Exim Win32 Proposal FAQ
-----------------------
Context: Tellurian is wondering about doing a Win32 port of Exim
1. Who is making this proposal? Do they know anything at all?
Well, we like to think so :-) Tellurian is a software development and
support company in Adelaide, South Australia. We do lots of cross-platform
work, including porting Unix software to Win32 for a mixture of private
general-purpose jobs. We have 10 or so different kinds of Unix running on
all sorts of architectures, and one each of most of the major Win32 OSs,
plus MacOS. This FAQ is a bit of market research. Tellurian might decide
to print out the results on the tearoom wall and do nothing further again
with the idea.
2. What is this about?
A port of the latest Exim 3 series to Win32. The port would work in exactly
the same way as the Unix version, taking into account the limitations of
Win32 platforms. It would, in addition, have some Windows-specific
features such as the ability to be controlled by a control panel applet,
and a nice pretty installer.
3. Why?
For similar reasons that Apache on Windows is popular. There are people
that can't get away from Windows for any of many reasons, or who don't
wish to leave Windows. And just as Apache/Win32 has no competition when it
comes to quality Windows web servers that are managed like Unix daemons,
so an Exim/Win32 would be an outstanding Windows MTA that works like a
Unix MTA. The throughput and general reliability might be less on Win32
than under Unix, but in terms of getting the job done well it would
probably rank in front of the current Win32 alternatives.
2. So why ask anything? Do the port and show us the code!
Tellurian is a business, and we would like to continue to make money. So
before doing the port, we need to establish some things. Tellurian would
like to have some sort of a feel for the answers to some questions about
what people would like to see, and what they would be willing to pay for.
I would very much like to hear from anyone who thinks they might purchase
an Exim-based product as described in the rest of this FAQ. We need to
get some kind of idea of potential demand. If it doesn't exist, then we
won't be doing the work. This FAQ is evidence that we think the demand might
exist. We could well be wrong.
3. Pay for? But this is GPL software!
Indeed. Nevertheless, that's what we mean. Windows users expect to pay
for software, and like to get nice reminders of having done so such as
packaging, and manuals, and holographic certificates and the like. That's
the way they are, and nothing anyone says on this list is likely to change
that fact. Plenty of Windows users could present sound business cases
for why they like to do things this way.
4. So what would the product consist of?
Here is what Tellurian is thinking of producing and selling:
a. An all-in-one executable with slick Windows installer.
This whole would be licensed much like any other Windows product -
by site, or by number of users or other such scheme.
Source to Exim capable of compiling on Win32 would of course
be available, but source to the Windows-specific bits probably
wouldn't be in the first instance. If the Win32 port is clean
enough hopefully it would get in the master Exim distribution. No
promises of course.
(Remember, our position is that if someone wants a really good
MTA they should just be using a Unix MTA.)
b. Quality Exim documentation, specific to Exim/Win32
c. Support packages for Exim/Win32, ie you pay Tellurian money and we
give some promises to fix your problems
d. Binary OEM licenses for the complete Exim/Win32 package with all
its Windows-specific tinsel and wrapping. Buy one of these and
you can resell the entire assemblage as often as you like
e. Source OEM licenses for ditto
4. Won't Exim 4 make all this obsolete?
Probably not. Philip says he isn't changing any fundamental design
parameters, and it is fundamental design that matters when doing a
port to a non-Unix platform. So Exim/Win32 users would be in the same boat
as any other Exim users, and be able to follow the same upgrade path.
As to timing; when will Exim 4 become a stable production release? My
guess is well into next year. By that time an Exim/Win32 should have been
long released and be in a state to track the upgrade to version 4. In a
way this is the best possible time because Exim 3 is frozen in terms
of features, last I heard. Stability helps a porting effort a lot.
5. What can I do to help?
Express your thoughts clearly (and politely), and most especially if you
think such a product might be of interest to you.
Regards,
--
Dan Shearer
dan@???
Tellurian Pty Ltd
