some thoughts re roadmap

[valan]

Hello PW-friends!

 

Let me share some thoughts re roadmap that emerged together with the development of mymechanic.ru site (currently 8K clients/30K visits per month). Initially it was a simple "info" site and PW was ideal for it, but when business has grown, I start to miss some functionality. I understand that this functionality may not be necessary for current PW sites or even PW concept, but anyway I would be grateful if you could share thoughts on the following:

 

1. simultaneous page editing capability. There are several roles that work with page and I have to remind them about queue in order to ensure that edits are saved properly. It is not effective. To unlock the capability, all fields and even admin layout should be able to update in "real-time" as soon as there are changes at server. Also there should be field-level lock/sync mechanism. I guess it is big core code refactoring.

 

2. event-driven admin (page editing, etc). With observers, controllers, routers etc - all you need when you develop frontend SPA. Of course, this can be done w/o PW internal features but I see great opportunity here for PW. By the way, check this framework - http://riotjs.com/ I believe it aligns well with PW ideas of best performance and simplicity.

 

​3. advanced filtering in listers. E.g. the dream is filters like this: {myfield1.subfeild1.subfield1.date1} > {myfield2.subfeild2.subfield2.date2} - with indefinite levels depth, fully customizable on left and on right sides of condition, even with PHP snippets option. It is something that actually I haven't seen anywhere else - something really unique and powerful.

 

I also have some other thoughts but I don't want to overload. These are top-3 in my mind. Thanks.

[LostKobrakai]

Thanks for your thoughts valan. I can certainly see where you're coming from as I'm working on a similarly scoped project, even if it doesn't currently have anywhere that number of users. I think your first point is quite a thing to get to with processwire and it's philosophy of not deleting / overwriting content without explicitly being told to do it (hitting the save button). Introducing functionality like you're describing would certainly need some kind of core field versioning (maybe even event based). This would also need to be deeply considered with the database performance/choice. 

Event drive admin sounds fun, but it's also a new hurdle for new comers. With a core js api on the roadmap it might become more easily to implement any custom framework into the admin on your own. If there will be a admin js framework to be introduced I'd rather root for vuejs, which from my perspective does have the greater community and drive behind it. But that's to be determined if it comes to that point.

Deeply nested selectors would certainly be a great addition.