Jean-Marc Lasgouttes wrote:
Abdelrazak Younes <[EMAIL PROTECTED]> writes:
On the implementation point of view, charstyle as inset has *many* benefit:
- shorter code
- cleaner code
- generic code
Yes, but we should never change the look and feel just because the
code is short clean and generic. It is our role to strive for short,
clean and generic code, but at the same time we should not force the
consequences of the on the user. The interface we expose should be a
consequence of our vision and be as much as possible independent of
mundane problems like implementation.
I know that what I write is just wishful thinking, but this thinking
is important. Software written by technicians (even good ones) is not
necessarily good software.
The original LyX, with its mix of latex-like structure and word-like
look-and-feel was successful because Matthias found the right balance
between the interface and the inner working. We all know how horrible
the code was, but it had an intuitiveness factor that was very cool
IMO. We should not loose that.
Sure, good look&feel is the most important goal but, at the same time,
any new feature should not add software layer that will takes *years* to
cleanup. Let's please not forget that too.
1) spans --- which is how font attributes work today;
And which is horribly complicated...
Try to forget about implementation details. It is not the good
argument for deciding what the UI should be.
I agree of course. What I am really saying is that the UI we want will
be easier to achieve with a sane implementation. I am pretty sure that
with enough effort we will reach that point without sacrificing on the
cleanliness of the code.
Abdel.
