Hi Victor,
Thanks for your excellent feedback. And also many thanks for your previous consultations on this (you had a very good work-session at JAB12 on that that I came too).
Here a few replies to the good questions you raised:
- The Goals section of the documents outlines the goals of the implementation, indeed not the goals of having "in-app installation", as those were expressed in the project's 2013 goals.
- As I see it, the main goal of such a simplified "App Store" (name to be changed ofc) is to make the life of the Joomla user much easier when he wants to install one or a few extensions. It just brings back Joomla to a leading position in that area. So the gaol in a few words is: "Simpler and friendlier extensions installation for the user". It is a user-centric approach. Side-effects are probably less frustration of the user (hopefully), better usability and better evolution of Joomla in Google Trends, and this should help to benefit the whole Joomla Community, not just users and extensions-developers.
- Terminology: That's why I'm proposing a "low-profile" terminology, like e.g. "Install from web" in the user-interface.
- I believe that the 3 basic models expressed in the document cover probably 99% of the extensions. Special model should be possible too. The user is (after review of the information provided) redirected to the extension developer website, if the download is not available for free without registration, and from there on all scenario are possible imho. We will also make use of all the Joomla great ways of extendability, so that new uses and models that we didn't think of should be possible.
- The "App Store" will not support updates, as those are already supported by Joomla or by third-party libraries, independently of the installation method.
- We are 5 weeks away from Joomla 3.2 beta 1, but there are some more weeks before Joomla 3.2 stable, so that there should be enough time to test and fix all bugs. Keep in mind that the Joomla CMS part is very reasonable in function and size. The bigger part is the server-part running on
joomla.org, and that can be very simple too in the first release, and then extended on its own schedule during the lifetime of Joomla 3.2 and 3.5. So the implementation, testing and bugs-fixing delays are reasonable imho.
- Once 3.2 beta 1 is out, all features are frozen until Joomla 4.0 in more than one year from now, and also the 3.5 LTS release will be up for another around 2-3 years. So that's why I believe that it is way better to provide that new feature now in 3.2 beta 1 in a modest and pragmatic way, and then to be able to extend it server-side as needed during the whole life-cycle, as well as to fix/improve minor issues CMS-side during minor releases, than to aim to a perfect solution for 4.0.
Open-source works best with small reasonable steps released more often, and that's the aim here with catching the 3.2+3.5 "train" before it has left. :-)
I'll not change the document under review during the review process here so that everyone speaks about the same document and not a moving target that needs to be re-read every day, but I might add a last, marked as "live", section based on feedbacks and replies in here as time permits, and integrate the changes more up in the document after the review here.
Best Regards,
Beat