[theforum] Decision time - please cast your vote beforenoonSaturday Nov 20th (12.00GMT)

Morgan Kelsey morgan at morgankelsey.com
Thu Nov 18 14:45:08 CST 2004


From: "Martin Paul Burns"


> Morgan wrote on 17/11/2004 17:45:51:
>
>> haven't past rewrites involved designated "teams" with leaders that had
>> "todo lists" who then passed out the work to the wonderfully willing?
>
> Nope. And it's not a valid comparison.
>
> 1) Previous rewrites have been deep coding a completely bespoke system,
> rather than on a modular framework application where the focus is on light
> customisation

i guess that depends on your definition of light.
i've never seen "a small job" in the internet world

> 2) Previous rewrites have been along the lines of "Bunch of folk get
> together without a *great* deal of pre-setting of requirements by evolt"
> even though the participants may have planned a lot. Generally it's seemed
> to those of us outwith the codefest that cool stuff just emerged.

that kind of seems to be what's happening now....but that's ok.

> [snip]

>
>> why is this any different?
>
> See (1) above.
>

huh? i thought we were trying to imporve?


>> i don't think anyone really cares (i dare say jeff included) what
>> platform/system/language evolt runs on as long as we duplicate the
> existing
>> functionality.
>
> Please see the 'suits you sir' thread: we have already accepted that we 
> may
> lose some functionality for the greater future benefit of a framework
> where:
> 1) We are not the only developers
> 2) There are 3rd party modules to hand
> 3) We're only working on the top 10% of the system.
>
> That discussion is dead, gone, and decided.

martin, when reading back that thread it appears to be a conversation you 
had mostly with yourself. there was no vote on what functionality we can 
"live without", and there is still no definition of what those functions 
are. i think a lot of us are a little nervous about a "we think we can do 
that" attitude. this is not meant to bash john and the current team of 
migraters, i think they are doing a great job -- but i'm sure they would 
appreciate a more complete feature list so they themselves can give us an 
accurate assessment.

as the aardvark pointed out, the last 10% of a job is the biggest, most 
draggiest, most pain-in-the-hiney part.

it also seems to me, that the majority of the core developers of the last 
rewrite (those that are still around, and i could be wrong) don't want to 
see any functionality or quality lost.

i think we should heed these folks, or, revisit the "re-skinning the old 
thing and running on linux" arguments that davidk proposed, if we're 
incredibly hot to get off of .jeff's server.
i also think the current team of developers is entirely capable of producing 
a high-quality product -- but i'm not so sure on the timeframes being 
proposed (again, with no specs).

>> jeff is charging us $50 a month? i think?
>
> We're also paying $100/month for a server that's sitting doing nothing.

i believe it serves as our mail server, and carries the load for BEO, that 
ain't exactly "doing nothing".

>
>> can't we afford that until a new site is *ready*?
>
> Where 'ready' is in theforum's opinion good enough to launch. This doesn't
> mean every last tiny thing that we have now. It certainly doesn't mean
> keystroke duplication. And it doesn't mean limiting ourselves to what's
> there now if we can just add a module that does cool stuff.
>

i haven't heard anyone propose keystroke duplication, but rather not losing 
features that have been carefully crafted by existing and past members. nor 
have i heard discussion about limiting future functions, just not losing 
what we currently have.


>> i think we can all see that evolt will *work* in drupal, but as jeff
> pointed out there is a huge amount of tweaking to be done, and we all kow 
> that
> takes a lot of time.
>
> Yes there's tweaking to be done. But think of a few things:
>
> 1) The majority of existing functionality was put together in a few
> weekends. We have 4 weeks. The hard bits are done.
>
> 2) Unless we have an impetus to do the tweaking, inertia will reign as it
> has done for the last few years. Moving to Jeff's system was supposed to 
> be
> a short term thing. What, 2 years ago nearly?
>

i agree, but i still think we need a features list.
and i don't think it's fair to make .jeff "the bad guy" here.


> 3) Going out with a bit of cruftiness would actually be a useful spur, not
> only to our current coding hotshots, but to all the others out there in
> evoltland that *could* contribute. We could use it as a "You think this
> could be better. Yes. Now come and *help*" - we could gain a whole new 
> raft
> of volunteers who *can* come and work on it. And gain community 
> involvement
> in the requirements setting, rather than presenting a fait accompli. Would
> you swap site perfection for a revitalised community? I know I would.

not sure i agree.
i think we have a constantly re-vitalizing community, we've had
lots of new talent emerge in the few years i've been here.

but we've also lost a lot of really valuable members for various reasons.
i'd really like to see a redesign happen that doesn't piss anyone off.

there's a pretty ugly attitude toward .jeff -- who -- unbeliveably -- is 
being
difficult back -- when we never even recovered the machines that matt
so generously took a weekend of his life out to physically move to our
previous host.

maybe the elder folks can form some type of a type of council that
signs off on the new work in some way....(just an idea)

> 4) Redesigning in public is fun, exposing the process - are we really
> saying we're so much better than Zeldman who does so regularly?

nope, and not a valid comparison.

evolt hasn't put this face on before -- it has always struck me as a 
mysteriously-professional site, never to throw an error or look 
out-of-whack. i think it may be a disservice to those who have developed for 
evolt in the past to take a flippant attitude at this point. and yes --  
think evolt is better than zeldman ;-)


nagrom
http://www.morgankelsey.com



More information about the theforum mailing list