Le 25/02/2016 19:00, Kurt Hornik a écrit :
>> Vincent Belaïche writes:
>
> Vincent,
>
> Thanks millions again for looking into this.
>
>
You are welcome, anyway I was probably the one to blame for the bug
:-$ .
It was good to learn about `set -e', and an occasion to resume some
discussion w
On 25 February 2016 at 18:00, Kurt Hornik wrote:
>> Vincent Belaïche writes:
>
> Vincent,
>
> Thanks millions again for looking into this.
>
> We have used the patch you sent first (I guess flavor one) to test
> against our complete R/CRAN build/check process (which includes using
> texi2dvi t
> Vincent Belaïche writes:
Vincent,
Thanks millions again for looking into this.
We have used the patch you sent first (I guess flavor one) to test
against our complete R/CRAN build/check process (which includes using
texi2dvi to build the reference manuals and other LaTeX docs for almost
8
Dear Gavin,
I have attached the patch in two flavors. Please apply your favorite
one. The first flavor is the same as the original patch. And the second
flavor keep the same code but just add some additional `true' just
before the function return in order to force the return status to 0.
The seco
On 23 February 2016 at 23:15, Vincent Belaïche
wrote:
>
> OK, I got it. The problem is in function filter_files. Unfortunately
> when we made the changes to solve the missing \openout in log problem,
> we did not make any tests with not any new xref files after the first
> run. When this happens (
OK, I got it. The problem is in function filter_files. Unfortunately
when we made the changes to solve the missing \openout in log problem,
we did not make any tests with not any new xref files after the first
run. When this happens (ie variable xref_files_new empty) the function
filter_files cras
Ooops... it proved out that I made the trial with a quite old texi2dvi
version. Now I have used a more recent one (not yet checked whether this
is the latest one) and I could reproduce the problem.
I let you know when I understand why there is one single run.
Vincent.
>Hello Kurt,
>
>I think t
Hello Kurt,
I think that this ~ thing is a separate problem. That should be
submitted as a tex2dvi bug, I have tried to do that but the bug
submission is closed on savannah for texinfo project (??).
I have just tried your second test1.tex file, and my texi2dvi does make
two runs, here is the log
> Vincent Belaïche writes:
Ok, so let's try a version without a tilde then:
test1.tex
Description: LaTeX document
It makes no difference for me: running texi2dvi test1.tex ends with
Package rerunfilecheck Warning: File `test1.out' has changed.
(rerunfilecheck)Rerun to get
Hello Kurt,
It seems that some \ref is missing before the reference. Well I have
added it, and the only problem which I can see is that the ~ is handled
like a normal tilde.
But this is because texi2dvi uses this command:
Running pdflatex '\catcode126=12 \def\normaltilde{~}\catcode126=13
\let~\
Hello,
We're facing the same texi2dvi issue in Fedora after update
to Texinfo 6.1.
See:
https://bugzilla.redhat.com/show_bug.cgi?id=1309702
I can also reproduce it with Kurt's latest minimal reproducer
(test1.tex) on F23:
$ texi2dvi --verbose ./test1.tex > test1-log.txt
/usr/bin/texi2dvi: Pr
> Vincent Belaïche writes:
> Dear Kurt,
> Is the issue that texi2dvi does not make enough runs?
Yes.
> Could you send also a minimal example of Rd2.tex, I could then make a
> trial on my machine to see if I can reproduce the issue
Gavin had also asked for a smaller example (the one I h
Dear Kurt,
Is the issue that texi2dvi does not make enough runs?
Could you send also a minimal example of Rd2.tex, I could then make a
trial on my machine to see if I can reproduce the issue
Vincent.
Le 22/02/2016 22:19, Gavin Smith a écrit :
> The log files you sent help a bit. They co
The log files you sent help a bit. They contain lines like:
Package hyperref Warning: Rerun to get /PageLabels entry.
This explains why no "New xref files" messages are being printed by
texi2dvi: the function xref_files_changed in texi2dvi sees that text
in the log files and doesn't look in t
> Gavin Smith writes:
> On 22 February 2016 at 18:00, Kurt Hornik wrote:
>> Sure.
>>
>> I get
>>
>> $ ls Rd2.*
>> Rd2.tex
>>
>> $ ./texi2dvi --verbose Rd2.tex >/dev/null
>> ./texi2dvi: Processing Rd2.tex ...
>> ./texi2dvi: Checking TeX \openout in log support...
>> ./texi2dvi: ./texi2dvi:
On 22 February 2016 at 18:00, Kurt Hornik wrote:
> Sure.
>
> I get
>
> $ ls Rd2.*
> Rd2.tex
>
> $ ./texi2dvi --verbose Rd2.tex >/dev/null
> ./texi2dvi: Processing Rd2.tex ...
> ./texi2dvi: Checking TeX \openout in log support...
> ./texi2dvi: ./texi2dvi: running latex --file-line-error '\input'
>
> Gavin Smith writes:
> On 22 February 2016 at 07:40, Kurt Hornik wrote:
>>> Gavin Smith writes:
>>
>>> On 21 February 2016 at 19:24, Kurt Hornik wrote:
I attach files Rd2.tex and Rd.sty. With these in cwd, running
texi2dvi Rd2.tex
returns 1 with warning
>>
On 22 February 2016 at 07:40, Kurt Hornik wrote:
>> Gavin Smith writes:
>
>> On 21 February 2016 at 19:24, Kurt Hornik wrote:
>>> I attach files Rd2.tex and Rd.sty. With these in cwd, running
>>>
>>> texi2dvi Rd2.tex
>>>
>>> returns 1 with warning
>
>> I tried running this but I got an error
> Gavin Smith writes:
> On 21 February 2016 at 19:24, Kurt Hornik wrote:
>> I attach files Rd2.tex and Rd.sty. With these in cwd, running
>>
>> texi2dvi Rd2.tex
>>
>> returns 1 with warning
> I tried running this but I got an error:
> ! LaTeX Error: File `inconsolata.sty' not found.
> C
On 21 February 2016 at 19:24, Kurt Hornik wrote:
> I attach files Rd2.tex and Rd.sty. With these in cwd, running
>
>texi2dvi Rd2.tex
>
> returns 1 with warning
I tried running this but I got an error:
! LaTeX Error: File `inconsolata.sty' not found.
Can you send that file as well please?
Dear maintainers,
Texinfo 6.1 has introduced a texi2dvi problem that will cause
considerable trouble for R (www.R-project.org), where on Unix texi2dvi
is used for generating pdf from latex and R's internal documentation
format.
I attach files Rd2.tex and Rd.sty. With these in cwd, running
21 matches
Mail list logo
