Re: broken doc build (orchestra.ly)

2016-06-04 Thread Knut Petersen

"make doc" succeeds again, at least here.

[newer versions]: make doc succeeds
dbaf1e56e37be0e204231c5bf1adcb14bd8ac3b8: fixes "make doc" failure here
4b4a2cae9953cff69159f10763e990f5e4265ddd: lilypond does not build
cccd7f9574c72ccecca00641d7a5ea3e8ce99cfd: still broken "make doc"
[...]
c1d7bc2217462a63bf5c5c6d6f6df5cb00099180: first broken "make doc"
[older versions]

cu,
 Knut

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Add ly:get-font-format to get the font format (issue 296350043 by truer...@gmail.com)

2016-06-04 Thread lemzwerg

My main purpose was the distinction between the TTC and OTC.
So this patch cannot distinguish between the PFA and PFB.


OK, but IMHO we should get rid of being dependent on the file
name suffix.


I'm trying the following `font-file-as-ps-string' with this
Patch Set 2.


Hmm, this looks like the original, unpatched code...


https://codereview.appspot.com/296350043/

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Issue 3945: fix (De)crescendo with unspecified starting volume in MIDI (issue 294700043 by nine.fierce.ball...@gmail.com)

2016-06-04 Thread nine . fierce . ballads

https://codereview.appspot.com/294700043/

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: make doc still fails - problem with lilypond-book - was: Still cannot make doc

2016-06-04 Thread Thomas Morley
2016-06-04 15:12 GMT+02:00 Masamichi Hosoda :
>> In addition, the following command sequence succeeds:
>>
>> lilypond-book aaa-lilybook-test.lytex
>> latex aaa-lilybook-test.tex
>> dvips aaa-lilybook-test.dvi
>> ps2pdf aaa-lilybook-test.ps
>> evince aaa-lilybook-test.pdf
>>
>> aaa-lilybook-test.ps is >16MB again, though.
>
> Thank you for your results.
>
> If I understand correctly, it's a ghostscript issue.
>
> Ghostscript might not be able to handle IPAex fonts that you used.
>
> Using lilypond-book with `--pdf' option invokes lilypond
> with `-dgs-load-fonts' option.
> LilyPond with `-dgs-load-fonts' option generates ps file
> which is embedded font filename only.
> Without the option, the ps file is embedded font glyphs.
>
> If the glyphs are embedded, ghostscript works fine.
> However, if filename are embedded, ghostscript may not work fine.
>
> In the case of IPAex fonts,
> embedded filename brings the ghostscript error.
> But, for IPA fonts which are another Japanese fonts,
> embedded filename works fine.
>
> The GUB environment uses IPA fonts instead of IPAex fonts.
> So there is no problem.

Thanks a lot for your research and findings!

> Would you try the following commands?
>
> $ sudo apt-get install fonts-ipafont-mincho
> $ sudo apt-get install fonts-ipafont-gothic
>
> IPA fonts are `IPA Mincho' (Japanese serif font)
> and `IPA Gothic' (Japanese sans serif font).
> By default, fontconfig takes precedence over the IPA fonts than IPAex fonts.

As of
http://www.mail-archive.com/bug-lilypond@gnu.org/msg40846.html
I had uninstalled Noto and instead installed IPA fonts _and_ IPAex fonts.
After uninstalling IPAex fonts - first the test-case worked and now I
have success on a full `make doc'.

Any chance to get a more meaningful error-message?
Or is there any criterion we could select such problematic fonts?


Again many thanks,
  Harm

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: make doc still fails - problem with lilypond-book - was: Still cannot make doc

2016-06-04 Thread Masamichi Hosoda
> In addition, the following command sequence succeeds:
> 
> lilypond-book aaa-lilybook-test.lytex
> latex aaa-lilybook-test.tex
> dvips aaa-lilybook-test.dvi
> ps2pdf aaa-lilybook-test.ps
> evince aaa-lilybook-test.pdf
> 
> aaa-lilybook-test.ps is >16MB again, though.

Thank you for your results.

If I understand correctly, it's a ghostscript issue.

Ghostscript might not be able to handle IPAex fonts that you used.

Using lilypond-book with `--pdf' option invokes lilypond
with `-dgs-load-fonts' option.
LilyPond with `-dgs-load-fonts' option generates ps file
which is embedded font filename only.
Without the option, the ps file is embedded font glyphs.

If the glyphs are embedded, ghostscript works fine.
However, if filename are embedded, ghostscript may not work fine.

In the case of IPAex fonts,
embedded filename brings the ghostscript error.
But, for IPA fonts which are another Japanese fonts,
embedded filename works fine.

The GUB environment uses IPA fonts instead of IPAex fonts.
So there is no problem.

Would you try the following commands?

$ sudo apt-get install fonts-ipafont-mincho
$ sudo apt-get install fonts-ipafont-gothic

IPA fonts are `IPA Mincho' (Japanese serif font)
and `IPA Gothic' (Japanese sans serif font).
By default, fontconfig takes precedence over the IPA fonts than IPAex fonts.

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Add ly:get-font-format to get the font format (issue 296350043 by truer...@gmail.com)

2016-06-04 Thread trueroad

LGTM, but note that FreeType's `FT_Get_Font_Format` doesn't really

return the

font format but the name of the module used to handle a font.  In

particular, it

doesn't make a difference between PFA and PFB.


My main purpose was the distinction between the TTC and OTC.
So this patch cannot distinguish between the PFA and PFB.


Such a function could be then used to rewrite `font-file-as-ps-string`

to avoid

any dependency on the file extension.


I'm trying the following `font-file-as-ps-string' with this Patch Set 2.

```
  (define (font-file-as-ps-string name file-name font-index)
(let* ((downcase-file-name (string-downcase file-name)))
  (cond
   ((and file-name (string-endswith downcase-file-name ".pfa"))
(embed-document file-name))
   ((and file-name (string-endswith downcase-file-name ".pfb"))
(ly:pfb->pfa file-name))
   ((and file-name (string-endswith downcase-file-name ".ttf"))
(ly:ttf->pfa file-name))
   ((and file-name (string-endswith downcase-file-name ".ttc"))
;; TODO: distinguish files which have extension `*.ttc'
;; whether TrueType Collection (TTC) fonts
;; or OpenType/CFF Collection (OTC) fonts.
(ly:ttf->pfa file-name font-index)) ;; TTC fonts
   ((and file-name (string-endswith downcase-file-name ".otf"))
(ps-embed-cff (ly:otf->cff file-name) name 0))
   ((and file-name (string-endswith downcase-file-name ".otc"))
;; The files which have the extension `*.otc' are OTC fonts.
(ps-embed-cff (ly:otf->cff file-name font-index) name 0)) ;; OTC
fonts
   (else
(ly:warning (_ "do not know how to embed ~S=~S") name file-name)
""
```


Wouldn't it be better to return a symbol instead?


I've uploaded Patch Set 2 that uses a symbol instead of a string.

https://codereview.appspot.com/296350043/

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel