Re: text in staff and markup text font=addlyric font

[Robin Bannister]

Jay Hamilton wrote: 
not exactly sure what that means in this case 


There is an example at 
http://lsr.dsi.unimi.it/LSR/Item?id=258
This sidesteps the question of horizontal space by setting ragged-right. 

Do you already have enough horizontal space for the text? 
Or do you expect Lilypond to make room for it? 


Cheers,
Robin


___
lilypond-user mailing list
lilypond-user@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-user

Re: weblily: security risk

[Bertalan Fodor (LilyPondTool)]

Note that there are existing effort on creating a service that can be 
used to render lilypond scores in Google Wave or in any web application. 
You can get its code at 
http://code.google.com/p/lilypondy/source/browse/#svn/trunk/lilywaveservlet


You can see some good ideas about request queue, caching etc.

Bert
___
lilypond-user mailing list
lilypond-user@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-user

Re: dynamic and midi velocity

[David Raleigh Arnold]

On Saturday 06 March 2010 03:12:11 David Kastrup wrote:
> David Raleigh Arnold  writes:
> > Ideally the volume should be preset for each instrument according to
> > ratios expressing the relative natural loudness of the instrument, IOW
> > set it and forget it, and only velocity should be used for dynamics.
> > That is obviously the purpose of the design.  Why fight it?  No good
> > can come from twiddling the volume in the middle of the performance.
> 
> I have a (not quite cheap) Ketron Keyboard with a volume pedal.  I was
> rather surprised to realize that putting it to zero would
> a) considerably lower the volume of notes already hit
> b) completely stop new notes from being sounded
> 
> I would have thought that it would act more or less as a simple analog
> volume control, but it would appear that it _partially_ affects notes
> already hit.
> 
> Interesting.  I have not checked the corresponding Midi messages,
> though.
> 

Ha!  The volumes should be part of the instrument specifications for
any lilypond/midi document, and dynamics should relate to velocity
alone.  To do otherwise was a major mistake in the midi implementation in
lilypond.  It would be a very good idea to quit trying to make
it work in some other way.

Of course the "solution" to the "problem" is to instantiate a new
voice, instrument and volume for every dynamic.  Surely a lot
of trouble for nothing.  Regards, daveA


___
lilypond-user mailing list
lilypond-user@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-user

Re: dynamic and midi velocity

[Hugh Myers]

Go right a head and implement your suggestion--- Lilypond is open
source you know...

---hsm

On Thu, Mar 11, 2010 at 8:38 AM, David Raleigh Arnold
 wrote:
> On Saturday 06 March 2010 03:12:11 David Kastrup wrote:
>> David Raleigh Arnold  writes:
>> > Ideally the volume should be preset for each instrument according to
>> > ratios expressing the relative natural loudness of the instrument, IOW
>> > set it and forget it, and only velocity should be used for dynamics.
>> > That is obviously the purpose of the design.  Why fight it?  No good
>> > can come from twiddling the volume in the middle of the performance.
>>
>> I have a (not quite cheap) Ketron Keyboard with a volume pedal.  I was
>> rather surprised to realize that putting it to zero would
>> a) considerably lower the volume of notes already hit
>> b) completely stop new notes from being sounded
>>
>> I would have thought that it would act more or less as a simple analog
>> volume control, but it would appear that it _partially_ affects notes
>> already hit.
>>
>> Interesting.  I have not checked the corresponding Midi messages,
>> though.
>>
>
> Ha!  The volumes should be part of the instrument specifications for
> any lilypond/midi document, and dynamics should relate to velocity
> alone.  To do otherwise was a major mistake in the midi implementation in
> lilypond.  It would be a very good idea to quit trying to make
> it work in some other way.
>
> Of course the "solution" to the "problem" is to instantiate a new
> voice, instrument and volume for every dynamic.  Surely a lot
> of trouble for nothing.  Regards, daveA
>
>
> ___
> lilypond-user mailing list
> lilypond-user@gnu.org
> http://lists.gnu.org/mailman/listinfo/lilypond-user
>


___
lilypond-user mailing list
lilypond-user@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-user

Re: weblily: security risk

[Han-Wen Nienhuys]

I have poked around a bit, and could not find obvious holes, but you
are exposing the unix system, so it is a bit scary.  For example, if
there are local exploits in the kernel, this system makes it a remote
explote for weblily,and the fact that uname is available (telling me
you are on 2.6.31.ec2.301) makes this easier

Also, the scheme system is complete, so your scripts may still do
other nasty things:

- start a webservers (perhaps not useful if your system is behind a firewall)
- do a fork-bomb  (through the primitive-fork function), crashing your machine
- delete files under /usr/share/fonts, stopping weblily from working
- start a program that connects to other machines from weblily

I assume that this machine is not dedicated for weblily, so the last
point means that the firewall for inbound connections is also pierced.

