Compilation of manual.tex (testers needed)
7 views
Skip to first unread message
Kishore G
Mar 25, 2026, 2:14:53 AM (12 days ago) Mar 25
to pencil-code-discuss
Hi all.
Currently, compilation of the manual (doc/manual.tex) is broken with the
latest version of texlive because we are using a very old version of input.sty
in doc/texinputs [2]. While replacing it by the latest version of input.sty
fixes the issue, I suspect it will break compilation with older LaTeX versions
[1].
I have tried to backport the fix for the issue in the branch
`indexsty_backport_fix` (on pencil-code.org). However, I have only tested it
with the latest version of texlive (2026). It would be helpful if users of
older texlive versions (and possibly other tex distributions) could verify
that compilation still works with my changes.
After checking out the `indexsty_backport_fix` branch, you just need to do
```
cd $PENCIL_HOME/doc
make clean
make
```
and check if manual.pdf is successfully generated. If it doesn't work, please
send me the error message and the output of
```
tex --version
```
[1] the latest index.sty says `\NeedsTeXFormat{LaTeX2e}[2024/11/01]`
[2] https://github.com/davidmjones/ltx-index/issues/1
--
கிஷோர் கோ. (Kishore G.)
https://kishore96in.github.io/
Currently, compilation of the manual (doc/manual.tex) is broken with the
latest version of texlive because we are using a very old version of input.sty
in doc/texinputs [2]. While replacing it by the latest version of input.sty
fixes the issue, I suspect it will break compilation with older LaTeX versions
[1].
I have tried to backport the fix for the issue in the branch
`indexsty_backport_fix` (on pencil-code.org). However, I have only tested it
with the latest version of texlive (2026). It would be helpful if users of
older texlive versions (and possibly other tex distributions) could verify
that compilation still works with my changes.
After checking out the `indexsty_backport_fix` branch, you just need to do
```
cd $PENCIL_HOME/doc
make clean
make
```
and check if manual.pdf is successfully generated. If it doesn't work, please
send me the error message and the output of
```
tex --version
```
[1] the latest index.sty says `\NeedsTeXFormat{LaTeX2e}[2024/11/01]`
[2] https://github.com/davidmjones/ltx-index/issues/1
--
கிஷோர் கோ. (Kishore G.)
https://kishore96in.github.io/
Axel Brandenburg
Mar 25, 2026, 2:59:36 AM (12 days ago) Mar 25
to pencil-co...@googlegroups.com
Hi Kishore,
For me, the compilation works (the latest change was from yesterday,
r42997 | ToxPuro | 2026-03-21 14:16:39 +0100). I just updated the pdf
file on the webpage, and the same also for quick-guide/quick_start.tex
(r43028 | ToxPuro | 2026-03-24 10:51:14 +0100).
My version is:
pdfTeX 3.141592653-2.6-1.40.22 (TeX Live 2022/dev/Debian)
kpathsea version 6.3.4/dev
Copyright 2021 Han The Thanh (pdfTeX) et al.
There is NO warranty. Redistribution of this software is
covered by the terms of both the pdfTeX copyright and
the Lesser GNU General Public License.
For more information about these matters, see the file
named COPYING and the pdfTeX source.
Primary author of pdfTeX: Han The Thanh (pdfTeX) et al.
Compiled with libpng 1.6.37; using libpng 1.6.37
Compiled with zlib 1.2.11; using zlib 1.2.11
Compiled with xpdf version 4.03
and I am compiling both with make.
Axel
> --
> You received this message because you are subscribed to the Google Groups "pencil-code-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pencil-code-dis...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/pencil-code-discuss/3407710.44csPzL39Z%40kishore-workpc-iitm.
For me, the compilation works (the latest change was from yesterday,
r42997 | ToxPuro | 2026-03-21 14:16:39 +0100). I just updated the pdf
file on the webpage, and the same also for quick-guide/quick_start.tex
(r43028 | ToxPuro | 2026-03-24 10:51:14 +0100).
My version is:
pdfTeX 3.141592653-2.6-1.40.22 (TeX Live 2022/dev/Debian)
kpathsea version 6.3.4/dev
Copyright 2021 Han The Thanh (pdfTeX) et al.
There is NO warranty. Redistribution of this software is
covered by the terms of both the pdfTeX copyright and
the Lesser GNU General Public License.
