-- [email protected] <[email protected]> wrote
(on Sunday, 04 November 2012, 07:53 PM +0100):
<snip>
> I have not been active member of the community because I have been
> busy working on our projects.  So I have not been part of the
> discussions about the design patterns etc. in ZF2.  I also couldn't
> believe we have to start over again. 

The point of a new _major_ release is to allow backwards breaking
changes. Sometimes previous decisions make it impossible to accommodate
new application designs -- or even existing ones! -- and the only way to
allow them is to break compatibility. This is not done lightly --
whenever a change like this has to be made, there's significant thought
that goes into it.

With ZF1, we had 5 years between major releases. This is a huge amount
of time, and during that time, we discovered many people were moving
_away_ from ZF1 because we could not accommodate the applications they
wanted to build. We thus chose to make a new major version that would
address the shortcomings in the framework.

This happens.

What we plan going forward is a shorter timeframe between major
releases. This may seem like an odd decision, as we're saying there will
be backwards incompatible changes more frequently. However, the idea is
that we can introduce _fewer_ breaks, and thus make migration to the new
version easier.

This is a plan for the _future_ though, and does not address your
concerns about ZF2.

> I think ZF2 has great ideas.  But there should be a way to migrate the
> existing code not multiply the time and effort of so many people by 0.

You've been able to write code that works release after release for 5+
years at this point. We had to draw a line in the sand somewhere.

We're working on compiling information on migration. The fact of the
matter, however, is that it's not likely going to be something that can
be automated, nor something you can likely manage to do in a few hours
(unless your site is quite small). But the end result should be quite
compelling -- a truly modular architecture, with the ability to
substitute services easily, and the ability to tie into the application
workflow in a much simpler fashion.

> The sad story in this industry is that if you follow somebody (we follow Zend)
> you are doomed to be always second.
> We have to buy again new books, attend seminars invest again huge amount of 
> time
> learning the new. things.
> It is like pushing the reset button of your computer. Every once in a while
> somebody pushes the reset button and we all start from scratch again.
> And we are not allowed to complain, because we get everything "for free".
> It is a whole industry around such projects. To keep it going there should be
> such restarts.
>  "Instead of learning, you better learn how to learn fast!"
> We better start refactoring everything to use ZF2, right now. And in 5 years 
> ...
> or even sooner they again will push our reset button. :)
> This is the way how it is!

It's clear you're frustrated. However, the web application industry is,
and always has been, a fast-paced one. Five years ago, if somebody had
told me that the dominant web development paradigm would be centered
around RESTful JSON services, I would have laughed; and yet, today,
that's true. Five years ago, I hoped we could cobble together a way to
make modules drag-and-drop for users; it's still not possible in ZF1,
due to the way we did the initial architecture. 

My point is that a combination of technology changes and architecture
will always limit what can be done with code released one year, two
years, 5 years ago. This is, unfortunately, the way it is.

Yes, you can complain about it. But it won't necessarily change how the
project adapts moving forward -- for the very reason that it needs to
adapt.


-- 
Matthew Weier O'Phinney
Project Lead            | [email protected]
Zend Framework          | http://framework.zend.com/
PGP key: http://framework.zend.com/zf-matthew-pgp-key.asc

-- 
List: [email protected]
Info: http://framework.zend.com/archives
Unsubscribe: [email protected]


Reply via email to