Re: master b8062be 3/5: Remove some compat code from eudc-bob.el

[Stefan Monnier <monnier@iro.umontreal.ca>]

> Right...  but these days, we don't use Unibyte buffers for text at all?

That's right.

> So is this still baggage that we have to carry around?

I'm pretty sure it's not necessary.
But I don't have a clear idea of how much breakage it would involve:
someone should try it out and see what breaks in the packages they use to
get a feel for how serious it is.

> Hm...  I guess if we have a unibyte ASCII buffer, somebody might still
> want to do EOL conversions...

I get the impression that the more problematic case will be process
filters: I think it is fairly rare to have ASCII text (really meant as
*text* where we want EOL conversion) in a unibyte buffer, but I think
it's pretty frequent to have ASCII *text* inside unibyte strings.
The difference is that the unibyteness of strings is determined
semi-automatically.

So we could end up with a half-way rule where coding-systems (i.e. EOL
conversion) doesn't apply to text coming from or going to a unibyte
buffer, but does apply to unibyte strings :-(


        Stefan