Hi Knut,
On Wednesday 2019/01/23 12:54 +0100, Knut Petersen wrote:
> fatal error: failed files: "65/lily-bab68f98.ly"
>
> So lilypond fails because gs failed to convert 65/lily-bab68f98.ly.
I reproduced the same error with PRs 53-60 merged on Ubuntu 14.04.
[...]
> Summary up to now:
>
>>> Is dir guaranteed to be a relative path?
>>
>> It seems that there is no guarantee.
>> But, `make check` invokes `output-distance` with relative paths.
>> http://git.savannah.gnu.org/gitweb/?p=lilypond.git;a=blob;f=GNUmakefile.in;h=da1f6f64e088756e07d20f8e1603ccc3a102053e;hb=HEAD#l340
>>
>>
Valentin Villenave writes:
> On 1/24/19, David Kastrup wrote:
>> texi2html(?) does have specific version requirements
>
> Yes, I meant texi2html! We’re in a bind similar to that of Guile: the
> old unmaintained version works and not the new one. (Well, minus the
> outrageous
>
Hello,
2019-01-24 15:37 GMT+00:00 David Kastrup:
> Valentin Villenave writes:
>
>> On 1/24/19, Knut Petersen wrote:
>>> There's the possibility to uses git submodules ... guile, ghostscript
and
>>> extractpdfmark are primary candidates.
>>
>> Not to mention the specific Texinfo version we rely
On 1/24/19, David Kastrup wrote:
> texi2html(?) does have specific version requirements
Yes, I meant texi2html! We’re in a bind similar to that of Guile: the
old unmaintained version works and not the new one. (Well, minus the
outrageous
Valentin Villenave writes:
> On 1/24/19, Knut Petersen wrote:
>> There's the possibility to uses git submodules ... guile, ghostscript and
>> extractpdfmark are primary candidates.
>
> Not to mention the specific Texinfo version we rely on.
texinfo.tex and supplementary files are already part
On 1/24/19, Knut Petersen wrote:
> There's the possibility to uses git submodules ... guile, ghostscript and
> extractpdfmark are primary candidates.
Not to mention the specific Texinfo version we rely on.
V.
___
lilypond-devel mailing list
Masamichi Hosoda writes:
This is a gs 9.26 issue and I cannot see how this might be related to what
we have hit here, so maybe Hosoda-san will be abel to figure why make
check is breaking without extractpdfmark installed.
>>>
>>> If I understand correctly, this patch solves the
>>> This is a gs 9.26 issue and I cannot see how this might be related to what
>>> we have hit here, so maybe Hosoda-san will be abel to figure why make check
>>> is breaking without extractpdfmark installed.
>>
>> If I understand correctly, this patch solves the error.
>>
>> ```
>> ---
Knut Petersen writes:
> Great. I can reproduce the problem.
>
> After that I created branch stabletest. stabletest is stable/2.20 with 3
> additional commits cherry-picked commits from master:
>
>commit 2c7277e0014b8d1d22ef5a1caa69a2f86bcfb964
>commit
On 24.01.19 14:04, David Kastrup wrote:
Masamichi Hosoda writes:
This is a gs 9.26 issue and I cannot see how this might be related to what we
have hit here, so maybe Hosoda-san will be abel to figure why make check is
breaking without extractpdfmark installed.
If I understand correctly,
Masamichi Hosoda writes:
>> This is a gs 9.26 issue and I cannot see how this might be related to what
>> we have hit here, so maybe Hosoda-san will be abel to figure why make check
>> is breaking without extractpdfmark installed.
>
> If I understand correctly, this patch solves the error.
>
>
> This is a gs 9.26 issue and I cannot see how this might be related to what we
> have hit here, so maybe Hosoda-san will be abel to figure why make check is
> breaking without extractpdfmark installed.
If I understand correctly, this patch solves the error.
```
---
On 1/24/2019 5:44 AM, Knut Petersen wrote:
It would be much easier and faster to set up a lilypond build
environment if the tools that are really needed would be submodules
I see. So ideally, people could be starting from scratch, and doing "git
clone --recurse-submodules" would bring in
On 24.01.19 12:21, Karlin High wrote:
On 1/24/2019 4:57 AM, Knut Petersen wrote:
Problem is that we don't have conditional prerequisites of the "install
this if you want to compile that" kind.
There's the possibility to uses git submodules ... guile, ghostscript and
extractpdfmark are
On 1/24/2019 4:57 AM, Knut Petersen wrote:
Problem is that we don't have conditional prerequisites of the "install
this if you want to compile that" kind.
There's the possibility to uses git submodules ... guile, ghostscript
and extractpdfmark are primary candidates.
Git Submodules... this?
On 24.01.19 11:26, David Kastrup wrote:
Therefore I think requiring extractpdfmark just to build LP binary is
a bit overkill.
Problem is that we don't have conditional prerequisites of the "install
this if you want to compile that" kind.
