« first day (2402 days earlier)      last day (2524 days later) » 

5:51 AM
Joseph Wright has added an event to this room's schedule.
 
6:35 AM
@Moriambar Ciao! All together to Answer the Unanswered party! By chance, it starts right after the end of the Champions League finals, otherwise, someone wouldn't have taken part to it :)
 
@CarLaTeX Well I registered just because I don't know exactly but it seems like fun. About champions league I don't know
 
@JosephWright There's a little typo in the event title (Unaswered instead of Unanswered)
 
Joseph Wright has removed an event from this room's schedule.
Joseph Wright has added an event to this room's schedule.
 
@Moriambar For what the event is, look here: tex.meta.stackexchange.com/questions/7165/…. For the Champions League, I know about it because I work 8 hours a day with two Juve fans!
 
@CarLaTeX I guess then tha Juve has reached the final then
@CarLaTeX the event sounds interesting
 
6:42 AM
@Moriambar Of course :) Register yourself again to the event, because Joseph has renewed the title (Thank you @JosephWright)
 
@CarLaTeX I already did
 
@Moriambar Bravo. I don't know how much will be my contribute, I could offer myself as upvoter :)
 
@CarLaTeX we'll see. I hope I'll be around for enough time (it could be all day or a couple of hours).
 
@Moriambar In the schedule it's written "lasts 120 minutes."
 
@CarLaTeX that can be good then
 
7:21 AM
@CarLaTeX Well, at that time I cannot participate. It's too late for me being productive.
 
7:58 AM
@TeXnician I think you can participate for half an hour only. Moreover, you can answer the unanswered even if there's no event :)
 
8:13 AM
@JosephWright busy this morning? :-)
 
@DavidCarlisle Not at the mo'
 
@JosephWright just catching up my gh mail:-)
 
@DavidCarlisle :)
@DavidCarlisle Will do xetex.def shortly
 
mostly looks OK but I think you picked up an old graphics.dtx?
$ diff ../l3svn/latex2e-public/trunk/required/graphics/graphics.dtx .
22c22
<           [2017/04/14 v1.1b  Standard LaTeX Graphics (DPC,SPQR)]
---
>           [2017/01/01 v1.1a  Standard LaTeX Graphics (DPC,SPQR)]
 
@DavidCarlisle Oh drat: I'll sort that
 
8:20 AM
@JosephWright thanks, I was worried initially that I'd never checked that in, but seems like I did:-)
 
@DavidCarlisle Hopefully sorted now
@DavidCarlisle One thing I'm still not 100% about is whether to use XForms to cache images when using (x)dvipdfmx. The pdfTeX driver does this, and it is workable for dvipdfmx, but would mean altering all of the scaling code ...
 
@JosephWright thanks
@JosephWright we went through hoops deciding on whether to do that for pdftex (I think the underlying code wasn't that stable originally so things kept changing) It's been a while since I looked at that but I assume it can make big savings in the relatively rare case that the same image is used several times in the same document (page backgrounds and headers etc)
 
@DavidCarlisle Yes: I did test it (dvipdfmx) and it works as advertised: one copy of the image. What's important if you do that is that the image is 'the same' each time. For dvipdfmx that would mean altering the clipping/scaling/... to use the pdftex.def approach. Probably zero-risk but then the pay-off is low (not that many docs use repeated images ...)
 
@JosephWright yes I think the motivating example at the time was a corporate logo in a page head on every page, but I don't see that much in tex documents and they tend to be small anyway, so I think the feature sounds more useful than it is (and it makes the code harder to read)
 
8:36 AM
@DavidCarlisle Leaning that way myself
@DavidCarlisle I suspect post-processing is likely to pick this up anyway
@DavidCarlisle For pdftex.def it's long-standing so best kept I
@DavidCarlisle What about bbox data? That's cached in current dvipdfmx.def (to avoid multiple calls to extractbb) as well as more-or-less by pdftex.def (as the image itself is cached).
@DavidCarlisle Perhaps I'll back-out in the L3 experiments on that basis
@DavidCarlisle What's notable is that this has zero impact on dvips or dvipdfmx or XeTeX, so suggests it's not generally vital: I guess I favour one set of functions which apply everywhere ...
 
