Re: User Experience Engineering

[Erik Sandberg]

Citerar Linda Seltzer <address@hidden>:
> I would greatly encourage the project to focus on the user interface and
> the user experience if this is to catch on in a large way.
> 
> Having to install separate editors (and who knows what bugs that will
> bring and what other mailing lists one will have to subscribe to...) or
> get into the system with DOS commands, and to understand what is wrong if
> the flags are wrong, etc. does not constitute user interface engineering.

Hi,

Your criticism is quite valid, IMHO. However, as far as I have observed,
lilypond does not seem to be ready for getting a full-blown GUI just yet. The
developers seem to be very aware of this; e.g., Han-Wen has written a
proof-of-concept GUI called ikebana, which is too slow for practical use. This
indicates that there are some problems in lilypond that need to be solved
before a useful GUI can be created. One problem is that it's difficult to know
what the right APIs should look like: if an officially supported GUI is written
too early, it is likely that lots of reworking needs to be done due to poorly
designed APIs.

My observations of the recent development, is that deep changes under the hood,
often tend to eliminate obstacles that prohibit the development of a GUI.

(and as others already mentioned, third-party products like Denemo and
Rosegarden could always implement lilypond-compatible GUIs).

Erik