[TYPO3-dev] blog_example gives error "A cache with identifier "cache_extbase_reflection" does not exist"

[Dmitry Dulepov]

Hi Sebastian!

Ok, here is my answer. I think I risk a lot by posting it but I need to
clarify my position. It burns me for more than a year now :(

Sebastian Gebhard wrote:
> I've got a question to you Dmitry which nags for some time now.
> What is your attitude towards FLOW3 and TYPO3 v5?

When I attended T3TD, I expected that we will get v5 very soon (in a couple
of months). I saw code examples, I saw a very raw unfinished v5 interface.
I was optimistic. But I am not any more.

v5 project has nearly 4 years of history. I believe 3 of those years are
taken by FLOW3 development. We do not have v5 after all these years. We do
not have a usable framework (see other posts). We have a slow and hard to
use code, which is alpha 8 (!!!) after 3 years of active development by
full time developers. There is no deadline for FLOW3 and no deadline for v5
project. This is what we have.

Let's compare it to another existing project. Varien, a small company
behind Magento, started developing Magento in 2007. This is the same time
when v5 team abandoned v5 and decided to go for a perfect framework. Unlike
v5 team, Varien did not went to invent their own framework. They took
existing Zend Framework and used it heavily to create Magento. Magento is
now the most popular ecommerce platform in the world. Varien is famous and
gets lots of benefits from their right choices. Now they can afford 55
employes working full time on open source solutions (see
http://www.linkedin.com/companies/varien).

Do you see the difference in approaches? Varien knew they need to create a
product. They did not have anybody behind them. They just did it. v5 team
had all favors of community, all love and trust. Instead of using existing
tools efficiently, they spend their time duplicating qualities of existing
frameworks. Don't believe me? Look at the caching framework in FLOW3 and
Zend Framework. Basically v5 team lost time duplicating features and
inventing the perfect wheel while others used the imperfect wheel to create
a highly successful product.

There are tons of great PHP framework around: Zend's, Kohana, Symphony,
Doctrine, etc. Just ask yourself: do we need a Yet Another Framework? No,
we don't! We need v5, not a framework. We must use code backed up by others
to develop our own code. We must not reinvent the wheel! We could use Zend
Framework because it is backed by Zend and a huge community. They develop
the framework and fix bugs, we develop v5. Wouldn't this be better?

I strongly believe that we could have TYPO3 v5 in one year *if* v5 project
reused existing frameworks. This approach worked for Varien and it could
work for TYPO3. In fact, we could be better than Varien because our
programmers are MUCH better! But we did not do it.

What to do now? I think v5 team should get a clear and strong deadline when
they must release v5. Not FLOW3 alpha1087 but v5! Not v5rc1 but a
completely ready v5. This deadline should be in this year. If they fail, we
should declare the project as failed. We should abandon it and use existing
frameworks to build v5 paying major attention to usability instead of
abstract perfectness of the code, DDD, DI and other rubbish, which clients
do not care about. It is time to be realistic. It is time to turn to users
and clients and follow their needs for a great UI, speed and intuitive
workflow.

I must tell that I like people behind FLOW3 as people. They are good guys.
It is a pleasure to talk to them. Also I am very fascinated by the work
that Jochen Rau did with Extbase. It is really impressive and I think we
all must honor him for this tremendous work.

We got really great programmers in TYPO3 and it is a shame not to use our
talents to create the greatest super CMS ever. We could do it. Can we still?

So you know now what I think of FLOW3. Now I can go to lunch.

-- 
Dmitry Dulepov
TYPO3 expert / TYPO3 security team member
Twitter: http://twitter.com/dmitryd
Read more @ http://dmitry-dulepov.com/