8:57 AM
@JosephWright reducing system calls seems like a good thing to do in general.
 
@DavidCarlisle OK, will stick as we are
 
9:20 AM
1
Q: Problem with latest version of babel

José Carlos SantosI have a problem with that latest version of babel (3.10). If I compile this file: \documentclass{article} \usepackage[ukenglish]{babel} \pagestyle{empty} \begin{document} Hello world! \end{document} Then I get the message LaTeX Error: Command \setlocale already defined. Why is that?

This is going to be a major issue with MikTeX, I suspect (@UlrikeFischer)
 
9:31 AM
@JosephWright @UlrikeFischer VVVVV
I tried copying bblopts.cfg in the working directory, but I don't get the error (on TeX Live). Same file list. — egreg 2 mins ago
 
@egreg OK, so perhaps a modified ukenglish.ldf?
 
9:49 AM
@DavidCarlisle Hopefully most of the suggestions for drivers are now done!
 
@JosephWright I should build.lua install, and see if anything breaks when answering questions from real people:-)
 
@DavidCarlisle Well I was thinking that this will need some testing: blog post, perhaps
@DavidCarlisle We'll of course find out at least one person is using one of the really weird options ...
 
@JosephWright more people would test it if we pushed out a release as soon as tl2017 is out:-)
 
@DavidCarlisle More-or-less the plan
@DavidCarlisle I'm hoping for a review buy the team (i.e. you!) then to send to CTAN and expect to update quickly as needed
 
@JosephWright oh build.lua install doesn't unpack the new graphics:-) (hmm might be easier to just do that by hand if I want those to be installed into latex/graphics from here)
 
10:02 AM
@JosephWright @egreg @DavidCarlisle looking at it now, but I have to check for update first. Currently I don't see the issue.
 
@JosephWright
(/home/davidc/texmf/tex/latex/graphics-def/pdftex.def
! Illegal parameter number in definition of \Gread@pdftex.
<to be read again>
                   2
l.154         \PackageError{pdftex.def}{File `#2
                                                ' not found}%
?
 
10:15 AM
@JosephWright pushed a change for that back, hopefully I'm hooked up to the right branch..
 
10:29 AM
@JosephWright @egreg looks like a simple "didn't update the format" problem.
 
10:53 AM
@JosephWright I think graphics needs gin@log something like
+% \begin{macro}{\Gin@log}
+%   \changes{v1.2a}{2017/06/01} {New macro} In the standard interface
+%   this prints to the log but can be changed via keys in \textsf{graphicx}.
+%    \begin{macrocode}
+\let\Gin@log\wlog
+%    \end{macrocode}
+% \end{macro}
+%
 
@UlrikeFischer Maybe the well-known user/admin issue?
 
@JosephWright and switch first second here:
\def\Gread@pdftex#1{%
  \expandafter\ifx\expandafter\relax\pdffilesize{#1}\relax
    \expandafter\@firstoftwo
  \else
    \expandafter\@secondoftwo
  \fi
 
@egreg Yes I guess so.
 
Hi. Anyone with Apache experience? I would like to enable file download (e.g. rar, zip) but also disabling directory listing at the same time. I couldn't do that without enabling directory listing enabled. Any suggestion?
I'm on Win w. Apache 2.2
 
@cempro Oh I am afraid you poked the wrong audience here. :) We are the TeX blokes. In any case, take a look at your httpd.conf, the file is usually well commented and it might be a matter of tweaking things to your own desire. :)
 
11:03 AM
@cempro this is an odd place to ask:-) but there shouldn't be any relation between directory listing and files, so log as you use the full url to the hidden zip file. Are you sure it's not blocked for other reasons (eg explicitly blocking zip is not uncommon)
 
100
Q: How do I disable directory browsing?

arthurprsI want to disable directory browsing of /galerias folder and all subdirectories Index of /galerias/409 * Parent Directory * i1269372986681.jpg * i1269372986682.jpg * i1269372988680.jpg

 
Thnx. I was able to disable directory listing from the httpd.conf. But not for file access. But let me check it again.
 