For more information about these matters, see the file
named COPYING and the pdfTeX source.
Primary author of pdfTeX: Han The Thanh (pdfTeX) et al.
Compiled with libpng 1.6.37; using libpng 1.6.37
Compiled with zlib 1.2.11; using zlib 1.2.11
Compiled with xpdf version 4.03
and I am compiling both with make.
Axel
> You received this message because you are subscribed to the Google Groups "pencil-code-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pencil-code-dis...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/pencil-code-discuss/3407710.44csPzL39Z%40kishore-workpc-iitm.
Kishore G
Mar 25, 2026, 3:30:34 AM (12 days ago) Mar 25
to Axel Brandenburg, pencil-co...@googlegroups.com
Hi Axel.
> For me, the compilation works
The issue is only there with texlive 2026 (you are using texlive 2022). I ran
into it after upgrading from texlive 2025 yesterday.
Can you check if the changes in the branch `indexsty_backport_fix` work on
texlive 2022? If so, I will merge them to master.
> For me, the compilation works
into it after upgrading from texlive 2025 yesterday.
Can you check if the changes in the branch `indexsty_backport_fix` work on
texlive 2022? If so, I will merge them to master.
Matthias Rheinhardt
Mar 26, 2026, 1:31:44 PM (10 days ago) Mar 26
to pencil-co...@googlegroups.com
Hi all,
for maxadvec we find in the code
dustvelocity.f90: maxadvec=maxadvec+p%advec_uud(:,k) ! k runs over the species!
heatflux.f90: maxadvec = maxadvec + c_spitzer/dxmin_pencil
hydro.f90: maxadvec=maxadvec+p%advec_uu
neutralvelocity.f90: maxadvec=maxadvec+p%advec_uun
I can't see a reason why there is summing-up: shouldn't the maximum of these advective velocities matter for the timestep instead of the sum as they all act on different quantities?
In particular for dust with a high number of variables, the result could be absurd.
Best,
Matthias
Von: pencil-co...@googlegroups.com <pencil-co...@googlegroups.com> im Auftrag von Kishore G <kishore...@gmail.com>
Gesendet: Mittwoch, 25. März 2026 08:30:24
An: Axel Brandenburg
Cc: pencil-co...@googlegroups.com
Betreff: Re: Compilation of manual.tex (testers needed)
Gesendet: Mittwoch, 25. März 2026 08:30:24
An: Axel Brandenburg
Cc: pencil-co...@googlegroups.com
Betreff: Re: Compilation of manual.tex (testers needed)
--
You received this message because you are subscribed to the Google Groups "pencil-code-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pencil-code-dis...@googlegroups.com.
You received this message because you are subscribed to the Google Groups "pencil-code-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pencil-code-dis...@googlegroups.com.
To view this discussion visit
https://groups.google.com/d/msgid/pencil-code-discuss/7064671.lOV4Wx5bFT%40kishore-workpc-iitm.
Axel Brandenburg
Mar 26, 2026, 3:44:19 PM (10 days ago) Mar 26
to pencil-co...@googlegroups.com
Hi Matthias,
It could be useful to have different options. In the special case of
advection together with sound and Alfven waves, the relevant wave speed
is +/- Uadvect + sqrt(cs^2+vA^2). This is why we worked with the sums of
squares. But it could be good to have a combination of switches, which
would allow us to study the relevant timestep in a variety of cases.
Note, in this connection, the difference between Eqs.(11) and (12) in
https://ui.adsabs.harvard.edu/abs/2020GApFD.114..162B/abstract
which is an example in radiation physics that illustrates the difference
between previously applied "reasoning", as applied to the Athena code,
and a detailed study.
Axel
On Thu, Mar 26, 2026 at 05:31:38PM +0000, 'Matthias Rheinhardt' via pencil-code-discuss wrote:
> Hi all,
>
>
> for maxadvec we find in the code
>
>
> dustvelocity.f90: maxadvec=maxadvec+p%advec_uud(:,k) ! k runs over the species!
>
> heatflux.f90: maxadvec = maxadvec + c_spitzer/dxmin_pencil
>
> hydro.f90: maxadvec=maxadvec+p%advec_uu
>
> neutralvelocity.f90: maxadvec=maxadvec+p%advec_uun
>
>
> I can't see a reason why there is summing-up: shouldn't the maximum of these advective velocities matter for the timestep instead of the sum as they all act on different quantities?
