[ic] "Interchange 6"? Really?

Jon Jensen jon at endpoint.com
Wed Sep 28 00:32:50 UTC 2011

On Tue, 27 Sep 2011, Paul Jordan wrote:

> I think David feels like your throwing away *his* knowhow, and the time 
> and money he has put towards ideas and concepts in the tech he built 
> around IC.

Nobody can throw away his skills. They may slowly atrophy if they are only 
useful on older technologies. We all have to keep up with change.

As to the question about whether it's "even Interchange" if it changes so 
much, just consider history. Minivend 4 was a big change from Minivend 3, 
which was a big change from Minivend 2. Perl 5 was a huge change from Perl 
4. It still felt like Perl, just much better.

> The thought that I pour tons of time and money into their projects and 
> have custom code built right now and for the next 2 years only to know I 
> am going to have to learn something new and rewrite everything - free of 
> charge - is a bit hard to swallow.

You don't have to learn anything new, and you don't have to write 
anything. It's one of the beauties of open source. If you host your own 
software based on open source on your own server, why would you have to 
rewrite it or learn anything new?

Consider the software stack of an average Interchange deployment:

Linux: One of the most popular operating systems on the planet, with 
several popular and well-maintained distributions to choose from.

Perl: Perl may not be the hippest language around right now, but it's been 
remarkably stable, is well-maintained, and has a more vibrant ecosystem in 
CPAN than ever before.

Interchange: Interchange 5 is in maintenance mode. A quick review of its 
commit history makes that clear. But it's stable, and makes relatively few 
demands even when you upgrade. Compare that to the huge amount of churn in 
the Ruby on Rails platform, or Django and Python -- in both cases, 
upgrading either the language or the framework brings serious amount of 
breakage. It's likely worth it, the cost of progress, but it's still a 
cost, and people don't pay much of those kinds of costs with Interchange 

PostgreSQL or MySQL: Both very widely used, stable, and maintained 
databases. Even with the uncertainty Oracle has brought about MySQL's 
future, you can choose from Percona, Drizzle, or MariaDB. MySQL is going 
to be around. And PostgreSQL is better than ever, progressing amazingly 

There are hundreds of serious Interchange 5 applications running out 
there, and all of us on the Interchange core team have a strong interest 
in keeping Interchange 5 maintained. We've been maintaining it because our 
customers need it, and that will continue to be the case. There's almost 
no chance that all Interchange 5 sites will be migrated to Interchange 6 
or anything else within the next 5 or even 10 years, because migrations 
are expensive, time-consuming, and difficult, and Interchange 5 works 

In fact, we could use help with the maintenance of Interchange 5. Very few 
people actually contribute to it, and there are lots of things to do, for 
example, improving compatibility with Perl 5.14 and UTF-8.

> However I can see that the community would want more than the roadmap. 
> Maybe someone can describe what it will be like when ready. What will 
> the code I use in the page be like to achieve [query], [if], [scratch], 
> or what kind of rework will all my usertags will need?

Talk is cheap. I'm not very interested in talking about what Interchange 6 
will be like, because we've done that plenty, and that keeps it in the 
realm of vaporware, where it's been for years. Obviously we have to 
discuss features and architecture to make something coherent, but that can 
be done at a small level, with lots of code writing. Without code, the 
talk is wasted.

Let's just try it out as it develops, and see. Better yet: Get involved! 
Write some code! Ask Racke where you can help and I'm pretty confident he 
will tell you how you can help. You can help guide Interchange 6's 
direction if you contribute.

Just like with any other true open source project.


Jon Jensen
End Point Corporation

More information about the interchange-users mailing list