April 2025 - ntg-context - NTG Mailing Lists

pdfuaid metadata info not added to xmp
by Pablo Rodriguez 05 Apr '25

05 Apr '25

Hans, compiling the following source: \setupbackend[format=pdf/ua-2] \setuptagging[state=start] \setupstructure[state=start] \setupbackend [format=PDF/A-3a, intent=sRGB IEC61966-2.1, level=0] \setupexternalfigures[location=default] \starttext \externalfigure[cow][label={this is a cow}] \stoptext I’m afraid that veraPDF warns about ua-1 missing metadata info (https://github.com/veraPDF/veraPDF-validation-profiles/wiki/PDFUA-Part-1-ru…) This should be the reason why veraPDF cannot autodetect that the output from the sample above is PDF/UA-2 (althought it passes this validation profile when specified). I see that lines 600 to 750 from lpdf-fmt.lmt include these metadata (even with WTPDF-1.0 accesibility and reuse), but somehow they don’t end up in the output PDF document. Could you take a look at this? Many thanks for your help, Pablo

2 2

update
by Hans Hagen 04 Apr '25

04 Apr '25

We pushed the first post texlive 2025 update. Apart from a few pending patches, we spend some time on what we call retrotex and retropost. retrotex: Due to external factors the decision was made to let luatex math default to a more conservative handling of fonts: we just assume that opentype fonts are implemented like traditional tex fonts. This makes math in the engine behave like xetex. The primitives that relate to controling some aspects of math rendering are obsolete (we also removed experimental additional italics mode code). This change doesn't affect ConTeXt (MkIV) because there we do things differently. The adapted engine will be used in texlive 2026 (but we already have it in the distribution). retropost: We had to spend some time on fixing a bug in the mp library, one that was introduced when the additional number models were implemented and that surfaced when testing files of a power user. It was a rather nasty one that only showed up when using equations with no viable solution. In the process we (MS and HH) had to consult the metafont books, because we basically had to check all relevant code, and then fell in love with bitmaps. That made us decide to bring back bitmaps in the mp engine even if it sounds somewhat going back in time. We actually had some other reasons but those will be revealed at the meeting. tariffs: Talking of the ctx meeting, we might have to make some decisions there. A recent post on the list indicated that some installations default to letter size paper format, which one canconsider a bad idea but let's put that aside. The fact that installations can adapt that way, makes that we can also implement a tex tariffs system (tts) in accordance with current and upcoming European guidelines. For instance we can make context run 25 percent slower when the main language is us-english or 50 percent when fonts originating at (spineless) big tech companies are used. It's not that hard to do this (via directives). We assume that at some point tex installations used in the usa have to provide similar means but we leave that to those responsible for the distributions. After all, they already have to deal with excluding files that show traces of support for commercial fonts. At the meeting we might also decide to default to Canadian English patterns or those used in GB. And who knows if and when we're forced to (or feel that we have to) remove software components from the engine that relate too much to big tech or come with unacceptable demands for conformance. The good news here is that we have no dependencies that can't be overcome. Maybe we should also support the euro as math marker or just ditch the dollar for that purpose. Mixed in with preparations for the upcoming BT we will come back to some issues reported on the list. We've been a bit silent for various reasons but Wolfgang made sure all went well which feels good. Hans & Mikael ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl -----------------------------------------------------------------

7 9

TaggedPDF & Screenreaders
by Henning Hraban Ramm 04 Apr '25

04 Apr '25

Here are a few recordings how screenreaders handle tagged or untagged PDFs (from LaTeX): https://latex3.github.io/tagging-project/documentation/wtpdf/fulldoc HR

