@PhelypeOleinik -- I haven't found the videos for Doug McKenna's talk, but a search has determined that the slides for the talk do exist, linked from the web page for the TUGboat proceedings issue: tug.org/TUGboat/Contents/contents35-2.html That might be better than nothing.
@barbarabeeton It would be a shame if these talks were lost. There seems to be many interesting topics there. But the article will do for now. Thanks a lot for looking into it!
@manooooh -- Definitely option 2, but there may be regional differences. What I have recommended is the style used by U.S. publishers. Also, the first line should really start to the left of the second and following lines; also a U.S. style.
@barbarabeeton oki doki, thanks! I would go for 2, toó. It is obvious, but for some teachers that I have the pleasure to be its student, it is not obvious. They not only choose option 1, but they also do not respect that in the first line there must go two expressions instead of 1
Another question: what is your favourite (or the correct): "I am a duckling, too" or "I am also a duckling"? This difference appears between US and British people?
@UlrikeFischer It seems as if I was just too eager with scoping commands. This fix seems to work github.com/hmenke/pgf/commit/… (at least it fixes your example and it still compiles the manual). Should I merge it?
@JosephWright, what if your boxes contain the same long paragraph and the only distinction between them is a single word? Wouldn't it make more sense to accept that word as an argument in a macro which sets/saves the box rather than creating two separate boxes? (Of course you can place a paragraph in a \def but it's different if the paragraph is actually a picture, which needs boxing to avoid nesting when placed inside another picture).
@bp2017 If you want to reuse the text, save as a macro. A box is a 'completed' piece of typesetting, you can't go back and change it, only reuse it 'as is'
@DavidCarlisle ;-). But I think the main reason for a box is so that is can be stored outside a tikz picture, and then changing it inside the tikz picture like this won't work.
@HenriMenke if I say yes, @DavidCarlisle will blame me if it breaks tikz ;-(.
@DavidCarlisle that's true for the first box. But what if @bp2017 wants to change and reuse the box inside the same tikz environment?
@DavidCarlisle I have been pondering about @AlexG's code here chat.stackexchange.com/transcript/message/50667436#50667436. That's an "ifthen" code right? And if .setopacityalpha is not known then it actually executes \special{ps: mark /CA 0.5 /ca 0.5 /SetTransparency pdfmark }?
@UlrikeFischer Yes, that's a real problem. In my institute we have a central Windows server with Distiller installed, though not the current DC, but Acrobat version XI. But it's enough for such tests.
@AlexG another question. pdfbase is using this syntax for the EMC (and similar for BDC): \special{ps:~SDict~begin~mark~/EMC~pdfmark~end}, while accsupp uses \special{ps:[/EMC~pdfmark}. Does it make a difference? Which do you recommend?
@AlexG I do have the distiller, I can run tests - it only needs more time if one has to consider this too ;-(.
@UlrikeFischer You can safely use the accsupp way. Pushing SDict on the dict stack doesn't make sense here, actually, as nothing is used from it, only /systemdict (which means built-in) commands mark and pdfmark are used. BTW [ and mark are the same PS commands. But I prefer mark because an unbalanced [ screws up syntax highlighting in my text editor.
@AlexG so \special{ps:~mark~/EMC~pdfmark}? ( the [ is bothering me too, I have to add a closing ] after a comment to get the brace matching on track again, if I can replace it, this would be fine)
@UlrikeFischer mark is a keyword in PS. Yes, there should be some space around it, as in any other programming language. But the amount of space doesn't matter.
@AlexG good. And for dvipdfmx, pdfbase uses code: \special{pdf:code~EMC} while accsupp uses content: \special{pdf:content~EMC}. Would this matter somewhere?
the documentation of dvipdfmx says that code differs "from dvipdfm’s content special in that it does not enclose contents with a q and Q (save-restore of graphics state) pair".
EMC doesn't change the graphics state. So I thought saving it before EMC and restoring it afterwards would not be necessary. I have never seen EMC been bracketed as q EMC Q in any PDF code.
@UlrikeFischer, @DavidCarlisle I missed on fix but I realise I want to hold that one back anyway (it's linked to LaTeX2e and the current branches, or rather I don't want to fix it until we merge dev in)
@DavidCarlisle I first should run the tests and harftex and look if the lastest changes broke something. But yesterday some tests failed on luatex because \kern0.28 (font) changed to kern0.28 again ;-(. I have no idea why. This (font) is appearing and disappearing on its own will.
