Upgrading ACP's core code.

[3Di]

Let's talk about "re-inventing" (core code) the ACP before to re-styling it.

I am aware of the discussions already made here and there and the archived RFC and everything, so please do not point me to those. It's time to re-code the ACP .. at least for 3.3.x?

What about? There are plans we are not aware of?

[DavidIQ]

I'm not saying that it should or shouldn't be done but what would be the point of re-writing the ACP? There should be a goal we want to reach if we were to do something like that.

[3Di]

I'm not saying that it should or shouldn't be done but what would be the point of re-writing the ACP? There should be a goal we want to reach if we were to do something like that.

I want to turn your question to yourself and place it under another point of view.

Let's try.. do you think the ACP core code does it belong to the main one?
I mean, don't you think that the ACP core code is buried in antiquity?

Feel free David to split this posts into a new topic with (please) a different title, I will be more than glad to partecipate on it.

[DavidIQ]

I've split off the topic now.

And seriously? You answered a question with a question? You're the one that brought up wanting to rewrite the ACP so it's up to you to provide reasons why it should be done. You can't just drop a bomb like that and just expect everyone to be in agreement just because.

So according to you this should be done because the code is "old". What is old about it and how would you improve it? What other reason(s) would you have for wanting to see this done?

[3Di]

I've split off the topic now.

Kind of you, thx.

And seriously? You answered a question with a question?

Yes I did, I know that's not "ethical" but I hoped me to read between the lines.

You're the one that brought up wanting to rewrite the ACP so it's up to you to provide reasons why it should be done. You can't just drop a bomb like that and just expect everyone to be in agreement just because.

I am not actually wanting to do this nor somebody else for "me", I would like it to be done though for phpBB. Indeed.

So according to you this should be done because the code is "old". What is old about it and how would you improve it? What other reason(s) would you have for wanting to see this done?

This problem affects also extensions, the way we have to code those. We have to use a type of coding style for ACP and another one for the rest of.. to be more accurate I can say it is not normal we have to globalise variables instead of using objects.

All of this affects phpBB therefore the way extension writers (like me) have to follow, I am a little bit experienced but's a no-no for those are new to this world (like not OOP coders or vice-versa).
What about to be able to create new styles for the ACP then? It's trivial I know.. but?

All of this being said, it is under the vision of every savyi coder the ACP is buried somewhere into the ancient thinghs, It's not a new nor a breaking-news. Sure it's not a bomb.

edit: I corrected some typos, in the meantime.

[MattF]

This problem affects also extensions, the way we have to code those. We have to use a type of coding style for ACP and another one for the rest of.. to be more accurate I can say it is not normal we have to globalise variables instead of using objects.

This is a problem.

The ACP, MCP and UCP are all decoupled from Symfony components. That means they do not support the container based DI system, so they require using things like globals, trigger errors and other things that are no longer considered PHP the right way.

And as 3Di points out. Writing ext code for the modules requires going backwards in time, to antiquated phpBB coding. We can't throw exceptions (and have a nice error page like outside the ACP), we can't use Symfony's Response object for simpler redirects and AJAX, etc.

There are plans to bring the ACP up to speed, however. But it's a lot of work and will take time.

[CHItA]

There are plans to refactor the ACP (as basically everything else that is in a front-facing file). Probably when someone will have some architectural plans, then you can expect to see an RFC posted somewhere in this forum.

As VSE said, the *CPs are pretty big beasts so it may take a lot of time/multiple major releases, however, once we implement the basic infrastructure extensions would probably be the first ones to be able to take advantage of that (as every extension author could refactor their own code only, to take advantage of the changes if they wish).

[hanakin]

also the ACP is the last of those three VSE pointed out that need re-worked as the UCP & MCP are seen by everyone and not just the board admins. The changes also hold back progressive enhancements that can be made to the design/interactions in a new front-end theme as well. AJAX!!!!!!

[3Di]

also the ACP is the last of those three VSE pointed out that need re-worked as the UCP & MCP are seen by everyone and not just the board admins. The changes also hold back progressive enhancements that can be made to the design/interactions in a new front-end theme as well. AJAX!!!!!!

Hello, well.. no need to post so many exclamation marks to enforce something that BTW it's already easily understandable. Any way.. let's talk about what? First, I wanted to point out that the issue doesn't just belong to phpBB itself but the way the extension writers have follow in order to accomplish something that can be considered a kind of mental torture, so to speak. If you reckon with this, you just should agree with us (at least VSE and me).

Regards.

[3Di]

There are plans to refactor the ACP (as basically everything else that is in a front-facing file). Probably when someone will have some architectural plans, then you can expect to see an RFC posted somewhere in this forum.

In fact the ACP is a front-facing file, technically speaking..

hanakin?

BTW thanks for the heads-up.

As VSE said, the *CPs are pretty big beasts so it may take a lot of time/multiple major releases, however, once we implement the basic infrastructure extensions would probably be the first ones to be able to take advantage of that (as every extension author could refactor their own code only, to take advantage of the changes if they wish).

Well, this bit about multiple major releases it's something discouraging.