There's the possibility to uses git submodules ...
"James Lowe" writes:
> No one really builds the doc apart from me (test patches) Phil
> (website) and maybe a few hardcore LP users.
I build it regularly, mainly because the HTML documentation can't hold a
candle to the usability of the Info documentation and I need it. It was
one of the
Hello,
2019-01-23 23:22 GMT+00:00 Knut Petersen:
> On 23.01.19 22:50, Thomas Morley wrote:
>> While I agree it has nothing to do with the current problem, it may be
>> an argument against _requirement_ of extractpdfmark
>> Though, I'm not familiar with this stuff, I may be wrong.
>
>
> The gs interpreter should not be 9.26. What files are opened during
> the gs run? [...]
Can you show the environment for this gs run? It seems as if
`.../target/linux-64/root/usr/lib' is missing from LD_LIBRARY_PATH...
> -> We either need to change lilypond's code in a way that PATH is
>
>> Yes, 430fcf895a49d8159413f194488b8474d4ae6be6 is also needed.
>
> Doesn't apply either.
OK, so please add 61d42c83db9abf567da0dcabec50df71d56d2bf9, too.
Werner
___
lilypond-devel mailing list
lilypond-devel@gnu.org
On 23.01.19 22:50, Thomas Morley wrote:
While I agree it has nothing to do with the current problem, it may be
an argument against _requirement_ of extractpdfmark
Though, I'm not familiar with this stuff, I may be wrong.
extractpdfmark saves more than 100 MB of disk space if you only look at
Am Mi., 23. Jan. 2019 um 14:30 Uhr schrieb James Lowe :
>
> Thomas,
>
> On Wed, 23 Jan 2019 10:58:14 +0100, Thomas Morley
> wrote:
>
>
> > I seem to remember Masamichi-san reported some problems with recent gs
> > and extractpdfmark, I can be wrong here, though. Right now I've not
> > the time
Am Mi., 23. Jan. 2019 um 10:58 Uhr schrieb Thomas Morley
:
> I'll redo tests with recent gs-master in the evening.
For completeness, make, make test-baseline, make check succeeds with
gs-9.27 (from most recent master) and extractpdfmark (self-compiled
from the repo)
Cheers,
Harm
Il giorno mer 23 gen 2019 alle 12:54, Knut Petersen
ha scritto:
Have a look at
/home/knut/sources/gub/target/linux-64/build/lilypond-git.sv.gnu.org--lilypond.git-master/out/lybook-testdb/snippet-names--3368173372264547087.log
Werner LEMBERG writes:
>>> -> I suspect that stable/2.20 fails because patch
>>> 2c7277e0014b8d1d22ef5a1caa69a2f86bcfb964 is missing ...
>>
>> Patch does not apply (upon cherry-pick) so I guess that this builds on
>> some previous patch. Would we be needing both then?
>
> Yes,
Thomas,
On Wed, 23 Jan 2019 10:58:14 +0100, Thomas Morley
wrote:
> I seem to remember Masamichi-san reported some problems with recent gs
> and extractpdfmark, I can be wrong here, though. Right now I've not
> the time to research. cc-ing him
>
> I'll redo tests with recent gs-master in the
>> -> I suspect that stable/2.20 fails because patch
>> 2c7277e0014b8d1d22ef5a1caa69a2f86bcfb964 is missing ...
>
> Patch does not apply (upon cherry-pick) so I guess that this builds on
> some previous patch. Would we be needing both then?
Yes, 430fcf895a49d8159413f194488b8474d4ae6be6 is also
Knut Petersen writes:
> ===
> SUMMARY
> ===
>
> We have two problems.
>
> -> I suspect that stable/2.20 fails because patch
> 2c7277e0014b8d1d22ef5a1caa69a2f86bcfb964 is missing ...
Patch does not apply (upon cherry-pick) so I guess that this builds on
some previous patch.
Hi everybody!
Below an analysis of lilypond-test problems. strace still is our friend ;-).
Lets start investigation ...
knut@golem:~/sources/gub> rm -f STRACE/* ; strace -v -f -ff -s 1024 -o
STRACE/TP bin/gub linux-64::lilypond-test
calculating
Am Mi., 23. Jan. 2019 um 10:36 Uhr schrieb James Lowe :
> This morning I tried on my work system with extractpdfmark and gs 2.6 and I
> could get a make check to work. Without it (but still using gs 2.6) make
> check will fail.
>
> So the issue is when extractpdfmark is not installed.
