* lmtx upload
@ 2021-01-05 10:01 Hans Hagen
2021-01-05 15:23 ` Pablo Rodriguez
0 siblings, 1 reply; 8+ messages in thread
From: Hans Hagen @ 2021-01-05 10:01 UTC (permalink / raw)
To: mailing list for ConTeXt users
Hi,
A new version has been uploaded. This one has:
\enableexperiments[fonts.compact]
for the brave to test.
As a reminder, as this has been around for a while now, if you really
want to stress your test, run with (top of file):
\overloadmode 4 % error
\overloadmode 3 % warning
in order to see if you (re)define primitives, core functionality,
instances and whatever can lead to conflicts. There are still some
commands to be tagged (and for some the category might change) but it's
a bit of a tedious job.
Btw, for metafun the warnings are enabled by default (for now), but
users might have noticed that already.
Hans
-----------------------------------------------------------------
Hans Hagen | PRAGMA ADE
Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl
-----------------------------------------------------------------
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!
maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage : http://www.pragma-ade.nl / http://context.aanhet.net
archive : https://bitbucket.org/phg/context-mirror/commits/
wiki : http://contextgarden.net
___________________________________________________________________________________
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: lmtx upload
2021-01-05 10:01 lmtx upload Hans Hagen
@ 2021-01-05 15:23 ` Pablo Rodriguez
0 siblings, 0 replies; 8+ messages in thread
From: Pablo Rodriguez @ 2021-01-05 15:23 UTC (permalink / raw)
To: ntg-context
On 1/5/21 11:01 AM, Hans Hagen wrote:
> Hi,
>
> A new version has been uploaded. This one has:
>
> \enableexperiments[fonts.compact]
>
> for the brave to test.
Hi Hans,
many thanks for your new release with the experimental feature.
I get a compilation time decrease from 64s to 58s in one of the
documents I have.
Unrelated to this, I’m afraid that the current latest doesn’t like my
definition of small caps and I don’t get them (either with or without
experiments) in many documents.
But I know I have to provide a sample that displays the issue.
Many thanks for the new upload,
Pablo
--
http://www.ousia.tk
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!
maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage : http://www.pragma-ade.nl / http://context.aanhet.net
archive : https://bitbucket.org/phg/context-mirror/commits/
wiki : http://contextgarden.net
___________________________________________________________________________________
^ permalink raw reply [flat|nested] 8+ messages in thread
* lmtx upload
@ 2023-01-04 11:22 Hans Hagen via ntg-context
0 siblings, 0 replies; 8+ messages in thread
From: Hans Hagen via ntg-context @ 2023-01-04 11:22 UTC (permalink / raw)
To: mailing list for ConTeXt users; +Cc: Hans Hagen
Hi,
I uploaded the first 2023 lmtx. There are not that many changes, most
relate to updating manuals. Currently we're looking at the metafun /
luametafun and lowlevel manuals - adding some accumulated todo's.
Keep in mind that when we update manuals that is is not automatically
reflected in new pdf's (processing happens outside the tree on temporary
paths) and only occasionally I push back a pdf (when it looks ok). So,
manual sources can be more up to date than manual pdfs. If you feel
something (fundamental and stable) is missing you can let us know here.
We gained some performance over the last month, mostly due to using
different low level mechanisms (there are some more not yet widely used
ones). The gain depends on (extreme) usage so quite likely most goes
unnoticed unless one has many runs. After all, there are no fundamental
changes in context, which is already reasonable optimized for ages (it's
kind of a game to make it faster without obscuring the code, which is a
criterium for rejection).
As usual, (new and old) bugs will be fixed,
Hans
-----------------------------------------------------------------
Hans Hagen | PRAGMA ADE
Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl
-----------------------------------------------------------------
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!
maillist : ntg-context@ntg.nl / https://www.ntg.nl/mailman/listinfo/ntg-context
webpage : https://www.pragma-ade.nl / http://context.aanhet.net
archive : https://bitbucket.org/phg/context-mirror/commits/
wiki : https://contextgarden.net
___________________________________________________________________________________
^ permalink raw reply [flat|nested] 8+ messages in thread
* lmtx upload
@ 2022-12-01 12:17 Hans Hagen via ntg-context
0 siblings, 0 replies; 8+ messages in thread
From: Hans Hagen via ntg-context @ 2022-12-01 12:17 UTC (permalink / raw)
To: mailing list for ConTeXt users; +Cc: Hans Hagen
Hi,
I uploaded lmtx. Most significate:
- A slightly different implementaqtion of font modes 'base' and 'none'
(hopefully no side effects).
- A few math updates (like primes done a bit differently, we made some
choices here).
