Re: component definitions

ok, here goes again.

> As you can see, the difference between the above approaches is precisely 7
> lines in the locations that I have specified. I have put the '? lines'

I am not sure what locations you have specified. But assuming it is 7 lines
more to write for the "new component" approach, one also has to add that
these are dispersed over 3 files, 2 of which contain 1 line. I cant help
it - I dont like it.

> Note that there is basically no difference between the two approaches when
> 'use' is concerned -- they both require one short line.

true.

> This is where we come from, actually. In the pre-2.2 world aliases were
> optional and Spindle, for example, did not use them at all. This is not
very
> compatible with the notion of libraries, however. Please have a look at
the
> mailing list archives on this topic, or perhaps Howard can explain his
point
> of view.

strange. I have come across other systems with nested namespaces (e.g., the
Java language), and none required to maintain unique global IDs for every
item, as tapestry does for components. I remember one system that does have
similar requirements, but I'd rather not mention it here..



-------------------------------------------------------
This SF.NET email is sponsored by: Geek Gift Procrastinating?
Get the perfect geek gift now!  Before the Holidays pass you by.
T H I N K G E E K . C O M      http://www.thinkgeek.com/sf/