Re: order of function arguments

[Janek Warchoł]

2013/4/23 David Kastrup <address@hidden>:
> Janek Warchoł <address@hidden> writes:
>
>> Hi,
>>
>> 2013/4/23 David Kastrup <address@hidden>:
>>> Janek Warchoł <address@hidden> writes:
>>>> Btw, issue 2540 made it possible to write
>>>>
>>>> \version "2.17.3"
>>>> { < \tweak Accidental #'color #red cis' fis'' > }
>>>>
>>>> which colored the sharp of the cis.  Is it possible to achieve the
>>>> same goal with newer versions?  I've tried some combinations with
>>>> 2.17.13 but they failed.
>>>
>>> If you don't want to read the manual, you can use convert-ly.
>>
>> Hmm.  I tried, but apparently Frescobaldi tricked me - it appeared as
>> if the conversion didn't change anything.  Apologies - i've tried
>> doing this directly with command line, which worked and gave me
>>
>> \version "2.17.3"
>> { < \tweak Accidental.color #red cis' fis'' > }
>>
>> which, quite frankly, is exactly what i hoped for!  Woohoo!  Now we're
>> talking.  So, current \tweak syntax is actually
>>
>> \tweak grob-property-path value grob-or-music
>> (not \tweak property value grob-or-music as i mistakenly thought)
>
> And if you really put a grob name last, you don't get a tweak but an
> override.

So, a \tweak gets, so to speak, "internally transformed to an
override"?  This is cool!!

>> and _that_ makes my day!  I wish that someone told me that when i
>> asked about it in https://codereview.appspot.com/8647044/#msg2 - it
>> would save us a lot of time.
>
> Uh, it was not at all relevant to your question?  You explicitly state
> there "I understand that grob-name is at the end because it's optional,
> and we want to omit it when we're using \offset as a tweak."
>
> You claim that you understand the reason, and you explicitly state the
> reason.  That does not exactly make it easy to guess that you don't
> understand the reason.

bang.
Well, in that sentence i had used the word "understand" in the meaning
"infer" (that's a valid use according to the dictionary
http://www.thefreedictionary.com/understand).  Apparently i should've
used some other word instead, like "suppose".
lol :P
It's amazing that i'd been almost right in that comment...  I mean, it
_is_ true that "Since David K's change that allowed to use
dot-separated list for specyfying grobs "together" with properties, we
could process both the property and grobname as one argument".

best,
Janek