@PauloCereda odd that stackoverflow has and tags but we don't in tex.sx. Even stranger that we don't have tags for and
 
@DavidCarlisle you are an expert in everything. :)
@DavidCarlisle oh no
 
@PauloCereda Especialmente línguas
 
11:06 AM
@DavidCarlisle we used to have other important tags: tex.stackexchange.com/revisions/153389/3
@DavidCarlisle ooh
 
:)
just one last question although this is not the best place
 
And it's quite documented, too!
@DavidCarlisle ^^ and you are the top answerer!
:D
 
if I enable Indexes option under the directory tag (httpd.conf), everything works. Should I leave indexes option there and create a seperate .htconf for the subdirectory?
*.htaccess not conf :)
 
@cempro Think so, .htaccess it's the best way to set up specific, custom settings. :)
 
@PauloCereda i should fix that tag wiki page. No mention of interesting recipies
 
11:11 AM
@DavidCarlisle you are mean
 
@PauloCereda except when they are disabled
 
@DavidCarlisle ooh good point
@DavidCarlisle in fact, I am not sure how it works in Windows.
 
11:26 AM
Config. was all right. Problem was related with the directory (ownership)
Thank you
Have a nice day ;)
 
@cempro you too! Always have a towel!
 
@PauloCereda Yeah! Why not on the main site, too? 😆
 
@CarLaTeX Your question is suitable for that tag, go for it!
 
@PauloCereda Indeed! I want the duck tag!
 
[tag:🦆] hmpf do you have to use
 
11:36 AM
@DavidCarlisle oh no
 
@PauloCereda "non ascii tags enabled on a per-site basis"....
4
A: Tag search and info won't work with non-ASCII tags

Brian NickelThis is fixed with the latest API build. (No app update required.) Non-ASCII tags are allowed on a per-site basis. When the app went to sanitize the tag for search, it checked against the API site, which doesn't have non-ASCII tags so it removed every non-ASCII character and in the first case e...

 
@DavidCarlisle ooh
 
11:51 AM
\documentclass{article}

\usepackage{graphics}

\begin{document}

\includegraphics{example-image}

\end{document}
@JosephWright I started to debug a bit but I suspect I'm out of sync somewhere for example I get ! Undefined control sequence. <argument> \Grot@sin for the above, could trace that back but no point if I have the wrong files anyway:-) Does the above work for you?
 
@DavidCarlisle I get past there but something else is wrong ;) Give me a couple of minutes
@DavidCarlisle Ah, got it now: working on the issues
@DavidCarlisle Didn't think I'd tested anything did you ;)
 
12:34 PM
@DavidCarlisle Should be fixed now
@DavidCarlisle One thing that we might also do is swap .png and .pdf: I can't be the only person who things the vector format should be checked first ...
 
12:53 PM
@JosephWright sounds reasonable
@JosephWright oh Ok you've made it install graphics.sty in graphics-def Ok for testing I guess, I just need to remove /home/davidc/texmf/tex/latex/graphics/graphics.sty as that's masking the one in /home/davidc/texmf/tex/latex/graphics-def/graphics.sty
 
@DavidCarlisle Yes, just for testing: remember I plan to 'throw away' the dev branch once everything is ready (and tidy up the commits, so they will say something like 'Heavily revised drivers')
 
@JosephWright yes I just mention it as if (as i have) you ever do a build install in required/graphics then it's a bit arbitrary which one gets used. But anyway I just deleted the duplicates for now.
 