2 1

nested context/lua calls
by Mohammad Hossein Bateni 03 Apr '25

03 Apr '25

Hello, I have some code like this: --------------------------------- \def\MakeSomething{... some code that returns number 12} \startluacode function WorkWithSomething(data) -- call MakeSomething to obtain 12 -- base on the value above make some decisions end WorkWithSomething(some_data) \stopluacode --------------------------------- If I use context.MakeSomething() inside the lua block, this is buffered till the end of the block, so I can't get the result and work with it. If MakeSomething is a command already "edef-ed", I can use tokens.getters.macro.MakeSomething to find the value. If MakeSomething has an argument which is determined within the lua block, then I can't use this hack. The ugly partial solution I have found is to change the signature of WorkWithSomething to receive an argument and then I replicate the parameters of MakeSomething on the TeX side: --------------------------------- \def\MakeSomething#1{...} \startluacode function WorkWithSomething(data,MakeSomethingOutput) -- we wanted to use some computation on data to obtain input to MakeSomething -- then call MakeSomething to obtain its result end \stopluacode \ctxlua{WorkWithSomething(some_data, "\MakeSomething{replicate processing on data to obtain input to this macro}")} --------------------------------- Is there a more elegant solution? Ideally I want a way to start a TeX process under Lua, and process the macros there before I get to the next line of Lua code. In my motivation, MakeSomething is \datasetvariable. I wanted to look up something in the dataset and use it for the rest of computation/typesetting. Thanks a lot, --Mohammad

3 4

Re: How to get TABLE to start on the same page, split when needed?
by Joel 03 Apr '25

03 Apr '25

\starttext \input knuth \input knuth \input knuth {\boldface This is instructions.} \bTABLE[split=yes, frame=off] \bTR\bTD ... \eTD\bTD \input knuth \input knuth \eTD\eTR \bTR\bTD ... \eTD\bTD ... \eTD\eTR \bTR\bTD ... \eTD\bTD ... \eTD\eTR \bTR\bTD ... \eTD\bTD ... \eTD\eTR \bTR\bTD ... \eTD\bTD ... \eTD\eTR \bTR\bTD ... \eTD\bTD ... \eTD\eTR \bTR\bTD ... \eTD\bTD ... \eTD\eTR \bTR\bTD ... \eTD\bTD ... \eTD\eTR \bTR\bTD ... \eTD\bTD ... \eTD\eTR \bTR\bTD ... \eTD\bTD ... \eTD\eTR \bTR\bTD ... \eTD\bTD ... \eTD\eTR \bTR\bTD ... \eTD\bTD ... \eTD\eTR \bTR\bTD ... \eTD\bTD ... \eTD\eTR \bTR\bTD ... \eTD\bTD ... \eTD\eTR \bTR\bTD ... \eTD\bTD ... \eTD\eTR \bTR\bTD ... \eTD\bTD ... \eTD\eTR \eTABLE \stoptext here is a minimum working example. If it sees that one sell is rather big, like in the case the first one is, it decides to put it on the next page, instead of just splitting that cell up. --Joel

2 1

How to get TABLE to start on the same page, split when needed?
by Joel 02 Apr '25

02 Apr '25

I was using tabulate, but ended up with some strange error I could not track down. So I switched to using TABLE, the error went away, but now I have a new problem: If the TABLE is really big, but fits on the next page, it seems to add a \pagebreak before, then shows the table. This is wasting way too much paper, it means I have instructions, then the activity appears on the entirely next page, with a big unneeded blank space given. I need the table to start, then split whenever the end of the page is reached. In the wiki, I found this comment from 2021: "Beware: Split tables often begin not on the page (or in the column, if you use them) where they could start, but on/in the next one.This is a known limitation due to the complicated calculation of remaining space on the page." Is this still an error in TABLE, or is there a fix? --Joel

1 0

Cron <context@cgl> /var/www/aanhet.net/context/bin/cron/context-mirror
by root＠ntg.nl 01 Apr '25

01 Apr '25

receiving incremental file list ./ ctan.lsr document-2.htm download-1.htm download-2.htm logo-ade.png logo-cts.png logo-pod.png rss.xml show-fil.pdf context/latest/ context/latest/cont-lmt.zip context/latest/cont-mpd.zip context/latest/cont-ppc.zip context/latest/cont-sci.zip context/latest/cont-tmf.zip context/latest/cont-tst.7z context/latest/cont-tst.tar.xz context/latest/cont-tst.zip sent 182,962 bytes received 44,180,101 bytes 4,669,796.11 bytes/sec total size is 626,536,639 speedup is 14.12 Running archiver: New dir: /var/www/aanhet.net/context//htdocs/archives/context-2025-04-01.14 277221071 /var/www/aanhet.net/context//htdocs/archives/context-2025-04-01.14/latest 126741221 /var/www/aanhet.net/context//htdocs/archives/context-2025-04-01.14/current 403962292 /var/www/aanhet.net/context//htdocs/archives/context-2025-04-01.14 403962292 total

1 0