- Limited support for colr format 1 as needed for the rather nice:
https://gitlab.com/smc/fonts/Nupuram/-/blob/master/docs/README.md
which is used like:
\startTEXpage[offset=1dk,align={normal,tolerant,stretch}]
\definefontfeature[colored][default][colr=9]
\definefontfeature[slanted][default][slant=.2]
\definedfont[name:nupuramregular*default] \input tufte
\par \blank
\definedfont[name:nupuramregular*slanted] \input tufte
\par \blank
% \definedfont[Nupuram-Color.colrv0.otf*colored] \input tufte
\par \blank
\definedfont[Nupuram-Color.colrv1.otf*colored] \input tufte
\par \blank
\definedfont[Nupuram-Dots.otf*default] \input tufte
\par \blank
\definedfont[Nupuram-Arrows-Color.colrv0*,colored] \input tufte \par
\stopTEXpage
(don't install variable fonts and instances alongside)
Hans
-----------------------------------------------------------------
Hans Hagen | PRAGMA ADE
Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl
-----------------------------------------------------------------
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!
maillist : ntg-context@ntg.nl / https://www.ntg.nl/mailman/listinfo/ntg-context
webpage : https://www.pragma-ade.nl / http://context.aanhet.net
archive : https://bitbucket.org/phg/context-mirror/commits/
wiki : https://contextgarden.net
___________________________________________________________________________________
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: lmtx upload
2021-06-24 22:18 ` luigi scarso
@ 2021-06-24 22:32 ` Hans Hagen
0 siblings, 0 replies; 8+ messages in thread
From: Hans Hagen @ 2021-06-24 22:32 UTC (permalink / raw)
To: mailing list for ConTeXt users, luigi scarso
On 6/25/2021 12:18 AM, luigi scarso wrote:
>
>
> On Thu, Jun 24, 2021 at 7:29 PM Hans Hagen <j.hagen@xs4all.nl
> <mailto:j.hagen@xs4all.nl>> wrote:
>
> Hi,
>
> I uploaded a new version. I got sidetracked by an experiment. We can
> read from zip files, but as I had some tar archives I wondered if we
> could read from them. They happened to be compressed. It uses an
> already
> present infrastructure (early mkiv I played a bit with reading from zip
> files).
>
>
> libarchive looks quite useful
> http://libarchive.org/ <http://libarchive.org/>
> (see
> https://docs.microsoft.com/en-us/virtualization/community/team-blog/2017/20171219-tar-and-curl-come-to-windows
> <https://docs.microsoft.com/en-us/virtualization/community/team-blog/2017/20171219-tar-and-curl-come-to-windows>)
But it's huge and add lots of dependencies and thereby complications;
often these libs have special platform optimizations (so one needs to
run configure scripts) ... in luametatex i try to stay away from that ..
but loading as external lib is of course an option; we already support
zip, gz, tar, lz4, lzma (xz), lzo, xstd but i doubt if in practice they
are used often (zip we need for pdf, gz for some graphics, xz because
it's used in texlive but most of it is also kind of fun to play with: as
less code as possible and then lua support code).
Hans
-----------------------------------------------------------------
Hans Hagen | PRAGMA ADE
Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl
-----------------------------------------------------------------
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!
maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage : http://www.pragma-ade.nl / http://context.aanhet.net
archive : https://bitbucket.org/phg/context-mirror/commits/
wiki : http://contextgarden.net
___________________________________________________________________________________
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: lmtx upload
2021-06-24 17:28 Hans Hagen
@ 2021-06-24 22:18 ` luigi scarso
2021-06-24 22:32 ` Hans Hagen
0 siblings, 1 reply; 8+ messages in thread
From: luigi scarso @ 2021-06-24 22:18 UTC (permalink / raw)
To: mailing list for ConTeXt users
[-- Attachment #1.1: Type: text/plain, Size: 568 bytes --]
On Thu, Jun 24, 2021 at 7:29 PM Hans Hagen <j.hagen@xs4all.nl> wrote:
> Hi,
>
> I uploaded a new version. I got sidetracked by an experiment. We can
> read from zip files, but as I had some tar archives I wondered if we
> could read from them. They happened to be compressed. It uses an already
> present infrastructure (early mkiv I played a bit with reading from zip
> files).
>
>
libarchive looks quite useful
http://libarchive.org/
(see
https://docs.microsoft.com/en-us/virtualization/community/team-blog/2017/20171219-tar-and-curl-come-to-windows
)
--
luigi
[-- Attachment #1.2: Type: text/html, Size: 1231 bytes --]
[-- Attachment #2: Type: text/plain, Size: 493 bytes --]
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!
maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage : http://www.pragma-ade.nl / http://context.aanhet.net
archive : https://bitbucket.org/phg/context-mirror/commits/
wiki : http://contextgarden.net
___________________________________________________________________________________
^ permalink raw reply [flat|nested] 8+ messages in thread
* lmtx upload
@ 2021-06-24 17:28 Hans Hagen
2021-06-24 22:18 ` luigi scarso
0 siblings, 1 reply; 8+ messages in thread
From: Hans Hagen @ 2021-06-24 17:28 UTC (permalink / raw)
To: mailing list for ConTeXt users
Hi,
I uploaded a new version. I got sidetracked by an experiment. We can
read from zip files, but as I had some tar archives I wondered if we
could read from them. They happened to be compressed. It uses an already
present infrastructure (early mkiv I played a bit with reading from zip
files).
