I would certainly not make the mbox mistake again if I was starting over. Nor would I make any any of a long list of other mistakes that I've noted down over the 18-year history of c-client! I would instead make entirely new mistakes. ;-)
That's what we need -- new complexities! Interesting that when people "rebuild" new applications from scratch to replace old ones - the argument is often that they are getting rid of all those old "mistakes".... entirely overlooking all the new ones being introduced.
I'd prefer to flush the code that supports /etc/c-client.cf entirely. I can't because of compatibility. There are many pitfalls associated with /etc/c-client.cf besides that one. Even if I filled in the one pit, there are still plenty more to trap the unwary. I worry that it may make things worse if it seems that I'm making it safer... :-(
Could be.... but I can say that, despite its pitfalls, c-client.cf is very useful IF used correctly. It allows you, for example, rebuild uw-imap (as you come out with new updates) without having to go in and modify src/osdep/unix/Makefile (or other source files) each time. That is a real and practical benefit. Also, flushing c-client.cf would not fix the "mbox" pitfall - as the createproto resides in the source Makefile anyway... and the same mistake would be made there.
As I've said many times before, I recommend against using /etc/c-client.cf because it is much too easy to shoot yourself in the foot. I've done so myself.
Yup - me too... as you already know. But I still very much like it.
_______________________________________________ Imap-uw mailing list [email protected] https://mailman1.u.washington.edu/mailman/listinfo/imap-uw