For standard malware purposes the scheme route is too weird for
automated attacks, but this gives attackers so much to work with that
the tiniest weakness will turn into a remote compromise.  Also, the
opportunity for vandalism is big; while taking down your server may
not leak your bank credentials, it will cause you headaches.

In short, I dont think this is a very good idea - I recommend using
--safe in addition to the jail, also because it will help us improve
the --safe mode, which is underused currently.


On Thu, Mar 11, 2010 at 9:17 AM, Weblily  wrote:
> Hi Graham,
>
> thank your for sharing your thoughts about weblily.net. Of cource, security
> is a concern I have on my mind and I'd be happy to get into discussion with
> you and other knowledgable people on security issues. And I will do my very
> best notto fall prey to all those evil people out there, granted.Though I
> must confess, the tonality of your first e-mail did not really sound very
> inviting. But let's forget about that.
>
> Please take into consideration: I am not a specialist on computer security
> and rather (sorry for that) only an enthusiast working with some fervour on
> his little project. But I'll happily listen to any good advice.
>
> I would really like to know about security problems on weblily.net and would
> definitely work on overcoming them with the limited means I have.
>
> To give some facts:
> * weblily.net uses Liferay als portal software
> * the editor runs as Liferay Web Content on the page
> http://weblily.net/web/guest/runlilypond
> * the editor is a Google GWT application, i.e. JavaScript using GWT RPC to
> communicate with a java servlet hosted on the same Tomcat Liferay is running
> on
> * the servlet runs as user tomcat
> * LilyPond is called from the java servlet using the --jail=lily,lily,...
> option, i.e. runs as user lily
> * Hopefully the user lily has write permission only for the
> /homel/lily/scores/ and the /tmp directory in the jail
> * /home/lily/scores is visible as http://weblily.net/app/scores
> * /tmp should not be visible from the outside at all, but who knows?
> * other paths, like the permalink and template directories are not visible
> within the jail
>
> A problem I do have are crashes and infinite loops of LilyPond. After 5 such
> events you will get a "Server overloaded" message and it will take about 30
> seconds before LilyPond will be running again. Of course, if too many people
> are working simultaneously on weblily.net this will result in the very same
> message.
>
> Another message you might occasionally see is "Engraver error", this is a
> nice way of saying: "Servlet crashed" .
>
> You, Jan and Han-Wen are invited to play around with weblily.net and to
> explore potential weaknesses as long as you will inform me as the  first
> person about problems you see and as logn as you will give me a chance of
> fixing it before you go public. And of course, your advice on how these
> problmes might be resolved is always welcome.
>
> In the hope of providing a useful service to the LilyPond community,
>
> Johannes
>
> PS: I am currently preparing an article about weblily.net for the LilyPond
> Report. Maybe this can be a starting point for discussing ideas about how
> weblily.net might become a useful tool for the LilyPond community.
>
> Am 11.03.2010 02:07, schrieb Graham Percival:
>>
>> I apologize for this email; I jumped to a false conclusion and
>> made a baseless accusation.  I now have no reason to believe that
>> weblily poses a risk.
>>
>> I'm sorry.
>>
>> - Graham Percival
>>
>>
>> On Wed, Mar 10, 2010 at 08:21:24PM +, Graham Percival wrote:
>>
>>>
>>> Mr. Weblily,
>>>
>>> I like your enthusiasm with your weblily project, but for Mao's
>>> sake please learn something about computer security.  The current
>>> website is completely insecure.
>>>
>>> This is not a theoretical concern.  It would take me approximately
>>> two minutes to delete everything in your /home/lily/ directory --
>>> not just material in /home/lily/scores/.
>>>
>>>
>>> I wouldn't do this, of course -- but if a non-expert like me could
>>> do this so quickly, I'm certain that an 

