« first day (5091 days earlier)      last day (30 days later) » 

cfr
cfr
1:36 AM
@DavidCarlisle possibly? (if you are trying to make me squirm, you are succeeding.)
 
4 hours later…
5:14 AM
@DavidCarlisle Ah, right - I missed that when removing a conditional testing for e-TeX - I'll fix
 
1 hour later…
6:29 AM
@barbarabeeton I have to admit, in my edits I never use it or the full multline
6:45 AM
@DavidCarlisle, @cfr Fix for the nesting-if business is with CTAN
@cfr Why've you deleted your answer about expansion?
@JosephWright hopefully I'll steal the tick :P
7:41 AM
Sign, my snap-firefox just died so bad that I no longer have my bookmarks or saved passwords. Nice start of the day. And snap-firefox refuses to even start.
7:53 AM
@daleif seems like it snapped.
8:13 AM
@Skillmon It more than snapped, it cannot find $HOME, so cannot start. And if you reinstall it, you just killed your setup as that is storred inside the snap dir, sign
8:26 AM
@cfr @JosephWright (anyone) could people say if kpsewhich afoot.sty returns arabtex/afoot.sty or ledmac/afoot.sty (I get arabtex)
arabtex
@DavidCarlisle I get arabtex. Are the files different?
@UlrikeFischer yes (blame @cfr)
@DavidCarlisle that's atypical, normally we blame @UlrikeFischer
@DavidCarlisle what did she do that you found that?
@Skillmon I'm a generous person, other can have some of my blame if they need it.
8:38 AM
@UlrikeFischer I inherited projects that were orphaned but are now less orphaned than they were :( github.com/bidi-tex/ledmac/issues/1
@DavidCarlisle I lost track of the bibi-repos, is that an active account for ledmac? (and why is ledmac in bidi??)
@UlrikeFischer super secret skype message
@JosephWright how about \timespan (I didn't check TeX Live though)? Also, I'd suggest supporting hh(:mm(:ss)?)? as valid input instead of ; (feels more natural for time expressions).
4
8:54 AM
I suggest: \time_in_oven so that David can easily typeset his duck recipes :P
@samcarter That should be the L3 name of the function: \time_in_oven:nnn
@Skillmon :D
9:10 AM
@samcarter Gert was disappointed that tardis wasn't considered - such a nice latin name.
@UlrikeFischer you should smuggle \tardis in as an alias via first aid :)
@samcarter If it will ever be considered again, it could be \retardis.
@DavidCarlisle Don't tease the duck!
@mickep It wouldn't need to be considered again, the tardis could go back to when it was first considered
@samcarter Oh, that is indeed true, but kind of cheating.
9:29 AM
@Skillmon Ooh, that's quite good
 
2 hours later…
11:14 AM
user image
6
12:01 PM
@Skillmon congrats
@daleif thanks. Some misanthropes might complain about comma placements
:)
12:32 PM
@Skillmon I'm in the territory at the moment, misses 55555 but is 16 away from 55755
@samcarter ooh secret ping :)
@PauloCereda oh secrets!
crap too many points ;-)
1:12 PM
@Skillmon Thanks. It took me a while to get back to this. But I must be doing something wrong. Here's a snippet:
\documentclass{article}
\ExplSyntaxOn
\tl_new:N \l_alan_cols_tl
\newlength{\mylen}

\NewDocumentCommand {\makecol}{m} {
\tl_set:Nn \l_tmpa_tl {\the #1}
\tl_set:Ne \l_tmpa_tl { { \exp_not:V \l_tmpa_tl } }
\tl_put_left:Nn \l_tmpa_tl {p}
\tl_put_right:Nn \l_alan_cols_tl {\l_tmpa_tl}
\l_alan_cols_tl
}

