Splitting ES up
By CK on Friday, 19 December 2014
Posted in Feature Requests (Not Supported)
Replies 10
Likes 0
Views 720
Votes 0
Hi Guys,

I've been meaning to ask this question for a while but haven't gotten around to it.

With every release, the extension becomes a bit larger and more complex, how about breaking it down to a couple different extensions and have a single core extension/framework that the other extensions are built one. this would help in 2 way

  1. Easier to make changes (add/enhance features [new extensions], implement bug fixes) to the whole suite of extensions.
  2. Less code needed on the server (reduces maintenance and attack surface).


Let me highlight an example. If I want to run a site that doesn't make use of groups and events, instead of turning off the feature (and keeping all the extra code on the site), I should just install the core extension, if in the future the needs of the community change and they need an event system, I install the event "plugin."

I'm not suggesting this for the immediate future, rather something to have in mind for future release (think about adding the videos and pages feature, instead of having to modify the entire extension, you can write an extension that plugs into the existing one).

What are you thoughts?
Joomla 3.4 I believe is going to have a similar method to what your describing. This has been on my mind too. I personally plan on using most of the built in major features of ES but know that many would like a way to go lighter weight. Perhaps the install package could start out as a whole, but then allow sections to be removed. When upgrading the system could check to see your preference and not reinstall those sections upon upgrade.

The good news is that once videos and pages are out, ES should theoretically not increase major sections of the component. Of course optimizations and little bells and whistles will still be wanted.
Josh Lewis
·
Friday, 19 December 2014 11:20
·
0 Likes
·
0 Votes
·
0 Comments
·
I would not vote for this because maintenance is going to be an extreme nightmare
Mark
·
Friday, 19 December 2014 15:28
·
0 Likes
·
0 Votes
·
0 Comments
·
I wonder how Joomla's going to pull it off? Your response is certainly understandable. We would not want to create a nightmare for you folks.
Josh Lewis
·
Friday, 19 December 2014 15:31
·
0 Likes
·
0 Votes
·
0 Comments
·
I definitely don't ant to create a maintenance issue, I still think that it should be considered (maybe for version 2), to the best of my knowledge, Project Fork and Community Builder works like this.

I also think that it might help speed up development, as you can focus on managing the individual "plugin" without worrying about "breaking" the entire extension. Just my thoughts.
CK
·
Friday, 19 December 2014 21:32
·
0 Likes
·
0 Votes
·
0 Comments
·
Thanks for the heads up on this Chaim
Mark
·
Friday, 19 December 2014 23:48
·
0 Likes
·
0 Votes
·
0 Comments
·
I wish ES breaks things up. It's just getting bigger and bigger and I only use 50% of the features it has to offer. Disabling the unneeded features doesn't reduce the resources since they are all compiled in one big bundle!
Neel
·
Sunday, 21 December 2014 14:18
·
0 Likes
·
0 Votes
·
0 Comments
·
Thanks for the heads up on this Neel. Unfortunately not everything can be broken down at this point of time as I am afraid it would be more of a maintenance nightmare for you guys and for us as well.
Mark
·
Sunday, 21 December 2014 16:29
·
0 Likes
·
0 Votes
·
0 Comments
·
Mark,

How about removing the migrations from the core install (or only install the migrator if the product is detected - such as JomSocial)? Another item would be integrations, why not move them to a plugin?

Thanks,
Chaim
CK
·
Wednesday, 08 April 2015 06:52
·
0 Likes
·
0 Votes
·
0 Comments
·
@CK Yeah, that would be good. I could understand that compatibility could break with upgrades due to it being separated from the core, however they could be fixed as needed. After all it doesn't appear that compatibility with ES plugins/apps breaks very often even with a lot of upgrades.

A lighter weight EasySocial would indeed be cool to see (we know that it will be heavy no matter what, but the lighter the better). I do however understand that if this somehow makes support trickier to deal with. CK is talking about the small stuff (for example I don't use K2), not the major components of ES which Mark already established that it would be a huge hassle.
Josh Lewis
·
Wednesday, 08 April 2015 08:15
·
0 Likes
·
0 Votes
·
0 Comments
·
Thanks for heading up on this CK and Josh, we will see what we can improve this in the future.
Arlex Wong
·
Thursday, 09 April 2015 01:54
·
0 Likes
·
0 Votes
·
0 Comments
·
View Full Post