winded. I just haven't read any documentation of anything yet, and will probably find what I'm looking for when I start reading.

@AndrewStacey My song sheets are typeset in LaTeX. :)

@Werner Ah. :)

Note that using sty2dtx for anything but a one-time conversion is useless, because the .dtx format is not mandatory on CTAN.

@AndreyVihrov I understand. Hence the fact that there's .sty files available for download in some packages.

I just think that the bundling using .dtx is neat.

@Werner If .dtx is your original format, then it is neat; if you convert your original format to .dtx just to post it on CTAN, then it is useless — you gain nothing.

@Werner Also, I think it's really only good as a tool for code documentation, not source documentation.

@AndreyVihrov Interesting to note.

@AlanMunn Why is this?

@Werner Sorry what I meant (if it wasn't clear) was not good for user documentation. Mainly because if your package/class has a reasonably comprehensive manual, you end up with a big chunk of regular latex that is a pain to edit and update.

@Werner Here's an implementation of the \names macros, where one defines a list macro with \definearray{letters}{X,Y,Z} and then can call \letters{} to get the number of elements or \letters{0} to \letters{2} to access the items:

\documentclass{article}
\usepackage{xparse}
\ExplSyntaxOn
\NewDocumentCommand{\definearray}{ m m }
  {
   \clist_new:c { g_vizipok_#1_clist }
   \clist_gset:cn { g_vizipok_#1_clist } { #2 }
   \cs_new:cpn { #1 } ##1
     {
      \tl_if_blank:nTF {##1}
        { \int_eval:n { \clist_length:c { g_vizipok_#1_clist } } }
        { \clist_item:cn { g_vizipok_#1_clist } { ##1 } }
     }
  }
\ExplSyntaxOff
\definearray{names}{Katie,Frank,Laura,Joe}
\begin{document}
\names{}\par\names{2}
\end{document}

@AlanMunn Do you think? After creating my own cls-file it works well and it's easier to maintain

@Werner Have you run the tool sty2dtx? After this open the dtx file and I am sure you know what I mean.

@egreg Thanks. Since the question seems to have been successfully answered using pgfmath (by @cjorssen), I'll add this to my answer with a reference to this chat.

@MarcoDaniel I guess it's a bit of personal choice, and for me it seems more effort than is worth it. I think @JosephWright is also on record as not being a big fan of it for user docs either (not surprisingly, given the size of e.g. the siunitx documentation.)

@MarcoDaniel No, I haven't. Once I do, I'm sure I'll see what it's all about.

@AlanMunn back when this was new we had emacs support for doc such that it was as easy to type doc format as direct code (you could make the % \begin{macrocode} nonsense come or go at will. Not sure what happened to that...

@Werner: To sum up, .dtx is good only if you actively work on it. That is, if your code was in the .dtx format from the beginning, or if you decided at some point to convert your code to the format (using sty2dtx), then went through and carefully annotated everything.

@Werner: If you have regular .tex code with comments, then automated conversions will only make it worse; it's always better to publish in the format you yourself used to develop the code. A really bad example is polyglossia: typeset its .dtx without \OnlyDescription to see what I mean.

@Werner It has the notable advantage of abstracting the definition of arrays. This is actually a clist in disguise, with the addition of accessing the length (a bit like Perl does).

@AndreyVihrov You're such a polyglossia h8tr :-)

@AndreyVihrov Indeed

@AndreyVihrov This makes sense. It's just that I see many new packages following this format, so I'm wondering whether I should follow that as well. Doesn't seem like I'm much of a leader. :-|

@DavidCarlisle It probably still exists. I haven't used emacs except occasionally in years, though.

@egreg Yes, the abstraction rulz.

@Werner I think also that as the distribution/installation of packages has improved at the distribution level (TeX Live and MikTeX) there's much less need for self-contained installation packaging.

@egreg A small macro like \addarray would be nice ;-)

@AlanMunn Good point. In fact, great point!

@MarcoDaniel Reaching for the stars, are we? :)

@Werner That fact combined with the fact that we have really nice version control systems that didn't exist when .dtx was developed, clinches it for me.

@AlanMunn Another thing I'm unfamiliar with: "version control systems".

@MarcoDaniel It's easy to add it.

It's clear I haven't written any packages. :(

@Werner Then get to this right now! hginit.com

@egreg I know. It's only a nice gimmick and the questioner will be happy ;-)

:3480881
     \NewDocumentCommand{\addarray}{m m}
      {
       \clist_gput_right:cn { g_vizipok_#1_clist } { #2 }
      }

Of course one should add some check to see if the array is defined.

\NewDocumentCommand \addarray { m m }
      {
        \cs_if_exist:cTF{  g_vizipok_#1_clist }
          { \clist_gput_right:cn { g_vizipok_#1_clist } {#2} }
          { \definearray { #1 } { #2 } }
      }

@egreg Something like this.

I recommend \clist_item:cn { g_vizipok_#1_clist } { ##1 - 1 } this definition.

In this way the ouput of \clist_length:c can directly use for the last argument.

@MarcoDaniel Perl arrays (and LaTeX3 clists) number from 0.

@MarcoDaniel Do you plan to go to the DANTE meeting?

Ok remind me again, how do you do links in chat? [text](url) doesn't seem to work.

@AlanMunn just paste into the chat, preferably on a line on its own

@StefanKottwitz With that syntax? Or just the bare url? Is it the "line of its own" that matters? If so, how did Andrey put the link that he did in the text?

@AlanMunn Does the url have the http:// prefix?

@PauloCereda Yes.

@AlanMunn Oh.

Testing Google with prefix.

@Werner In addition to what Andrey suggested, I'm partial to github.com

@PauloCereda Ok I get it. It's just the bare url. What you can't do is specify the link text, I guess.

Meta badges are sometimes very amusing: I've just earned the "enlightened" badge for arguing that we retain the package-writing and class-writing tags as separate.

@AlanMunn just the bare url - for known sites a preview may be shown

@StefanKottwitz Right. I finally figured that out. Thanks. So you can't specify a separate link text in chat, I guess.

Friends, can any of you help me with Some problem I am facing about TeXmaker and its Bibliography menu?

(sad that none helps)

I used the Bibliography menu and clicked on Journal article.

I got a layout to fill.

I filled in the entries.

But, on compiling, it shows up that there is an undefined reference and also those data I put in show up as-is -where-is. They have not been processed into nice form.

@AlanMunn Would you like to look at the .pdf file?

I'm not completely familiar with TeXMaker, but the Bibliography menu allows you to enter .bib entries. So you first need to save the entry as <yourbibname>.bib. Then in your document file you need to have \cite commands to the entries in the .bib file, and have a \bibliographystyle{<style>} command and a \bibliography{yourbibname}` command. Is this what you did?

You never compile the .bib entries themselves. The .bib entries are simply a database that is referred to by the citation (\cite commands.). (Ideally you probably want to use the natbib package for this.)

I put in the bib entry directly into TeX file, I think @AlanMunn

The following might be helpful: en.wikibooks.org/wiki/LaTeX/Bibliography_Management

@KannappanSampath No, that's not how it works. I have to go now (sorry) but it looks like you need to read a bit on how this is supposed to work. Check out the link I just posted. If you're still stuck, post a minimal working document as a question here on the site (not in chat), and you'll probably get a response. But do read the link I posted first and see if that helps you.

Thank you for your time. I will oblige. :)

I think I have learnt where the mistake lies.

Thank you all. :)