1:13 PM
@DavidCarlisle Something of a temporary hack
@DavidCarlisle Seeing @wipet reminds me to test on plain :) A few things need adjusting, I guess
 
@JosephWright one thing i was never sure of looking at HO's comments about some things added for plain was if I should just extend miniltx or if peopele are rellying on the exact versions there, things like % * Implementation of \providecommand is insufficient in miniltx.tex seem a bit odd if I have write access to all files...
 
@DavidCarlisle The ones I've found are problematic are the lack of \on@line and \MessageBreak, both of which are provided by infwarerr (so are in pdftex.def)
@DavidCarlisle In light of github.com/latex3/graphics-def/issues/4, I guess I need to add something for nat<thing> and the bb options, or perhaps we just ignore the entire 'these keys are daft' issue ...
 
1:53 PM
@DavidCarlisle Does “alternatoves” have something to do with Jabberwocky?
'Twas brillig, and the slithy toves
 
@egreg :)
 
@JosephWright Great piece of poetry, that one!
 
@DavidCarlisle If we do want to disable keys, perhaps best to do it by having a generic function in graphicx, but I'm not entirely sure this is the right approach. For example, if you set page for a bitmap it's just ignored, which is to me reasonable.
@egreg Indeed
@DavidCarlisle I see \\ strikes (yet) again: it's there just to be a rep-farm for you, isn't it ;)
 
@JosephWright I'd stop pdftex.def loading extrenal packages given half a chance:-)
 
@DavidCarlisle Er, that's partly what I've done :)
 
2:05 PM
@JosephWright yes
@JosephWright good a reason as any
@egreg I prefer the Italian poets
 
2:33 PM
@JosephWright speaking of miniltx I could try a bit harder to default a correct driver file:
 
@DavidCarlisle ???
 
oh perhaps that's OK, it's just a default default so gets fixed up by the standard graphics.cfg usually
@JosephWright taking to myself but pinging you:-)
@JosephWright \on@line I could easily add to miniltx, \MessageBreak is a bit tricker safe enough to add some groups to \def\PackageError#1#2#3{\errhelp{#3}\errmessage{#1: #2}} to set newlinechar to be ^^J and \MessageBreak to use that?
 
@DavidCarlisle More or less what HO does: both are defined inside a group
 
@JosephWright ok let's do that then should I check one in to graphics-def dev branch?
 
@DavidCarlisle Fine
@DavidCarlisle Hopefully we can get a few people to look over the ideas (@UlrikeFischer, @egreg, @HeikoOberdiek, ...)
@DavidCarlisle I guess some fallback is needed for the case where the update is made only to the .def, at least for a little while
 
3:03 PM
@JosephWright If there is some package I can attach through a local tree I can test the drivers and tell when something explodes ... or do they need new formats?
 
@UlrikeFischer they don't need a new format just any input path that puts that code in front of the code on latex/graphics, latex/graphics-def and plain/graphics-pln
@UlrikeFischer then just see if anything involving colour or graphics still works;-)
 
@DavidCarlisle That's easy, I only need a texmf-tree with the files. (and with texlive and the new auxtree option of tlmgr it should also be easy for others to tests).
 
@UlrikeFischer bah humbug I just installed them into my existing ~/texmf local tree (ohh except for miniltx, i should add that to build.lus)
 
@DavidCarlisle Installing new files is not the difficult part, it is removing them when they are no longer needed -- every few weeks I find some old long forgotten file lurking in my local texmf and grinning evily.
@DavidCarlisle Btw: I didn't know that Humbug came from the english into the german.
 
