hi philip,
>> strolling through the spec, it remains not-immmediately-obvious for some :-S
>
> Sorry. :-) That's kind of like the mathematician who stares at a problem
> for a few moments, then writes down the answer without proof, because
> it's "obvious", isn't it?
heh.
that's the basis for most peer reviewed papers isn't it? ;-)
> I didn't mean to appear so arrogant.
not at all! =)
i'm treading water as fast as possible to keep up!
>> is there any documentation, a reference table perhaps, that simply specifies
>> "this $var is set at/after the XXX acl?"
>
> 'fraid not. What particular variables are you interested in? There are
> so many different types. For example, $primary_hostname is always set.
> Ones like $local_part and $domain are set to different things at
> different times, and I hope I have documented all of them.Then there
> are those like the $tod_xxx ones that are never really "set" because
> they change every time you use them.
good, but fuzzy, question. answer is probably: the set of vars that are
mentioned inthe spec, and that balances value with effort.
that said, let's use my latest example as an example ...
it would be nice to be able to simply look at a table/etc and learn:
$local_part: not set until RCPT
sure it's 'obvious' (no, not poking any badgers ...), and has been discussed
b4, but so has just abt everything ELSE in exim. hence the 1/2 million (or so
;-) ) pages of docs/books/wikis/mails/lists/etc/etc. if it were there, i
would've saved myself a fair amt of time.
fwiw, imho, doc/OptionsList.txt -- where i first looked -- woud've been a nice
place to find it.
sure, now i won't make THAT mistake again. but i don't think the doc mission
is 'learn everything by trial & error'. rather, the REASON, i buy the book in
ADDITION to reading the spec is exactly because it IS targetd for startup/new
users.
> To create a table such as you describe, one would have to scour the code
> for every setting of every variable
probably true
> which would be a big job.
agreed. but to be fair (mainly to myself, at the moment ...), so is scouring
the 'universe' for the info each-n-every time ...
cheers,
richard