Re: Partials and upbeats

[Carl Sorensen]

On 3/10/10 10:04 AM, "Phil Holmes"  wrote:

> Mats said:
>>> Quoting Phil Holmes :
 Is it correct that the \partial command applies to all voices below the
 one in which it is issued,
> 
>>> No, the \partial command sets some properties on the Score level, so the
>>> order shouldn't matter. Hard to say why you see such a difference
>> without having access to your full test file.
> 
> This was from a while back, but I've now created some snippets that show the
> effect of the order on the reported barcheck error.
> 

 
> If I now modify the file to move the \partial to the first Staff/Voice, I
> get the following error:
> 
>  warning: barcheck failed at: -1/4
>   R4   |  % 1



>>> 
> 
> So it appears from this that it is important which voice or stave has the
> \partial command...
> 

Apparently so.  I suspect (but haven't proven) that this is true because
bar-checking is performed in an iterator outside the

Personally, I think that the warning is "technically" correct, since R
indicates a full-measure rest, and a full measure rest should start at
measure-position zero.

However, since we need to use full-measure rests to enable the empty staves
to be removed, it would seem that a partial measure is also a "full measure"
as far as full-measure rests are concerned.

But for the current status of bar-checks, putting a full-measure rest in a
partial measure should always produce the barcheck warning, IMO.

Thanks,

Carl





___
lilypond-user mailing list
lilypond-user@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-user

Re: Line breaks and text positioning

[Graham Percival]

Read the Learning Manual, "Less tweaks with longer processing" or
something like that.  It's at the end of the manual.

In the 2.13 docs, this section might have moved to Usage.

Cheers,
- Graham

On Thu, Mar 11, 2010 at 05:34:51PM -, Phil Holmes wrote:
> There is what looks like a problem in how LilyPond calculates line breaks 
> when there is some text to display.  The example I've attached shows 
> markup but I know it also applies to Tempo indications.  Is this a known 
> bug (I have searched and can't find it)?
>
> I know it can be worked around with \break, but as a general rule it  
> shouldn't be necessary to do this.
>
> --
> Phil Holmes
>


> \version "2.12.3"
> #(set-global-staff-size 17)
> 
> {
>  \clef "treble"
>  \time 4/4
>  b'4 b'4 b'4 b'4  |  % 1
>  b'4 b'4 b'4 b'4  |  % 2
>  b'4 b'4 b'4 b'4  |  % 3
>  b'4 b'4 b'4 b'4  |  % 4
>  b'4 b'4 b'4 b'4  |  % 5
>  b'4 b'4 b'4 b'4  |  % 6
>  b'4 b'4 b'4 b'4  |  % 7
>  % \break
>  b'4 b'4 b'4 b'4^\markup {\normalsize \italic "Some text to test linebreaking 
> calculation"}   |  % 8
>  b'4 b'4 b'4 b'4  |  % 9
>  b'4 b'4 b'4 b'4  |  % 10
>  b'4 b'4 b'4 b'4  |  % 11
>  b'4 b'4 b'4 b'4  |  % 12
>  b'4 b'4 b'4 b'4  |  % 13
>  b'4 b'4 b'4 b'4  |  % 14
>  b'4 b'4 b'4 b'4  |  % 15
>  b'4 b'4 b'4 b'4  |  % 16
>  b'4 b'4 b'4 b'4  |  % 17
>  b'4 b'4 b'4 b'4 \bar "|."
> }

> ___
> lilypond-user mailing list
> lilypond-user@gnu.org
> http://lists.gnu.org/mailman/listinfo/lilypond-user



___
lilypond-user mailing list
lilypond-user@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-user

Re: your MultiMeasureRest utility

[Alexander Kobel]

me - stanford wrote:

Alexender,

*Here's some more feedback for you *(whenever you get around to looking 
at this code again).  I've attached all the relevant files:


%=

I'm not sure if you remember the barcheck issue I was getting?  It 
looked like this:


* Bach_Singet_dem_Herrn_source_music.ly:85:23: warning: barcheck failed 
at: -1/4 *

* ** \mark \default  *
*R2.*15 ** ** % N = 137 *


Hi, Randy,

it seems like this was kinda solved without me having a look at it at 
all.  In a recent message on lilypond-user (the "Partials and upbeats" 
thread), Carl Sorensen says:

But for the current status of bar-checks, putting a full-measure rest in a
partial measure should always produce the barcheck warning, IMO.