@JosephWright I think I have worked out the bdc/Properties stuff. It works in all four backends.
@AlexG that's quite possible. With a lot of this code one has to keep in mind that it is often old, and perhaps simply had no alternative at the time it was written.
@UlrikeFischer check whether the raw log changed or if the l3build normalisation is different it tries to hide a lot of luatex differences there but it's quite fragile
@DavidCarlisle oh, you are right, I updated l3build in between. The raw log shows \kern0.28 (font), but the l3build log \kern0.28. But what could have changed to give the difference?
@UlrikeFischer I wonder who we could blame for l3build changes?
@UlrikeFischer are you sure you are picking up the luatex-specific normalization? not changing the (font) part would be consistent with just using the standard normalisation
@DavidCarlisle I'm just downloading the newest l3build to check if my manual changes messed up something. But the problem is that the font part is changed (now), so if it is this problem, I must have messed it up when I generated the tlg.
@HaraldHanche-Olsen No, that's just @DavidCarlisle: for LaTeX3 issues, blame me, for LaTeX2e ones, blame ... Frank or if really desparate Leslie Lamport :)
@UlrikeFischer There's also a normalisation change we need, but I realised it should be held until the LaTeX2e trees are sorted (i.e. after the dev release)
@DavidCarlisle Speaking of that, Frank keeps saying utf8andspace depends on you fixing some stuff ...
@UlrikeFischer It's been a while since we needed any l3build changes: I really still need to sort stuff for biblatex I guess, but other than that it's pretty stable now
@JosephWright I'm still interested in running latex+dvips tests. Currently you can't even load regression-test as it inserts dvipdfmx-specials (and I'm quite willing in building dvipdfmxlatex and dvipslatex ;-)).
@UlrikeFischer Hmm, OK. Do you want to log that as a feature request? We should think about the implications
@UlrikeFischer I've got to make changes for expl3-in-the-format too, so that would be a good opportunity. (That might not be in l3build core, but it is going to be needed somewhere)
Is it possible to have LaTeX (or another program) to expand a the macros in a .tex file, and output the expanded LaTeX source?
If no such program exists: are LaTeX macros processed simply as string substitutions, before the document is interpreted? If so, it might be easy enough to write my own...
@JosephWright apparently if you had an l3 milestone, triaged all the issues that pblocked that, closed those issues, you could release latex3 and go home
@FaheemMitha yes or at least it logs all the actions of its normal use so it isn't like \meaning or \shoe it really sets a section, but logs what it does
@FaheemMitha \tracingall can produce a lot (and I mean a lot as in fill your disk) of tracing so only apply it to small parts, putting it at the top of a beamer document will produces gigabytes of log
@FaheemMitha it's not like \meaning or \show that you apply to a token it is a switch that turns on tracing which is why I suggested starting a group typesetting a heading and then closing the group to stop tracing )
[paulo@satyagraha ~] $ snap install powershell
error: This revision of snap "powershell" was published using classic
confinement and thus may perform arbitrary system changes outside of the
security sandbox that snaps are usually confined to, which may put your
system at risk.
If you understand and want to proceed repeat the command including
--classic.
[paulo@satyagraha ~] $
@PhelypeOleinik Seriously, you and @egreg can handle the 'user promotion' part of the work, leaves @UlrikeFischer and me to do all the PDF stuff and @DavidCarlisle to do ... things
@FaheemMitha if you search for repostings of the above you will find it was posted even more times than the "mean" one, always with the answer no (not always truthfully) so don't blame me:-)
@JosephWright I got all testfiles for luaotfload sorted for luatex + harftex, and now it fails for the experimental luatex (which is another path) as it is called luatex but should use the harftex.tlg -files ;-(. Life is hard ...
@PhelypeOleinik Some @PauloCereda jokes are ununderstandable even to the people mentioned in them. I was expecting to find some prior chat message in which I mentioned Grêmio, but no luck.
@DavidCarlisle -- Rectangular is okay, but it should be covered with thin fresh tomato slices and pesto, or with slivers of artichoke hearts and black olives.
