Hi Russ,
I'd like to add my voice to the firestorm too...
>I've found a couple of places where Dan decries patches:
>
>http://msgs.securepoint.com/cgi-bin/get/qmail9812/214/1/2/1/3/2/1/2/1.html
(which says at the end)
DJB>You are of course free to distribute patches---but you're hurting the community
DJB>when you do it. Patches are a support nightmare, to the extent that they're
DJB>actually used; and they make it much more difficult for the author to find out
DJB>what the users actually want.
I have a lot of sympathy for that view, given that Dan gave us
qmail! At the same time, people are doing things with qmail
to make it work in their weird corporate setups, or for fairly
specific tasks, for which qmail is not designed, nor is it likely
to move in that direction. It's important that qmail can be
deployed in these places as well as "Ordinary" setups, since
qmail's kudos and spread is enhanced.
I would also like to mention one of Dan's pages on legal rights, which
specifically mentions patches...
According to the CONTU Final Report, which is generally
interpreted by the courts as legislative history, ``the right to
add features to the program that were not present at the time
of rightful acquisition'' falls within the owner's rights of
modification under section 117.
(that's an extract, there's more). The URL for this is,
http://cr.yp.to/softwarelaw.html
>Because a patch implies that something is wrong, and needs
>to be fixed.
For some people yes, but as others have replied, some rewording
of the patches section may minimise this impression - as well as
helping most of the readers of qmail.org who are the sysadmins
running qmail, sometimes needing a particular tool or patch - and
qmail.org is a brilliant central repository for them.
As most people on the qmail list will be aware, there are some
peculiar setups out there, and according to local needs and
policies, different add-ons will be needed. I feel patches are the
best way to provide these: They tend to be small and to-the-point.
They also require some tech expertise to use, but if people are
running qmail in anger ( = "Real world scenarios"), they hopefully
have this tech expertise to start with - if not, it's not the fault of
you as qmail.org maintainer.
When I have a strange requirement, the first place I look is
qmail.org, followed by the archives - to ensure I don't re-invent
the wheel. What you've given us, the qmail community, with
qmail.org is a resource that helps us to avoid exactly that - it's
good to see what other people do to integrate qmail into their
qmail-hostile environments. Without those itsy-bitsy patches,
a lot of people would be stuck, not really knowing if they can
get qmail working (perhaps modified) in their particular setup.
I think there is a case for some reworking of the qmail.org page -
specifically to increase the prominence of the first few paragraphs,
perhaps some bullet points for the source / mailing list / archive: At
present a [too] casual reader may just skim through these paragraphs,
not realising how important the links they provide are, and reach
instead the boxed text areas, which are more visually catchy.
(I volunteer myself for a sample reworking if this is desired).
Regarding Dan's specific comments about authors trying to
work out what users want (see above):
>From time to time on the list there is a "Wish list for qmail",
which normally bogs down in fairly tech-y discussions. Maybe
Dan could comment on whether he would consider producing
a new version of qmail to incorporate some of the
things on www.qmail.org - presumably some would be as
"Options". If he has that interest, I'm sure the list would be
only too interested to offer their opinions on which "Options"
would be most desired - and people on the list might also
contribute to a group effort to knock some of these "Options"
into better shape (the quality of patches and add-ons is variable),
so that Dan would have cleaner/tighter source to base his work
on (and presumably it'd be in C rather than Perl - so some of
the Perl add-ons would need "Translation"). Maybe you could
raise this idea with Dan, if he's not listening in on this discussion
already...
Whatever you decide, thank you for providing and maintaining
www.qmail.org - it's where I caught the qmail bug in the first place,
and I haven't looked back since.
Please don't do it!!!!
cheers,
Andrew.
