The future of Xitami (was Re: [Xitami] PHP, Virtual Domains & New Features in Xitami)

[Ewen McNeill]

In message <3C7FFF3D.6050200@freesbee.fr>, Francis Turner writes:
>Ewen McNeill wrote:
>>In message <3C7FCB0B.981AD47B@alum.rpi.edu>, evan heller writes:
>>>1) When is the next major/minor release for xitami?
>>
>>I have absolutely no idea.  Xitami doesn't have a release schedule as
>>such, and doesn't really have anyone with time to concentrate on its
>>development.  I suggest not holding your breath.
>
>What would it take to get Imatix to find the time? I assume that money 
>would be required to attract someone's attention.

The issue is, at base, that no one is working on development of Xitami
at anything even approaching full time (or really "at all" most of the
time to be honest).

The only "iMatix" contribution to Xitami at present is (a) providing the
site/bandwidth to host the Xitami site, and (b) my occassional posts
here (and those are over and above all my paid work -- there's no "paid
to support Xitami" going on here).  (There are a small number of people
who have paid for Xitami support in the past, who do get support, but
this number of people has not historically been enough to fund even a
part time developer, and it doesn't go towards, eg, paying me.)

There are basically two ways to go from here:
(a) Xitami development can be taken over by "the community" like a
    "traditional" open source project.  Everything that is needed to
    maintain Xitami is now open source as far as I remember (apart from
    the SSL support -- and that could be replicated by a skilled, 
    determined programmer in a week or two if they put their mind to it).

(b) iMatix somehow figures out a way to make money out of Xitami, so
    that it can afford to "dedicate" someone to working on Xitami.  (To
    be a practical proposition it'd need to fund at least half the time
    of a reasonably skilled developer; if it funds less than that
    they'll be given other work to do which will gradually take up all 
    their time (work expands to fill the time available).)

There has been various discussion at iMatix about ways to make money out
of Xitami so as to fund on-going development.  Xitami hasn't been used
in any major projects iMatix has done in at least the last two years;
even prior to that it was largely a "technology demo" rather than an
attempt at an independant product.  It's been rather more successful
than was originally planned.

Funding Xitami development from sales of Xitami/Pro, or commercial
licensing/support has been tried, and at least the way it's been done to
date, it simply does not provide enough money to fund more than
problem-focused development (eg, a customer has a specific problem and
that is investigated).

Unless there's some way to dramatically increase the number of
sales/support licences bought, it's not a practical model to fund
development.  Given the huge number of other webservers around,
including several being given away by people with Very Deep Pockets (eg,
IIS by Microsoft), a lot of people understandably simply use what they 
get for free.

Personally I think the only viable model to move forward is the
"community development" model, ie, reasonably sizable contributions from
non-iMatix people.  There have been some very good contributions in the
past, but not in sufficient quantities/focus to get to the point of a
new release/new features.

I should note, though, that management at iMatix HQ are not in favour of
"community-driven open source" style development, to the point where
it is likely that the public access to the "open source" CVS archive
at iMatixHQ will be removed, and access to the software limited to
occassional releases.

That isn't necessarily a problem: a "community driven" Xitami
development could be hosted somewhere else, based on what is currently
publically available.  I could possibly even host it on one of my
(personal) servers on the Internet.

The pay-for-development model could possibly be mixed with the
"community driven" model by, say, allowing people to sponsor a
particular feature (eg, offering to pay money to add something to the
program).  So people who want something in particular but aren't
developers themselves could still have a way of getting that feature.

Finally, FWIW, "half a developers time" is worth at least US$2500/month
(to the business; the developer may not necessarily be getting twice
that amount each month, but there's also all the overheads of providing
the developer somewhere to work, etc.)

If someone can think of a way to provide a regular US$2000/month or so
towards Xitami development, I'm all ears.  I'd like to see Xitami continue
to develop, but the only way to do that is for people to work on it --
and the only way to do that in the context of a business if for there
to be some advantage (eg, money) in doing so.

>Also on a somewhat related note: if I were to develop and sell a setup 
>that included xitami web server would you like a cut of the money and 
>would you then be willing to provide me or my customers with support?

This is definitely something that we can talk about.  Providing the
money can cover the costs of providing the support (give or take a bit)
it's quite probable something can be arranged.

Email me off list (ewen@imatix.com) and we can talk about it.

Ewen

-- 
Ewen McNeill, Technical Consultant, iMatix Corporation  www.imatix.com