traffic:lily usage
[[traffic:lily_usage]] last edit on
Sep 10, 2005
3:24 PM
by musicovore
topmenutraffic:topmenu
(syntax, encoding, requirements, output, ...)
Archive link Wrong characters with jEdit
Date 30 Aug - 6 Sep
About 20 messages.
LilyPond versions up to 2.4 used Latin1 text encoding. This has changed since LilyPond 2.5 (quoting http://lilypond.org/doc/v2.6/Documentation/topdocs/out-www/NEWS.html):
"LilyPond now uses Pango and FontConfig for selecting and rendering UTF-8 input in non-TeX backends."
This means that text markup have to be encoded in UTF-8 to be rendered in correctly, since the TeX backend is deprecated and no more supported.
This discussion shows how difficult the transition between Latin1 and UTF-8 can be, as Tom "stk" wrote:
"I have used Latin1 character encoding for the last 15 years for handling
text in English, French, Spanish, and German. My (Unix) e-mail
client uses Latin1 (ISO-8859-1). I use a Latin1 text editor for LilyPond
and thus avoid the cursed false-single-quote problem, and I do not want to
incur the hazards of unicode character-encoding.
[...]
There is considerable coding to enable Latin1 in the .ly and .scm files in
the LilyPond distribution; how can this coding be made to actually
function?"
Mats Bengtsson replied:
"What you have found in the source code files are some left-overs
from version 2.4 and earlier, where LilyPond only knew about Latin1.
If you browse through the mailing list archives, you can also find
out why this was not a satisfactory solution if you want to promote
the program outside western Europe.
If you find this added flexibility in LilyPond 2.6 so annoying, then
you could use some program that converts a Latin1 coded file into
UTF-8 coding and even make a script file that first does the
conversion and then calls LilyPond. Unfortunately, I don't know
Windows well enough to provide any specific hints but there should
be several possibilities available."
Tom "stk" replied:
"But I still think LilyPond should allow the option of putting a command
at the top of a *.LY file marking it as either Latin-1 or unicode. All
the code for allowing Latin-1 exists; it was used in LP 2.4 according to
Mats. There are plenty of computer users in North America who take it for
granted that all Windows software can understand Latin-1."
...and added in another message:
"(1) That's certainly one huge mass of "left-overs". For a language
that allows coloured labelled noteheads, Gregorian notation, tenor
clef, and Lord knows what else, it is amazing that the developers
would choose to *remove* an existing capability.
(2) Deleting a capability is not "promotion". Making unicode an available
people in Western Europe and North America whose other software all
understands Latin-1. (As a side remark, when I communicate with people in
Romania or other Eastern European locations, I have to use Latin-2;
software [mainly dictionaries] downloadable from those locations uses
Latin-2, *not* unicode. Unicode may or may not become predominant
someday, but it certainly is not now.)"
Han-Wen Nienhuys replied:
"A. LilyPond actually _does_ support the Latin1 character set, as Latin1
and Unicode coincide on the first 256 codepoints.
B. LilyPond does not support Latin1 encoding. This is because
1. It's not possible to detect the encoding of a file. Supporting
alternate encodings implies that users have to specify the encoding via
the command line. This is error-prone, and leads to confusion for newbies.
2. If we do latin1, why should we not do latin2. And if we do latin1
and 2, why not Big5? EBCDIC? UTF-16? tibetan-iso-8bit? Where does it
stop?
C. Unicode, not Latin1, is the future. Using UTF-8 gives us a much
better chance of catching that half-billion in the future, as well as
the 4.5 billion who don't use latin1 today."
Read the (long) thread for further replies.
- JM
Archive link http://thread.gmane.org/gmane.comp.gnu.lilypond.general/17329
Date 8 Sep
3 messages.
LilyPond from 2.6 has officially a "fully fonctionnal SVG backend".
Daniel Johnson tested SVG files generated by LilyPond 2.6.3 and 2.7.7 with a lot of viewers and editors.
There seems to be problems with renderers, especially for multi-page output.
Han-Wen Nienhuys reply explains the problem:
"We have tested lily with inkscape. As inkscape only supports single
page-documents, we don't know how multi-page documents should behave.
The question is also who is correct. If the SVG spec was unequivocal,
all SVG renderers would produce the same image. It's the question which
renderer is correct."
- JM
copyright
LilyPond usage
Here is information about LilyPond usage, especially changes between versions(syntax, encoding, requirements, output, ...)
