> > If I can find the list admin password, I'll go and set it so the
> > Reply-To header points to the list. Anyone here not like that?
>
> Me. I really don't like it, even though I make sometimes a fool out of
> myself by replying the wrong way. Reply-To-patching takes a way choice -
> you can't easily reply off-list any more, as you have to rewrite addresses
> manually.
That's cool then; I'll leave it as it is. You write most of the posts
here at the moment, so your opinion counts ;-)
> > unicode than latin-1. I'm thinking that I might need to post Japanese
> > text one day, and I don't think it encodes well in latin-1 :-)
>
> Yep, Unicode would be my choice, too. Only problem is that strings then
> are not transparently accessed from the database. And changing to
> PostgreSQL for example (as that supports Unicode in the database) isn't an
> option, since people not always have one available.
>
> Hmm. Maybe Gadfly is an option? It's quite small, usefull and fast for
> small stuff, easy to use as is Metakit (ok, not _that_ easy as is Metakit,
> but quite easy), and I think it might support Unicode. ZODB would be
> another option, but would be a bit _fat_.
Another way to do it would be to write a thin wrapper on top of
MetaKit that watches for string accesses and converts them. That way
we just need to wrap a few of MK's classes rather than do the big
plumbing job of switching databases.
Alternatively, can we compile MK to work with unicode natively?
Cheers,
Phil :)
Hi!
> If I can find the list admin password, I'll go and set it so the
> Reply-To header points to the list. Anyone here not like that?
Me. I really don't like it, even though I make sometimes a fool out of
myself by replying the wrong way. Reply-To-patching takes a way choice -
you can't easily reply off-list any more, as you have to rewrite addresses
manually.
And with the right client (Squirrelmail for example) you get nice "Post to
List" and "Reply to List" buttons - one just has to choose the right one
;-)
So I'd say keep it as it is, we only get a chance to lough at the expense
of somebody from time to time ;-)
> unicode than latin-1. I'm thinking that I might need to post Japanese
> text one day, and I don't think it encodes well in latin-1 :-)
Yep, Unicode would be my choice, too. Only problem is that strings then
are not transparently accessed from the database. And changing to
PostgreSQL for example (as that supports Unicode in the database) isn't an
option, since people not always have one available.
Hmm. Maybe Gadfly is an option? It's quite small, usefull and fast for
small stuff, easy to use as is Metakit (ok, not _that_ easy as is Metakit,
but quite easy), and I think it might support Unicode. ZODB would be
another option, but would be a bit _fat_.
bye, Georg
