[TYPO3-v4] The future of ExtJS 4 in TYPO3 4.7

[Christian Hennecke]

Am 11.01.2012 16:47, schrieb Oliver Hader:

Hi Oliver,

please see below.

[snip]

> That means, that we have these possibilities:
>
> a) postpone the final 4.7 release a lot and hope things will get better
>     - forward port all components to ExtJS 4
>     - introduce a great backward compatibility layer for
>       ExtJS 3 extensions
>     - create documentation on how to use ExtJS 3 and ExtJS 4
>       side by side
>
> b) revert the ExtJS 4 changes in the Git master branch of the TYPO3 Core
>     - get back a stable TYPO3 4.7 with working Workspaces,
>       Form Wizard and Recycler
>     - accept the "collateral" fact that the work on ExtJS 4
>       migration is lost, which means to disappoint developers
>       who invested a lot of time in what already has been achvieved
>     - find (better) alternatives for the mentioned "media" and
>       "vidi" components
>     - find alternatives to ExtJS in general in the long-term view -
>       however, ExtJS is good for widgets, but not as a one-fits-all
>       JavaScript solution
>
> It's really understandable that it must be frustrating for somebody
> having put a lot work into the ExtJS4 migration by reading these lines
> above. However, the end of these tasks cannot be predicted and as
> always: time and resources are limited.
>
> So, our suggestion is to revert the ExtJS 4 parts and continue with a
> clean TYPO3 4.7. However, if somebody out there has a good alternative
> on how to proceed with ExtJS 4, please share your thoughts and ideas. If
> there are no major objections and no solution that really saves us, we
> will proceed very soon with reverting back to ExtJS 3.

Some thoughts from an integrator's and extension developer's perspective:

- Stability is key. But we have TYPO3 4.5 with long-term support, and 
support for 4.6 could be extended. So a prolonged development phase 
wouldn't be too much of a problem if the reason is communicated clearly.

- OTOH, a long wait for 4.7 could be perceived as stand-still.

- Improved support for HTML5 stuff in the near future would be nice. 
While we can do without it for many projects, it will be mandatory for some.

- Judging from the "hope" bit in your statement, following the first 
path wouldn't necessarily guarantee the desired outcome.

- Would a large compatibility layer defeat the cleaning done with 4.6?

- I am a bit worried about the effects the whole thing might have on the 
BLE project - and possible similar future projects.

Personally, I'd prefer the second route. Advancement might be slower, 
but there would be some. And it would provide time for in-depth 
evaluation of future perspectives.

Cheers,

Christian