\begin{document}
\setlength{\mylen}{30pt}
\makecol{\mylen}
\end{document}
1:36 PM
@AlanMunn I'm not sure what you expect to get, but shouldn't that be \tl_put_right:No \l_alan_cols_tl {\l_tmpa_tl}?
1:58 PM
@UlrikeFischer I'm trying to build a tokenlist that contains {...}. So I would like the final object to contain p{30.0pt}
@UlrikeFischer The idea would be that this would be passed to \newcolumntype and then used as a table specification.
@Skillmon the comma is misplaced
@DavidCarlisle I knew it! <3
@Skillmon I advise having 6 digits rather than 5, then correct comma placement is easier.
@DavidCarlisle that's very helpful, thanks!
@AlanMunn \def\makecol#1{\edef\l_alan_cols_tl {p{\the\dimexpr#1\relax}}
2:10 PM
@AlanMunn You're not expanding \the #1, so you'll only get \the\mylen in your MWE. Also you can have that much easier: \tl_set:Ne \l_tmpa_tl { p { \the #1 } } \tl_put_right:NV \l_alan_cols_tl \l_tmpa_tl
@AlanMunn but to get a more stable and perhaps prettier interface I'd do \tl_set:Ne \l_tmpa_tl { p { \dim_eval:n {#1} } } on the first step.
@Skillmon slightly more idiomatic L3 than mine:-)
@DavidCarlisle slightly, yes :)
@DavidCarlisle also yours misses \protected
@Skillmon the brave don't need protection
@AlanMunn you can even combine the two steps: \tl_put_right:Ne \l_alan_cols_tl { p { \dim_eval:n {#1} } }.
@DavidCarlisle that's why I got four children.
@Skillmon @DavidCarlisle @UlrikeFischer Thanks all. Off to teach now. I'll absorb all this wisdom later. :)
cfr
cfr
2:27 PM
@JosephWright thanks!
@JosephWright somebody sensible answered, so I deleted mine ...
@DavidCarlisle arabtex/afoot.sty
@DavidCarlisle not sorry.
@UlrikeFischer it's the one listed on ctan.
2:48 PM
@cfr you have mail
3:10 PM
@cfr yes I know but ctan could be confused too (or not up-to-date) ;-)
@cfr thanks I'll see if i can get the ledmac one renamed
3:34 PM
Was \count0 a scratch counter? Seems to have found another problem with it in tikz abs pos vs memoir. Will investigate
@daleif well it can be, but it is also the page counter so has to be used with care (or better, not used)
@daleif basically you need to make sure you never ship out a page while using it or its value will get looged as [42] or whatever in the log file
3:53 PM
If it's all the page counter then it might explain why it is affected by the value 2.
cfr
cfr
@UlrikeFischer I more meant: I didn't actually check or anything.
@daleif \setcounter{page}{99} sets \count0
cfr
cfr
@DavidCarlisle ok. will try to find it later ...
@cfr it's not very exciting:-)
4:19 PM
@Skillmon Ok I'm still confused:
\documentclass{article}
\usepackage{array}
\ExplSyntaxOn
\tl_new:N \l_alan_cols_tl
\newlength{\mylen}

\NewDocumentCommand {\makecol}{m} {
\tl_put_right:Ne \l_alan_cols_tl { p { \dim_eval:n {#1} } }
\let\mycols\l_alan_cols_tl
}
\ExplSyntaxOff
\begin{document}
\setlength{\mylen}{30pt}
\makecol{\mylen}

\makecol{\mylen}
\mycols
\newcolumntype{A}{\mycols}
\begin{tabular}{A}
Foo & Bar
\end{tabular}

\end{document}
@AlanMunn \expanded{\noexpand\newcolumntype{A}{\mycols}}
@AlanMunn or dispense with \mycols and just directly use \newcolumntype in your constructor.
@DavidCarlisle Ok that works. I thought \newcolumntype was like \newcommand, but it just overwrites any previous specification.
@AlanMunn well it is, but tabular (once array package is loaded) doesn't expand its argument so you can not do \def\foo{p{3cm}}... \begin{tabular}{\foo} that works in the format version but not once you load array as by design it does not expand the argument, so you can put things in >{...] without them expanding too early
4:34 PM
@DavidCarlisle So if this code needed to be general it would be better to load array and use the \newcolumntype version, because if array is loaded, I couldn't do the simpler version at all?
@AlanMunn I'd be far too polite to mention it but @UlrikeFischer would point out that the array package documentation explicitly mentions this expansion issue on page 5
@AlanMunn you can define a macro but then at the end if you are using newcolumntype (in which case array must be loaded) you need to expand its argument
@AlanMunn you are using put_right so are you planning to build up several columns at the same time? that's easier if you use the intermediate macro, as you show here
@DavidCarlisle Yes, the idea is the iteratively construct columns from a sequence.
@AlanMunn I would always use \newcolumntype. Note that you are not restricted to letters, you can use commands \newcolumntype{\alancols}{...}
@AlanMunn yes so there is no documented equivalent for put_right for \newcolumntype so you'd have to add to the underlying macro
@UlrikeFischer Oh, I hadn't realized that either. But now I think I have all the pieces I need.
@DavidCarlisle @UlrikeFischer Thanks again.
5:21 PM
@samcarter help <3
 
3 hours later…
8:46 PM
@AlanMunn ooh no!
@AlanMunn it just needs to be a single N-type token, no other restrictions are implied by \newcolumntype.

« first day (5091 days earlier)      last day (30 days later) »