[ic] "Interchange 6"? Really?

Paul Jordan paul at gishnetwork.com
Thu Sep 29 19:49:18 UTC 2011

> From: jeff at downtowndevelopmentplan.com
>> #1 We're doing a ton of work right now, knowing it will have to be 
>> redone.
>> So the point in time we can do our current development in IC6, the
>> better, and this depends on the next point...
>> #2 How do you convert something large that is constantly being worked on?
>> You can't. You need a bridge so they can coexist during a migration
>> period otherwise you'll never have the chance to convert unless you
>> pay a third party a ton of money to convert everything for you on
>> short order.
> I think that I consider a particular IC version as a point-in-time 
> collection
> of routines that I'll use for development. When a client needs a feature
> added, the IC version doesn't impact my ability to deliver that feature. 
> The
> only requirement is that you have to backport security fixes and any 
> notable
> bugfixes. I have 4.8 running for some customers and 5.7 for a couple of
> others. At some point, we'll use 6.0 for new customers, after it's had a 
> bit
> of vetting. In your case, I think 6.0 needs that vetting before you'd be 
> able
> to make the decision that you're looking to make.

You are not understanding the concern - there is no "when a client needs a 
feature". If a system is always being improved - in other words, there is 
zero time in which development is not happening, then in order to deliver 
the enhancements, you will always be coding for IC5.x. How do you change to 
IC6.x in this scenario when you can't have parts in 5 and parts in 6?  Are 
you suggesting spending another decade of time and money painting myself 
into a corner with IC5.x?

I'd rather close up shop than have separate systems that cannot seamlessly 
interact for years at a time. That's disgusting. My clients that have third 
party integrations have priority on those integrations to be duplicated in 
IC - to remove ourselves from ghetto non-interactive integrations. We have 
one right now that is a network marketing engine - not like IC affiliates, 
but like Avon. It's huge and complex. The client can't drop half a million 
and just have it done, so we have a migration schedule to bring it in IC. 
However, as soon as the switch is done, I will know it should be in IC6.x - 
that is depressing. Why? Because there is no end to the work we have in the 
new system - so when do we switch to 6.x!

There needs to be a bridge. With a bridge, I can have seamless interaction 
from 5.x to 6.x and 6.x to 5.x and take 10 years to migrate if I wish. It 
seems to me there only needs to be session sharing.

I'll admit I am probably a little more sensitive to this than others, 
because I just-so-happen-to-be in the midst of a complete rewrite of our 
systems. Our client-specific admins are now completely in Jquery, and all of 
the IC code is being revamped. I had saw this as a strengthening of 
investment. But now, it's something that is going to have to be repeated. 
Right now I can rewrite everything and Seebase 2.x and 3.x can coexist, 
nobody has any awkwardness. I'll accept a rewrite into Seebase 4.x, but it 
has to coexist, otherwise I don't see how I could do it and maintain high 

Jon's given me hope. If they can come up with a plan and the community can 
send him gold, or beer, or virgins, or whatever, this will be an exciting 
time for everyone. If there cannot be a bridge, I'd have to seriously 
consider our future.


More information about the interchange-users mailing list