>
> In particular for dust with a high number of variables, the result could be absurd.
>
>
> Best,
>
> Matthias
>
>
> ________________________________
> Von: pencil-co...@googlegroups.com <pencil-co...@googlegroups.com> im Auftrag von Kishore G <kishore...@gmail.com>
> Gesendet: Mittwoch, 25. März 2026 08:30:24
> An: Axel Brandenburg
> Cc: pencil-co...@googlegroups.com
> Betreff: Re: Compilation of manual.tex (testers needed)
>
> To view this discussion visit https://groups.google.com/d/msgid/pencil-code-discuss/ae0d308a54de44c0899a7554af8a6e7f%40su.se<https://groups.google.com/d/msgid/pencil-code-discuss/ae0d308a54de44c0899a7554af8a6e7f%40su.se?utm_medium=email&utm_source=footer>.
It could be useful to have different options. In the special case of
advection together with sound and Alfven waves, the relevant wave speed
is +/- Uadvect + sqrt(cs^2+vA^2). This is why we worked with the sums of
squares. But it could be good to have a combination of switches, which
would allow us to study the relevant timestep in a variety of cases.
Note, in this connection, the difference between Eqs.(11) and (12) in
https://ui.adsabs.harvard.edu/abs/2020GApFD.114..162B/abstract
which is an example in radiation physics that illustrates the difference
between previously applied "reasoning", as applied to the Athena code,
and a detailed study.
Axel
On Thu, Mar 26, 2026 at 05:31:38PM +0000, 'Matthias Rheinhardt' via pencil-code-discuss wrote:
> Hi all,
>
>
> for maxadvec we find in the code
>
>
> dustvelocity.f90: maxadvec=maxadvec+p%advec_uud(:,k) ! k runs over the species!
>
> heatflux.f90: maxadvec = maxadvec + c_spitzer/dxmin_pencil
>
> hydro.f90: maxadvec=maxadvec+p%advec_uu
>
> neutralvelocity.f90: maxadvec=maxadvec+p%advec_uun
>
>
> I can't see a reason why there is summing-up: shouldn't the maximum of these advective velocities matter for the timestep instead of the sum as they all act on different quantities?
>
> In particular for dust with a high number of variables, the result could be absurd.
>
>
> Best,
>
> Matthias
>
>
> ________________________________
> Von: pencil-co...@googlegroups.com <pencil-co...@googlegroups.com> im Auftrag von Kishore G <kishore...@gmail.com>
> Gesendet: Mittwoch, 25. März 2026 08:30:24
> An: Axel Brandenburg
> Cc: pencil-co...@googlegroups.com
> Betreff: Re: Compilation of manual.tex (testers needed)
>
> --
> You received this message because you are subscribed to the Google Groups "pencil-code-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pencil-code-dis...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/pencil-code-discuss/7064671.lOV4Wx5bFT%40kishore-workpc-iitm.
> You received this message because you are subscribed to the Google Groups "pencil-code-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pencil-code-dis...@googlegroups.com.
>
> --
> You received this message because you are subscribed to the Google Groups "pencil-code-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pencil-code-dis...@googlegroups.com<mailto:pencil-code-dis...@googlegroups.com>.
> --
> You received this message because you are subscribed to the Google Groups "pencil-code-discuss" group.
> To view this discussion visit https://groups.google.com/d/msgid/pencil-code-discuss/ae0d308a54de44c0899a7554af8a6e7f%40su.se<https://groups.google.com/d/msgid/pencil-code-discuss/ae0d308a54de44c0899a7554af8a6e7f%40su.se?utm_medium=email&utm_source=footer>.
Matthias Rheinhardt
Mar 26, 2026, 4:00:28 PM (10 days ago) Mar 26
to pencil-co...@googlegroups.com
Hi Axel,
I'm not questioning either of the plus signs in Uadvect + sqrt(cs^2+vA^2). What I refer to is that Uadvect itself is already a sum, e.g. in a two fluid setup with dust:
velocity(fluid1) + velocity(fluid2) + \sum_i velocity{dust species i)
This can't be meaningful as each of these velocities acts on a different field.