\starttext
\dorecurse{100}{
\input {tar:///e:/temporary/test/foo.tar?name=test.tex}\par
\input {tar:///e:/temporary/test/foo.tar.xz?name=test.tex}\par
\input {tar:///e:/temporary/test/foo.tar.gz?name=test.tex}\par
}
\stoptext
The tar files are handled by lmtx itself, the gzipped variant uses the
basic decompressor of the engine (different in lmtx than in mkiv but
that should not matterl i only tested lmtx). The xz one is a bit more
special as it uses the external xz program unless one has limlzma
installed, e.g. in:
tex/texmf-win64/bin/lib/luametatex/lzma/liblzma.dll
(mtxrun --generate makes it locatable).
In all three examples performance is quite ok. It might a nice solution
for cases where one has many small files cluttering the system.
It actually might be more efficient to use tar.xz instead of zip files
but then we also need to ship the lib and that is not on the agenda. I'm
in no hurry with such alternatives (most of todays bandwidth is wasted
with ads and related crap anyway).
Concerning math .... we have several fonts, and the public ones don't
really play the game like cambria does. So, th emore bugs become
features, the more we're stuck with it. Now, once fonts are stable
(including bugs) we cna think of goodie files that fix things but an
other option is to tell the math machinery in what aspects as font is
not really opentype. Read: when to follow a traditional code path.
For the moment that is just an experiment and mostly meant for myself to
explore things but it is likely to stay. Contrary to other variabls that
control the math engine, here we have a bitset (adding dozens more
commands makes no sense):
\setupbodyfont[newcomputermodern-book]
\starttext
\startTEXpage[offset=10pt]
$
\showglyphs\showmakeup[math] f = 10
$
$
\mathcontrolmode\plusone
\bitwiseflip\mathfontcontrol\orditalickernmathcontrolcode
\bitwiseflip\mathfontcontrol\charitalicnoreboxmathcontrolcode
\bitwiseflip\mathfontcontrol\orditalickernmathcontrolcode
\bitwiseflip\mathfontcontrol\charitalicnoreboxmathcontrolcode
\bitwiseflip\mathfontcontrol\overrulemathcontrolcode
\bitwiseflip\mathfontcontrol\underrulemathcontrolcode
\bitwiseflip\mathfontcontrol\radicalrulemathcontrolcode
\bitwiseflip\mathfontcontrol\fractionrulemathcontrolcode
\bitwiseflip\mathfontcontrol\accentskewhalfmathcontrolcode
\bitwiseflip\mathfontcontrol\accentskewapplymathcontrolcode
\bitwiseflip\mathfontcontrol\accentitalickernmathcontrolcode
\bitwiseflip\mathfontcontrol\delimiteritalickernmathcontrolcode
\bitwiseflip\mathfontcontrol\charitalicwidthmathcontrolcode
\bitwiseflip\mathfontcontrol\boxednoitalickernmathcontrolcode
\bitwiseflip\mathfontcontrol\nostaircasekernmathcontrolcode
\bitwiseflip\mathfontcontrol\textitalickernmathcontrolcode
\showglyphs\showmakeup[math] f = 10
$
\stopTEXpage
\stoptext
Here, when \mathcontrolmode larger than zero, the \mathfontcontrol
bitset will control code paths. When the value is zero, a normal
opentype font is assumed. There might be some more options but it's a
bit tricky (maybe I already introduced some issues, who knows).
A font can also communicate a bitset and this is supported by a font
feature in lmtx:
% \definefontfeature[newcm-mathextra][mathcontrol={%
% over_rule,%
% under_rule,%
% radical_rule,%
% fraction_rule,%
% accent_skew_half,%
% accent_skew_apply,%
% accent_italic_kern,%
% delimiter_italic_kern,%
% ord_italic_kern,%
% char_italic_width,%
% char_italic_no_rebox,%
% boxed_no_italic_kern,%
% no_staircase_kern,%
% text_italic_kern,%
% }]
the newcm-mathextra is for instance applied with that font when defined
(as usual with features), Again there is some detailed control and more
can be added. So we have a control variable at the tex end and one that
goes with a font. The 'bits' have funny names but such are the
internals. Eventually I'll wrap this up in a progress article.
