preliminary questions..

hi, all..

for awhile, i was reading along with the tutorial, and trying to
actually DO it all, but it seemed like i was spending more time
fiddling, and not understanding.. so i just read the tutorial like a
book, and everything made WAY more sense.

i am thinking about using seaside on a project i am starting in house,
but i have a few reservations that i am hoping everyone will shatter
for me..

a little background:

we run a small development team.. 3 designers and a programmer (me)..
we currently use rails. we have done several rails projects, and have
kinda fallen into a groove...
we use git to keep each other (and our testing server) in sync. we use
capistrano to shove everything to the servers.
we use mysql.

the good part is, we are not married to any of these platforms.. but
the upcoming project DOES have to have a mysql backend as another
client will need to access the data. most likely directly.

the project is a web app. we started doing web sites in rails, and
have found that our projects are beginning to be more and more like
apps. this new project is an app.

here are some of the concerns i have:

after reading about GLORP, it seems like it's alot of work to get up
and running. if that is all there is, i can deal with that. what
happens in the case where the database structure changes? when fields
are added, changed, deleted, etc? it is really no big deal with rails
and migrations, but how much work is this with glorp?

team working - one of the designers on our team also does some of the
programming work (displaying data, etc). how tough is it to keep a
team like this together with monticello? i would like to be able just
automagically suck in all changes in programming and design whenever
they happen, without having to think too much about it..

deploying - we use capistrano to push everything out to a testing
server and a production server. this also restarts the servers, etc.
this is WAY easy.. effortless.. what is it like to deploy an
application?

workaday things - it seems that with rails, a LOT of the things you
have to do for every project (authentication, scaffolding, crud, etc)
are readily available via plugins. before i start to write anything
that remotely sounds like something someone else might use, a quick
search usually reveals something ready to install that someone else
did..

okay, so if everything is so great, why do am i looking at other frameworks?

a few things:

while ruby is a great and fun OO language, rails still looks at the
web as a series of pages. for a great number of applications, this is
a totally workable scenario. once things start getting a little more
intense, this approach totally gets in the way of itself. the seaside
way is much more what i am looking for.

the debugger in smalltalk is awesome. the debugging and development
workflow are good enough that i am not even worried that i can't use
emacs to develop. in rails, i am still using alot of really silly old
school debugging methods.

if persistence via an external database weren't an issue.. the agility
of just banging out objects everywhere for everything seems WAY more
speedy than developing while trying to make everything talk nicely to
each other.

anyway.... i would really value any input i can get.. thanks!



--
___
peace,
sergio
www.village-buzz.com
photographer, journalist, visionary
_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

On 17-Feb-09, at 10:03 PM, sergio_101 wrote:

> after reading about GLORP, it seems like it's alot of work to get up
> and running. if that is all there is, i can deal with that. what
> happens in the case where the database structure changes? when fields
> are added, changed, deleted, etc? it is really no big deal with rails
> and migrations, but how much work is this with glorp?

You looked at www.glorp.org  ?  or just ask Alan Knight directly.



>
>
> team working - one of the designers on our team also does some of the
> programming work (displaying data, etc). how tough is it to keep a
> team like this together with monticello? i would like to be able just
> automagically suck in all changes in programming and design whenever
> they happen, without having to think too much about it..

With any smallktalk librarian type of function you have to think a bit  
to
prevent someone checking in a change that trashs everyone's image and
messes up their work day morning.

I can say with Sophie we worked across 8 or so time zones with a bunch
of experience developers and didn't have any problems once we decided
the protocols involved in promoting work items we individually we were
responsible for into the common MC repository.

We even did branches for fundamental  month long programming changes and
remerging at the end.

Failure to run the Project's SUnits before promoting broken changes  
was the biggest issue.
Hint if the tests don't run you *can't* press that save button....

--
=
=
=
========================================================================
John M. McIntosh <[hidden email]>
Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
=
=
=
========================================================================



_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

thanks for the input!

checking into all this now..
_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside