Compose sequences that produce more than one character

[Simos Xenitellis]

Hi All,
I am opening a new side-discussion about layouts, and specifically the issue
about compose sequences that have to produce more than one character.

Some background:

$ echo '<asterisk>                      : "X.Org rocks!"' >> ~/.XCompose
$ setenv GTK_IM_MODULE xim
$ gedit

When you press '*' in gedit, you get 'X.Org rocks!'.

A more serious use for this feature is for compose sequences such as

<dead_acute> <t>     : "t́"          # is actually 't' and the
diacritic mark acute (0x301). It's two characters!

Currently, keyboard layouts for scripts that no precomposed Unicode
characters exist, use the diacritic marks verbatim.
For example, there is some key on the keyboard that produces a
verbatim acute '0x301'.
So, if you press a, and then verbatim acute, you get á, which is
actually two characters.
If you use GNOME, you can insert this 'verbatim acute' with
Type a, then Ctrl+Shift+u, then type 301, them press Space.
The issue here is that a user can type many such acutes, and get nice
results such as
the a and the w in
http://www.flickr.com/photos/simosx/2278982091/
This means that the layout cannot limit the number of consecutive
diacritic marks.

Is there a rational not to have compose sequences that output more
than one characters?
Or, is it something that just did not caught on yet?

Simos