Best,
Matthias
Von: pencil-co...@googlegroups.com <pencil-co...@googlegroups.com> im Auftrag von Axel Brandenburg <axelbra...@gmail.com>
Gesendet: Donnerstag, 26. März 2026 20:44:11
An: pencil-co...@googlegroups.com
Gesendet: Donnerstag, 26. März 2026 20:44:11
An: pencil-co...@googlegroups.com
To view this discussion visit
https://groups.google.com/d/msgid/pencil-code-discuss/acWMizQRT3HMfNCi%40nl6.
Axel Brandenburg
Mar 26, 2026, 4:13:59 PM (10 days ago) Mar 26
to pencil-co...@googlegroups.com
Hi Matthias,
Yes, I agree, but at the moment, we have only a limited number of
categories, between which we take the maxima of the inverse time steps.
The current thing was obviously a hack, which still works in standard
situations when one criterion is much more restrictive than others.
An option to do better may be good, especially if someone really runs
into problems with the current version.
Axel
On Thu, Mar 26, 2026 at 08:00:24PM +0000, 'Matthias Rheinhardt' via pencil-code-discuss wrote:
> Hi Axel,
>
>
> I'm not questioning either of the plus signs in Uadvect + sqrt(cs^2+vA^2). What I refer to is that Uadvect itself is already a sum, e.g. in a two fluid setup with dust:
>
> velocity(fluid1) + velocity(fluid2) + \sum_i velocity{dust species i)
>
> This can't be meaningful as each of these velocities acts on a different field.
>
>
> Best,
>
> Matthias
>
> ________________________________
> To view this discussion visit https://groups.google.com/d/msgid/pencil-code-discuss/acWMizQRT3HMfNCi%40nl6.
Yes, I agree, but at the moment, we have only a limited number of
categories, between which we take the maxima of the inverse time steps.
The current thing was obviously a hack, which still works in standard
situations when one criterion is much more restrictive than others.
An option to do better may be good, especially if someone really runs
into problems with the current version.
Axel
On Thu, Mar 26, 2026 at 08:00:24PM +0000, 'Matthias Rheinhardt' via pencil-code-discuss wrote:
> Hi Axel,
>
>
> I'm not questioning either of the plus signs in Uadvect + sqrt(cs^2+vA^2). What I refer to is that Uadvect itself is already a sum, e.g. in a two fluid setup with dust:
>
> velocity(fluid1) + velocity(fluid2) + \sum_i velocity{dust species i)
>
> This can't be meaningful as each of these velocities acts on a different field.
>
>
>
> Matthias
>
> ________________________________
> Von: pencil-co...@googlegroups.com <pencil-co...@googlegroups.com> im Auftrag von Axel Brandenburg <axelbra...@gmail.com>
> Gesendet: Donnerstag, 26. März 2026 20:44:11
> An: pencil-co...@googlegroups.com
> Gesendet: Donnerstag, 26. März 2026 20:44:11
>
> --
> You received this message because you are subscribed to the Google Groups "pencil-code-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pencil-code-dis...@googlegroups.com<mailto:pencil-code-dis...@googlegroups.com>.
> To view this discussion visit https://groups.google.com/d/msgid/pencil-code-discuss/93ccb41f89e348cfbde8d933d13c93ce%40su.se<https://groups.google.com/d/msgid/pencil-code-discuss/93ccb41f89e348cfbde8d933d13c93ce%40su.se?utm_medium=email&utm_source=footer>.
> --
> You received this message because you are subscribed to the Google Groups "pencil-code-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pencil-code-dis...@googlegroups.com<mailto:pencil-code-dis...@googlegroups.com>.
Matthias Rheinhardt
Mar 26, 2026, 7:15:54 PM (10 days ago) Mar 26
to pencil-co...@googlegroups.com
I think it is less about problems, but more that due to the sum, the timestep becomes unnecessarily small.
Indeed, I could see this for samples where Uadvec matters for the timestep.
M.
Von: pencil-co...@googlegroups.com <pencil-co...@googlegroups.com> im Auftrag von Axel Brandenburg <axelbra...@gmail.com>
Gesendet: Donnerstag, 26. März 2026 21:13:51
To view this discussion visit
https://groups.google.com/d/msgid/pencil-code-discuss/acWTfwq4hTMCgVXZ%40nl6.
Reply all
Reply to author
Forward
0 new messages