[TYPO3-english] worried about 4.x (6.x)

Thomas Skierlo pubtsk1 at pix-pro.eu
Fri Oct 12 15:17:23 CEST 2012


Hi,

spoiler ahead - apologies for the long text -- but it has to be this 
way. Don't read if you don't have the time.

it's perfectly clear that any thread like this must get the emotions 
boiling, because one party is blaming the other for not taking the right 
decisions or doing a bad job. I am quite sure that all developers are 
working hard to improve the technology and to get TYPO3 to the next 
level. This list, as well as the product TYPO3, wouldn't exist, if many 
devoted people wouldn't have done a perfect job for many, many years -- 
and are still doing today.

To get an understanding of the integrator side please listen to my 
personal history of TYPO3 usage. I started to learn TYPO3 basics in 
2008, building up a first, small, dev-site, using marker based 
templates. While I did that I noticed the limitations, and that there 
was another way to do things, which might be more powerful. After 
reading "Futuristic template building" I instantly knew that this was 
the way to go in future. Only TV seems to offer a way that one 
(sub)content element is aware of other content elements within the same 
data structure. You definitely need something like this to create FCEs 
like tabs or accordions or even grid-elements with one empty column, 
which must be handled differently to columns which aren't empty if the 
markup should be semantic. From there on I lived with the CON aspect of 
elements not being movable to other installations easily. Except this I 
never reached any limits with TV.

At that time the first rumors came up about a new TYPO3, which should be 
developed in parallel to the existing 4.x branch and which should offer 
a migration path for older versions once finished. A Manifesto, which is 
much more than a letter of intent, was ensuring that I took the right 
decision with TYPO3, because I noticed that people did care about 
migration paths. At that very moment one alternative could have been to 
build up a perfectly new CMS, which no migration paths at all. People 
could have moved on with v. 4.x and than face a new learning curve for 
the new technology, once this is ready for production.

At that time I definitely preferred the Manifesto way, because I felt 
that old-school ext.developers would stop further development and 
support for "old" extensions instantly if all work would be lost after a 
year or two. Even with the Manifesto still in place many of them stopped 
support for their old extensions, which directly leads to the current 
situation of TYPO3 being the best car in the world, with only a lack of 
2 tires. Everybody involved is perfectly sure that it will be the best 
car on the marked again if someone would build the 2 tires missing. One 
of the two tires missing is a working replacement for TV. Now you might 
say: Hey, we gave you Extbase and Fluid, a Fluidtemplate object - and 
FCE-killers like gridelements oder FED/Flux from the community. At least 
I heard those words a couple of times, so I decided to stop my 
commercial work a couple of month ago and move on to the new templating.

The FLUIDTEMPLATE object works as a direct replacement of the TV page 
template. Migration can be done withing a day without loosing anything. 
BON. Only thing which is missing is a decent way to get some constants 
into the Fluidtemplate, without defining them as variables. I asked for 
help on the lists last month, but never got a hint or a solution. 
Anyhow: Consider the page template done by means of Fluidtemplate. 100% 
TV replacement. And even fun to do.

Next in line was the "Grid" which modern, semantic and responsive 
layouts like Twitter/Bootstrap need -- as well as all websites with a 
modern look and feel. So I installed the first so called FCE killer: 
Gridelements. With this extension one can build multicolumn content 
elements which do their job, but since every column is handled 
individually it seems not to be possible to manipulate content of one 
column depending on the content of another column. Tried to find help on 
the lists, but didn't get any. Tried to offer my help to one of the team 
members (I tried it only once by email), but didn't get any reaction. 
When it comes down to real grid-elements (=fixed multi-column 
structures), I would give the current Gridelements ext. a 80% as TV 
replacement, the BE view get's a 120%.

My next try was to use Gridelements for a real FCE: An accordion. To 
build up the 2-tier markup you either have to visit the database 
multiple times (which is bad) and use another custom element (accordion 
content) instead of "any content element" (which isn't good either) OR 
you will not succeed. Sections are not supported, and sections are 
needed to build some advanced FCEs. Current trunk version would get a 
20% as TV-FCE replacement from me. Spent 3 weeks with gridelements. Work 
still unfinished.