Jut
Hello,
On Tue, 22 Jan 2019 21:39:13 +, pkx1...@runbox.com wrote:
> Hello,
>
> On 22/01/2019 21:30, Thomas Morley wrote:
> > Am Di., 22. Jan. 2019 um 08:24 Uhr schrieb Werner LEMBERG :
> >
> >> The last command present in both my and your log file is
> >>
> >>gs -sDEVICE=png16m \
> >>
Hello,
On 22/01/2019 21:30, Thomas Morley wrote:
Am Di., 22. Jan. 2019 um 08:24 Uhr schrieb Werner LEMBERG :
The last command present in both my and your log file is
gs -sDEVICE=png16m \
-dGraphicsAlphaBits=4 \
-dTextAlphaBits=4 \
Am Di., 22. Jan. 2019 um 08:24 Uhr schrieb Werner LEMBERG :
> The last command present in both my and your log file is
>
> gs -sDEVICE=png16m \
> -dGraphicsAlphaBits=4 \
> -dTextAlphaBits=4 \
>
> -slilypond-datadir=input/regression/out-test-baseline/share/lilypond/current \
>
Hello,
Sorry to top post but I have had the same error as Thomas using my 'work'
instance of Ubuntu 16.04 - at home I also use 16.04.
I try to keep the two 'OSes' the same but recently I had to reinstall my 'work'
instance as new and now, even after applying the latest git master, I get the
Hello,
On 22/01/2019 14:40, James Lowe wrote:
Hello,
Sorry to top post but I have had the same error as Thomas using my 'work'
instance of Ubuntu 16.04 - at home I also use 16.04.
I try to keep the two 'OSes' the same but recently I had to reinstall my 'work'
instance as new and now, even
Am Di., 22. Jan. 2019 um 08:24 Uhr schrieb Werner LEMBERG :
>
> >> Can you repeat the whole thing without any options for parallel
> >> computing and show us the output of
> >>
> >> make check &> make.check.log &
> >>
> >> please? I could then compare it directly with my log file.
> >
> > I
>> Can you repeat the whole thing without any options for parallel
>> computing and show us the output of
>>
>> make check &> make.check.log &
>>
>> please? I could then compare it directly with my log file.
>
> I downgraded ghostscript to 9.26, still failing. So likely not a
> gs-problem.
>
>
>> > Failed with:
>> >
>> > Error: /undefinedfilename in --file--
>> > Operand stack:
>> >
>> > (input/regression/out-test-baseline/share/lilypond/current/fonts/otf/emmentaler-20.otf)
>>
>> I can't reproduce that. I followed exactly your recipe (but
>> omitting `-j3 CPU_COUNT=3'), and it
Am So., 20. Jan. 2019 um 22:49 Uhr schrieb Werner LEMBERG :
>
> > I did from scratch:
> >
> > rm -fr build/
> > sh autogen.sh --noconfigure
> > mkdir -p build/
> > cd build/
> > ../configure
> > gitk
> > git status
> > make -j3 CPU_COUNT=3
> > make -j3 CPU_COUNT=3 test-baseline
> > make -j3
> I did from scratch:
>
> rm -fr build/
> sh autogen.sh --noconfigure
> mkdir -p build/
> cd build/
> ../configure
> gitk
> git status
> make -j3 CPU_COUNT=3
> make -j3 CPU_COUNT=3 test-baseline
> make -j3 CPU_COUNT=3 check
>
>
> Failed with:
>
> Error: /undefinedfilename in --file--
> Operand
Hi Knut,
On Sun 2019-01-20 11:17 +0100, Knut Petersen wrote:
> # Now build stable
>
> knut@golem:~/sources/gub> time make LILYPOND_BRANCH=stable/2.20
> lilypond
>
> # Unfortunately that fails with
[...]
> mkdir /home/knut/sources/gub/uploads/webtest/v2.21.0-
>
Am So., 20. Jan. 2019 um 12:07 Uhr schrieb :
>
> Hello,
>
> On 20/01/2019 10:17, Knut Petersen wrote:
> > Hi Werner
> >>> I don't know what the difference is between the two 'make' functions
> >>> but at the moment I cannot do a full set of tests on patches.
> >> Whatever :-) I found a problem
Hello,
On 20/01/2019 10:17, Knut Petersen wrote:
Hi Werner
I don't know what the difference is between the two 'make' functions
but at the moment I cannot do a full set of tests on patches.
Whatever :-) I found a problem with `output-distance.py', which I
hot-fixed in staging. Please retry!
Hi Werner
I don't know what the difference is between the two 'make' functions
but at the moment I cannot do a full set of tests on patches.
Whatever :-) I found a problem with `output-distance.py', which I
hot-fixed in staging. Please retry!
output-distance.py ... I think my system hit the
> I've been struggling today to 'make check' as part of the normal set
> of commands I use to patch test.
>
> I've gone back, one by one, and this is the commit that 'breaks' my
> patch testing
>
> Issue 5450: relocate.cc: Introduce new command `set?'
> commit
Werner,
I've been struggling today to 'make check' as part of the normal set of
commands I use to patch test.
I've gone back, one by one, and this is the commit that 'breaks' my
patch testing
--
Issue 5450: relocate.cc: Introduce new command `set?'
author Werner Lemberg
Fri, 21
47 matches
Mail list logo