3:21 PM
@UlrikeFischer never happened to me, I'm sure:-)
@UlrikeFischer I'd have guessed the other direction (since it has no obvious etymology in English:-)
 
[paulo@cambridge Tese] $ pdfinfo tese.pdf | grep Pages
Pages:          11
@DavidCarlisle ^^
 
@JosephWright perhaps although I'd be tempted to just make \messageBreak` be \space as a fallback rather than defining \PackageError in the driver files
@PauloCereda I do hope you are working to base 256
 
@DavidCarlisle Alternative is to reword the messages so they don't need them, of course
 
@JosephWright that's what I did originally:-)
 
@DavidCarlisle Might be the path of least resistance
 
3:35 PM
@DavidCarlisle :)
 
@JosephWright I don't think it would be too bad to define them as space and empty if not already defined it's only the log file and only for people using plain who update latex/graphics-def but not plain/graphics-pln and that's not going to be many people
 
@DavidCarlisle We can always back out the change if we don't fancy it
 
@JosephWright arguably it ought to be in color and graphics if it goes anywhere so work for all drivers \ifx\@onllne\@undefined...`
:37727369 $ pdfjam -o tese2.pdf  tese.pdf tese.pdf tese.pdf tese.pdf tese.pdf tese.pdf tese.pdf tese.pdf tese.pdf tese.pdf tese.pdf tese.pdf tese.pdf tese.pdf tese.pdf tese.pdf tese.pdf tese.pdf
$ pdfinfo tese2.pdf | grep Pages
Pages:          198
 
3:53 PM
oh
 
@JosephWright OK I see you just avoided in the end:-)
@PauloCereda you see, it's easy when you know how
 
@DavidCarlisle :)
 
 
3 hours later…
6:42 PM
o.O
 
7:26 PM
@egreg Releasing to CTAN ...
 
@JosephWright is this good?
 
@PauloCereda We will see ;)
 
@JosephWright Godspeed
 
@PauloCereda Do you like regexes?
@PauloCereda Fingers-crossed
 
@JosephWright Let me put this way, I know them. :)
@JosephWright What can I be of service? :)
 
7:41 PM
@PauloCereda Ours are special ;)
 
@JosephWright oh my
 
@PauloCereda We have catcodes :)
 
@JosephWright Regex on the surface. :)
@JosephWright: which reminds me: I need to poke Frank about some projects. :)
 
Also includes \tl_show_analysis:N, which is very handy
 
@JosephWright ooh
> Less talkey talkey, more makey makey.
^^ interesting slogan for my thesis :)
 
8:10 PM
Hola' to everybody
@PauloCereda @JosephWright good evening and greetings
 
@Sebastiano 'ello
 
@PauloCereda @JosephWright I have a little problem with the size of a character
\begin{equation}\label{Lorentz2}
\left\{\begin{aligned}
x&\rightleftarrows x'\\
y&\rightleftarrows y'\\
z&\rightleftarrows z'\\
t&\rightleftarrows t'\\
v&\rightleftarrows -v\\
\end{aligned}\right.
\end{equation}
How Can I obtain a little \rightleftarrows? It is very large compared to the characters
@JosephWright @PauloCereda I have solve myself the problem. sorry
haloooooooooooo
 
8:25 PM
@Sebastiano really you should not ping people to ask questions, just ask then people may answer if they have time
5
 
8:39 PM
@PauloCereda @DavidCarlisle commitstrip.com/en/2017/05/29/trapped?
3
 
8:55 PM
@JosephWright At work, I feel like the guy in your strip... :):):)
 
@CarLaTeX I'm a TeX programmer, so much more the second case
 
@JosephWright I don't know which is worse :):):)
 
@AlanMunn Yeah! Honorary Italian citizenship for you!
 
9:10 PM
@CarLaTeX I'm sure this must happen to @DavidCarlisle all the time.
 
@AlanMunn LOL
 
@JosephWright Is \int_set:Nn \l_tmpa_int { \int_to_roman:n {#1} } supposed to work?
 
@egreg Oh, I was thinking from!
@egreg Corrected
 

« first day (2402 days earlier)      last day (2524 days later) »