So it looks like the problem is not inherent to my code, though of 
course the method used there is quite ugly and hack-ish, and it breaks 
if you try to use anything but MultiMeasureRests in the BarNumberStaff 
without manual tweaking of the global counter variables.
But in essence: One should not bother to care about the warnings as long 
as it works.



Cheers,
Alexander


___
lilypond-user mailing list
lilypond-user@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-user

Re: Line breaks and text positioning

[James Bailey]

On 11.03.2010, at 18:34, Phil Holmes wrote:

There is what looks like a problem in how LilyPond calculates line  
breaks when there is some text to display.  The example I've  
attached shows markup but I know it also applies to Tempo  
indications.  Is this a known bug (I have searched and can't find it)?


I know it can be worked around with \break, but as a general rule  
it shouldn't be necessary to do this.


--
Phil Holmes

_ 
__

lilypond-user mailing list
lilypond-user@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-user
There's a function to have lilypond respect margins with regards to  
text. It's off by default.



___
lilypond-user mailing list
lilypond-user@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-user

Re: Partials and upbeats

[Graham Percival]

On Thu, Mar 11, 2010 at 10:58:13AM -0700, Carl Sorensen wrote:
> 
> On 3/10/10 10:04 AM, "Phil Holmes"  wrote:
> 
> > If I now modify the file to move the \partial to the first Staff/Voice, I
> > get the following error:
> > 
> >  warning: barcheck failed at: -1/4
> >   R4   |  % 1

IIRC I added this to the tracker 3 years ago.

> Personally, I think that the warning is "technically" correct, since R
> indicates a full-measure rest, and a full measure rest should start at
> measure-position zero.

We could argue about what the desired program output should be (I
don't like the current behavior), but since nobody worked on this
issue in the past 3 years, I see no benefit to arguing about what
it should do now.

Cheers,
- Graham


___
lilypond-user mailing list
lilypond-user@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-user

Re: Line breaks and text positioning

[Kieren MacMillan]

Hi Phil,

> I know it can be worked around with \break, but as a general rule it 
> shouldn't be necessary to do this.

Here's a snippet to consider:

\version "2.12.3"
#(set-global-staff-size 17)

\paper { indent = 0 }

\score {
  {
\clef "treble"
\time 4/4
\repeat "unfold" 31 { b'4 }
b'4^\markup { \normalsize \italic "Some text to test linebreaking 
calculation" }
\repeat "unfold" 40 { b'4 }
\bar "|."
  }
}

\score {
  {
\override Score.PaperColumn #'keep-inside-line = ##t
\clef "treble"
\time 4/4
\repeat "unfold" 31 { b'4 }
b'4^\markup { \normalsize \italic "Some text to test" }
\repeat "unfold" 40 { b'4 }
\bar "|."
  }
}

\score {
  {
\override Score.PaperColumn #'keep-inside-line = ##t
\clef "treble"
\time 4/4
\repeat "unfold" 31 { b'4 }
b'4^\markup { \normalsize \italic "Some text to test linebreaking" }
\repeat "unfold" 40 { b'4 }
\bar "|."
  }
}

\score {
  {
\override Score.PaperColumn #'keep-inside-line = ##t
\clef "treble"
\time 4/4
\repeat "unfold" 31 { b'4 }
b'4^\markup { \normalsize \italic "Some text to test linebreaking 
calculation" }
\repeat "unfold" 40 { b'4 }
\bar "|."
  }
}

I think the third example is a "bug": if the markup/tempo/rehearsalmark is 
going to cause the spacing to be so poor, the line should automagically break 
so that the item happens "mid-system" (like the last example).

Cheers,
Kieren.

> 
> --
> Phil Holmes
> 
> ___
> lilypond-user mailing list
> lilypond-user@gnu.org
> http://lists.gnu.org/mailman/listinfo/lilypond-user



___
lilypond-user mailing list
lilypond-user@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-user