In addition to the above there are a few fixes for issues discussed on
the mailing list.
I'm still trying to get tracing a bit nicer, so here is another trick:
\starttext
\def\MyPrimitiveCommandA#1{\hbox to 3cm{#1}}
\untraced\def\MyPrimitiveCommandB#1{\hbox to 3cm{#1}}
\bgroup
\tracingall
\MyPrimitiveCommandA{oeps}
\MyPrimitiveCommandB{oeps}
\egroup
The first one gives:
2:3: \MyPrimitiveCommandA #1->\hbox to 3cm{#1}
2:3: #1<-oeps
2:4: {\hbox}
2:5: {entering adjusted hbox group (level 2) at line 10}
and the second one:
2:3: \MyPrimitiveCommandB
2:4: {\hbox}
2:5: {entering adjusted hbox group (level 2) at line 11}
not that users do much tracing, but sometimes it helps not to see
clutter in the log for (e.g.) low level helpers that act like
primiteves. Taks:
\setbox0\hbox{test me}
With "\showbox 0" we get:
> \box0=
1:3: \hbox[hbox][color=1,colormodel=1], width 38.69073, height 7.37595,
depth 0.13193, direction l2r
1:3: .\list
1:3: ..\glyph[32768][color=1,colormodel=1], language (n=1,l=2,r=3),
hyphenationmode "79F3F, options "80, font <2: newcm10-book @ 12.0pt>,
glyph U+000074 t
...............................
But with "\showboxhere 0" gives
\hbox[hbox][color=1,colormodel=1], width 38.69073, height 7.37595, depth
0.13193, direction l2r
.\list
..\glyph[32768][color=1,colormodel=1], language (n=1,l=2,r=3),
hyphenationmode "79F3F, options "80, font <2: newcm10-book @ 12.0pt>,
glyph U+000074 t
...............................
WHen we say:
\meaning\showboxhere
it reports itself as for instance \relax does. This because it is
defined with the \untraced prefix (it's also unexpandable and protected
against overloading when one has overloadmode set).
\untraced\permanent\protected\def\showboxhere{\showbox nolevels content
online } % no group/input info, only content, and forced online
(There are cases where the meaning will show up, but at least it adds a
little to the abstraction.)
Hans
-----------------------------------------------------------------
Hans Hagen | PRAGMA ADE
Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl
-----------------------------------------------------------------
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!
maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage : http://www.pragma-ade.nl / http://context.aanhet.net
archive : https://bitbucket.org/phg/context-mirror/commits/
wiki : http://contextgarden.net
___________________________________________________________________________________
^ permalink raw reply [flat|nested] 8+ messages in thread
* lmtx upload
@ 2021-01-11 16:44 Hans Hagen
0 siblings, 0 replies; 8+ messages in thread
From: Hans Hagen @ 2021-01-11 16:44 UTC (permalink / raw)
To: mailing list for ConTeXt users
Hi,
I uploaded a new lmtx. Nothing spectacular as I'm still playign with
compact font support (as explained in previous mails). Most seem to work
now.
New magic:
\startTEXpage[offset=10pt]
\vl\glyph `M\vl\quad
\vl\glyph raise 3pt `M\vl\quad
\vl\glyph left 3pt `M\vl\quad
\vl\glyph right 2pt`M\vl\quad
\vl\glyph raise -4pt right 4pt`M\vl\quad
\vl\glyph left -1pt right -1pt`M\vl
\stopTEXpage
This is part of some experiments with manipulating glyph dimensions
instead of font kerns. These extensions will stay (who knows when it
comes un handy) but we will probably stick to font kerns as default.
\startTEXpage[offset=10pt]
\showfontkerns
\showglyphs
\strut average\par
\vskip-2ex
\enabledirectives[fonts.injections.method=advance]
{\red \strut average}\par
\stopTEXpage
Both methods have their pros and cons.
Hans
-----------------------------------------------------------------
Hans Hagen | PRAGMA ADE
Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl
-----------------------------------------------------------------
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!
maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage : http://www.pragma-ade.nl / http://context.aanhet.net
archive : https://bitbucket.org/phg/context-mirror/commits/
wiki : http://contextgarden.net
___________________________________________________________________________________
^ permalink raw reply [flat|nested] 8+ messages in thread
end of thread, other threads:[~2023-01-04 11:22 UTC | newest]
Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-01-05 10:01 lmtx upload Hans Hagen
2021-01-05 15:23 ` Pablo Rodriguez
2021-01-11 16:44 Hans Hagen
2021-06-24 17:28 Hans Hagen
2021-06-24 22:18 ` luigi scarso
2021-06-24 22:32 ` Hans Hagen
2022-12-01 12:17 Hans Hagen via ntg-context
2023-01-04 11:22 Hans Hagen via ntg-context
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).