[Dovecot] bleah

[Tom Allison]

Not to be a bitch, but I've asked several times about some basic 
questions or examples to clarify the essentials for configuration.

What do some of the parameters mean?  Expecially the ones that are 
mentioned only one in a paragraph and no where else in the documentation 
or wiki.  (ie: login_process_per_use)

And there have been a few others as well. that I won't bother to repeat.

But I've been sitting on this list for a while yet and have found that 
when it comes to configuration, none of the emails posted have been 
answered or addressed.  The closest I got was a suggestion to use views 
instead of tables in postgres.  But technically nothing was ever said 
about dovecot itself.

So I have to ask.  Is the sketchy documentation in the source file, the 
web page, wiki, and limited reponse on the mailing list a matter of "I 
haven't gotten around to it yet" or a deliberate move to promote the 
commercial support potential for Procontrol?

I've run into this kind of environment twice before (cyrus and razor) 
and in the case of cyrus they just didn't believe in doing the 
documentation but suggested I write it myself.  That may be the case 
there as well ("I haven't gotten around to it yet".  That's OK but cyrus 
has the difficulty of no one still not providing any answers to how 
anything was done, configured, functioned, or why.  So documentation was 
kind of a reverse-engineering project in itself.

In the case of razor it's a mix of no time and they would rather support 
the company than the project.

I went back and re-read the emails that I had posted earlier and am 
pretty sure that they weren't too stupid (ie: how do I test an imap 
connection), difficult, or obscure (well, maybe one).

I'm just trying to find more information on how to correctly configure 
this server such that I can retain the concept of keeping security in mind.

I'm trolling through the archives and there doesn't seem to be much 
discussion from anyone else.  So maybe I'm really stupid and can't do 
authentication schemes.

I'm sure by this point I've really pissed a few people off and for that 
I mostly apologize.  I'm a little bent here myself.  It just that I 
didn't plan on having to set up sniffit to track my IMAP traffic in 
order to verify that things were actually working securely.

I'm going to plug along and do it myself and probably get something to 
work.  I have plain_text authentication working now, but was really 
hoping to actually get something secure.  At this point I really don't 
expect an answer, but do hope that after 1.0 is released there will be 
time available to get some documentation in order.

IMHO the level of documentation that is available on a software project 
is critical to it's use.