Re: New read/eval Scheme syntax inconsistent in handling existing code

[David Kastrup]

Carl Sorensen <address@hidden> writes:

> On 11/21/11 1:35 PM, "David Kastrup" <address@hidden> wrote:
>>I don't throw out functionality as a rule.  ly:export was an exception
>>to that rule because its presence was fundamentally incompatible with
>>maintaining predictable timing for #.  But I backed up this step with a
>>set of convert-ly rules of rather high quality and coverage (including
>>the entire Lilypond code base).
>  
> Let me take the time (much belated) to thank you for that set of
> convert-ly rules.  I disagree that they are of rather high quality and
> coverage; I believe they are of extremely high quality and coverage.

Well, they remove $ from inside of # from inside of #{ ... #} without
bothering to convert $$ into $ (an oversight that I realized just now).
So if you have been writing

#{ ... #( ... $$defaultlayout ... )  #}

you get defaultlayout instead of (proper) $defaultlayout.  Granted, it's
not like the $x scheme variables are documented all too thoroughly
(i.e., not at all), but sometimes accessing them is hard to avoid
nevertheless.

Maybe I'll try fixing that before we reach a stable release, as then the
effects could become more visible.

On the other hand, the kind of person mucking around with $defaultlayout
and its ilk is likely to not wait for stable releases...

-- 
David Kastrup