All qa1.7 needs is more hooks! Lack of hooks makes plugin wrting sometimes impossible

Question

I would suggest asking developers what hooks they need to be provided to them for qa1.7 instead of asking users what features they need. Developers can add many features provided hook sare there!

There are many things we cannot do with current architecture examples:

1- hiding the regular registration and keeping only facebook/social login visible (usage: to avoid spam)

2- applying stying to notification messages being displayed to users

you name it...

My point is that now that time and resources are limited lets spend it on more important stuff, Architecture not the look and feel or new features ( because time is limited for the core developers of this qa, Scott and Gidden)

 

Answer 1

I believe this post could be subject to discussion.

Firstly, I do agree on the fact that hooks are missing. However, I don't fully like the part about asking developers. That is definetly something developers should evaluate, modify the core in a generic way (not just to satisfy their need but rather to satisfy a more generic need, if possible) and then create a pull request with the change justifying it. EG:

https://github.com/q2a/question2answer/pull/16

https://github.com/q2a/question2answer/pull/61

 

Those refer to hooks particularly as firing events. However, there can be other hooks such as using different new plugin types (eg: plugins to control point assignation or AJAX calls, which they just don't exist), adding (a lot of!) new template methods to the qa-base-theme.php file (which is considerably related to my advice of creating an API-like core that doesn't output any HTML at all http://www.question2answer.org/qa/31514/what-do-you-want-in-question2answer-1-7?show=31529#a31529) or even providing new plugin overrides to some functions that they just aren't overrideable and there is no justification to that (EG: https://github.com/q2a/question2answer/issues/60).

 

Secondly, and in this I fully agree with you, providing better support for developers would increase plugin quality and variety, as I mentioned in my request for v1.7. However, (there is always a However) I'm not sure users would like a new Q2A release so focused on architectural changes. Users are looking for new features and if they can't see it then it is useless. In the short term users won't definitely like that. In the long term they will but they just are not aware of that. Many users just don't know what the core is or how it works that's why some requests like max's http://www.question2answer.org/qa/31514/what-do-you-want-in-question2answer-1-7?show=31517#a31517 (which I do believe it would be great to be able to vote comments, by the way) get so many votes and other more technical just don't get voted up. That's why mixing user feedback with developer feedback is not good (which is my #7 advice in that post).

 

Thirdly, I don't quite get the hiding the logins part. Do you mean just hiding it or really blocking it? I guess hiding could be done through the theme or a plugin layer. Blocking the login itself, I think, it could be done with a process plugin. Regarding the notification styling is clearly an issue that would be solved by decreasing the amount of HTML output by the core so that it delegates that to the theme/plugins.

Comments

Thanks for your answer. In response to your "secondly" point I should say the only reason that I said hooks are more important is because of lack of time that Scott is experiencing. I was saying it is more important to spend time on architecture than feel and look when time is limited :)
Secondly: I would suggest shorter answers, so we spend more on development not discussion. Some action, some deed ;)

---

also:
1- hiding the regular registration and keeping only facebook/social login visible (usage: to avoid spam)