Input
traffic:sep05user1Character encoding of LilyPond input files: Latin1 against UTF-8
Section LilyPond usageArchive link Wrong characters with jEdit
Date 30 Aug - 6 Sep
About 20 messages.
LilyPond versions up to 2.4 used Latin1 text encoding. This has changed since LilyPond 2.5 (quoting http://lilypond.org/doc/v2.6/Documentation/topdocs/out-www/NEWS.html):
"LilyPond now uses Pango and FontConfig for selecting and rendering UTF-8 input in non-TeX backends."
This means that text markup have to be encoded in UTF-8 to be rendered in correctly, since the TeX backend is deprecated and no more supported.
This discussion shows how difficult the transition between Latin1 and UTF-8 can be, as Tom "stk" wrote:
"I have used Latin1 character encoding for the last 15 years for handling
text in English, French, Spanish, and German. My (Unix) e-mail
client uses Latin1 (ISO-8859-1). I use a Latin1 text editor for LilyPond
and thus avoid the cursed false-single-quote problem, and I do not want to
incur the hazards of unicode character-encoding.
[...]
There is considerable coding to enable Latin1 in the .ly and .scm files in
the LilyPond distribution; how can this coding be made to actually
function?"
Mats Bengtsson replied:
"What you have found in the source code files are some left-overs
from version 2.4 and earlier, where LilyPond only knew about Latin1.
If you browse through the mailing list archives, you can also find
out why this was not a satisfactory solution if you want to promote
the program outside western Europe.
If you find this added flexibility in LilyPond 2.6 so annoying, then
you could use some program that converts a Latin1 coded file into
UTF-8 coding and even make a script file that first does the
conversion and then calls LilyPond. Unfortunately, I don't know
Windows well enough to provide any specific hints but there should
be several possibilities available."
Tom "stk" replied:
"But I still think LilyPond should allow the option of putting a command
at the top of a *.LY file marking it as either Latin-1 or unicode. All
the code for allowing Latin-1 exists; it was used in LP 2.4 according to
Mats. There are plenty of computer users in North America who take it for
granted that all Windows software can understand Latin-1."
...and added in another message:
"(1) That's certainly one huge mass of "left-overs". For a language
that allows coloured labelled noteheads, Gregorian notation, tenor
clef, and Lord knows what else, it is amazing that the developers
would choose to *remove* an existing capability.
(2) Deleting a capability is not "promotion". Making unicode an available
- option* would be promotion, but deleting the existing ability to use
people in Western Europe and North America whose other software all
understands Latin-1. (As a side remark, when I communicate with people in
Romania or other Eastern European locations, I have to use Latin-2;
software [mainly dictionaries] downloadable from those locations uses
Latin-2, *not* unicode. Unicode may or may not become predominant
someday, but it certainly is not now.)"
Han-Wen Nienhuys replied:
"A. LilyPond actually _does_ support the Latin1 character set, as Latin1
and Unicode coincide on the first 256 codepoints.
B. LilyPond does not support Latin1 encoding. This is because
1. It's not possible to detect the encoding of a file. Supporting
alternate encodings implies that users have to specify the encoding via
the command line. This is error-prone, and leads to confusion for newbies.
2. If we do latin1, why should we not do latin2. And if we do latin1
and 2, why not Big5? EBCDIC? UTF-16? tibetan-iso-8bit? Where does it
stop?
C. Unicode, not Latin1, is the future. Using UTF-8 gives us a much
better chance of catching that half-billion in the future, as well as
the 4.5 billion who don't use latin1 today."
Read the (long) thread for further replies.
- JM
Output
traffic:sep05user2SVG backend - resulting files invalid?
Section lily_usageArchive link http://thread.gmane.org/gmane.comp.gnu.lilypond.general/17329
Date 8 Sep
3 messages.
LilyPond from 2.6 has officially a "fully fonctionnal SVG backend".
Daniel Johnson tested SVG files generated by LilyPond 2.6.3 and 2.7.7 with a lot of viewers and editors.
There seems to be problems with renderers, especially for multi-page output.
Han-Wen Nienhuys reply explains the problem:
"We have tested lily with inkscape. As inkscape only supports single
page-documents, we don't know how multi-page documents should behave.
The question is also who is correct. If the SVG spec was unequivocal,
all SVG renderers would produce the same image. It's the question which
renderer is correct."
- JM
copyright
Contents of this Wiki is protected by the Creative Commons Attribution-NonCommercial-ShareAlike 2.0 license