Next was FED/Flux: Did like the possibilities, but didn't like the 
general concept from the very beginning. Wrapping private directives my 
IDE doesn't understand around Fluid directives my IDE doesn't know 
either, kind of parser for a parser. Started with the page template 
following the detailed explanations on the very good and helpful website 
-- but didn't succeed. No output at all except fatals. Maybe I've only 
installed a broken version from trunk? Don't know. Spent 3 or 4 days 
until I moved on to getting the job done by an own extbase extension. I 
never loved the idea of "writing an extension" just to get a custom 
content element -- in my naive thinking this basic scaffolding should be 
part of the CMS base functionality.

Now, after spending a couple of month with extbase I still didn't 
succeed in building my own content elements. Had issues with caching and 
localization. Tried to analyze other, basic, extbase extensions from TER 
to get a way to handle BE modules, but didn't find many suitable for my 
learning. I am quite sure that it's possible to do the job with extbase, 
but once you got stuck the time needed to get a step further is much too 
long to still behave responsible.

Another possible player might be "WEC content elements" way back from 
2010. It addresses some of my current problems and it supports sections. 
But the project is not available on forge, has no issue tracker and 
seems not to be active any more. Didn't try it. Was getting too tired 
trying stuff.

During the last days I took a first look at Typo3 Neos. My primary 
aspect of this first look was to find deeper similarities to extbase 
from the current 4.x versions. My first impression is that it might be 
much easier to start all over again without any extbase knowledge from 
4.x, except DDD and MVC principles. At least Fluid looks similar. 
Awareness of content seems to be built-in. Very promising. But: Not 
ready for production. Can learn it today, but wouldn't use it for a 
while for real life projects. If I now move on with my current problems 
to Neos, I still have no solutions for them in the 4.x range. Besides 
that: Does anybody still believe in any later migration path? From 6 to 
Neos? I don't. Typoscript2 has not similarities with the current 
Typoscript and markers are (luckily) unknown. So we might be in that 
very situation which would have been the only alternative in the early 
times of the Manifesto.

One thing seems to be pretty clear: There's no room for TV in the Neos 
world, so my motivation to get it to version 6.x wouldn't be too big, 
since it will be the final -- call it burial - level. On the other hand, 
there is no current alternative. At least I couldn't find any. Now 
substitute TV with any other important extension name, and you'll get 
closer to the real problem we are all facing today.

Please don't be too harsh with people getting nervous about TYPO3. 
Better listen to their words by filtering out any emotions which might 
spoil the real meaning. Many people did invest a fortune to "understand" 
TYPO3. To learn, how it can be used to even build up an own business. If 
they feel, that their personal invest is taken at risk, they are getting 
nervous. Some will tell you about that feelings, others will not, just 
moving on to other products silently.

I would have loved to participate much more in the community, either by 
member fees or personal labor. Just imagine my 4-month struggle with a 
working FCE replacement, which isn't been over yet. What a waste of 
time, one might think. Would love to give back my solution to the 
community, but couldn't find any. Would love to give my time to the 
community, but can't spare any because I have to keep up with new 
techniques.

What I am missing is a TYPO3 team "above" the core team -- let's call 
them "Big Picture Ensurers", "Keepers of Usability" or "Tamers of 
Scientists" (especially the last task is often the most unpleasant, but 
mostly needed one). Is the current official version of TYPO3 INCLUDING 
mission critical community extensions still able to build up a modern, 
2012 website? Is it still a suitable base for integrators, or has it 
moved to a state which can only be handled by developers? If so, 
wouldn't it be fair to share this very important information? And 
please, don't mention 4.5LTS. It is not really a long term solution. 
What might help with the problems might be another LTS commitment. At 
least this worked better than the Manifesto :-)

I'm quite sure that extbase/Fluid is excellent. Probably FED/Flux is 
excellent too. Why not combine both into one, or even add Gridelements, 
for BE presentation? I didn't find "my" solution in any part of the 
active players group.

Why don't give the outside world a break by slowing down any further 
development of new technology before the old one had a decent chance to 
keep up? At least until all basic scaffolding can be done with on-board 
tools or system extensions. Until that time only good old TemplaVoila 
seems to be able to give all the answers needed when it comes to templates.

Can't imagine that I'm the only one with this FCE problem today. Where 
can I join any kind of FCE task force or at least get in touch with 
others struggling with the same beast right now?

Would have loved to post my words outside an indexed world of search 
engines, but for people like me this list is the only place available 
for discussions of essential matter. I spent too much time with TYPO3 to 
have any defeatist motives -- which would take all my personal invest at 
the risk of total loss.

Kind regards,

Thomas Skierlo



More information about the TYPO3-english mailing list