magento-hackathon/HealthCheck

Why not join the force with Hackathon_MageMonitoring?

sylvainraye opened this issue · 17 comments

We developed a similar extension to get the overview of a Magento installation. You could probably merge your project with this project https://github.com/magento-hackathon/Hackathon_MageMonitoring ?

What do you think about?

I have known about Mage_Monitoring, but I think the biggest difference between these projects are the following:

  • we provide a flexible plugin-framework without observers or widgets and everyone is able to hook their extension into our framework and our checks or add checks on their own - even without telling us
  • we want to give a kind of quantitive info about the shops state (product constellation, custom reports, template/css/js structure and quality), not a statistical approach like the features I know of your project, which are certainly not all features ;)

That said, I for myself (which is just a quarter of our team ;) ) would never refuse to join these projects, but I think it will be very difficult, since the different architectures and approaches. Furthermore we think of also delivering an API and make our project be able to act as a "nagios for Magento" and I don´t know how this all will combine in a nice way. :D

If we find a proper way of combining - why not? :)

What do @achimrolle @trickzZ and @BWunderlich think about this? :)

Oh, if you talk about joining forces, you should also look at https://github.com/liip/LiipMonitorBundle and https://github.com/zendframework/zenddiagnostics even if not directly connected to magento.

@Flyingmana thanks

@kortwotze I think both extensions are complementary and we can take advantage for both parts. I see your point regarding the flexibility and regarding it we had a better approach in the develop branch thanks to @edannenberg at the commit magento-hackathon/Hackathon_MageMonitoring@2bb195a
The branch is still a work in progress but tabs and widget are configurable via config.xml from any modules and new one can be added w/o big deal.

About the kind of data you want to provide, it's fine for me. Both projects are open source and it's free to be improved or extended. Everybody is welcome to join and enrich the tools in the way they wish, thanks to that all developers can benefit of it. It means all ideas or concept of your project could also be integrated into MageMonitoring, it goes in the same direction, providing information about the current Magento installation.

I'd like to combine both projects to not have to decide which one I wish to install :-)

It's up to you and also I'd like to get some feedbacks from @edannenberg, @Zifius and @yaro-atwix

I agree with @diglin. When i saw this module yesterday i was scratching my head a bit as they are so similiar in what they try to achieve. Should be pretty straight forward to merge them into something both sides are happy with. Maybe arrange an irc session?

@kortwotze that kind of flexibilty has been the main focus for us as well in the last iterations, and besides some cosmetics it is pretty much done in dev branch. Implementing 1 class with minimal config.xml declaration enables any module to add widgets, watchdogs or even new tabs. Adding generic magento api support for watch dogs in magemonitoring should be fairly trivial. Besides that we pretty much provide the same feature set.

@diglin @edannenberg Well okay. Thanks for the update. I guess, it´s pretty much obvious to join forces. Otherwise it´s going to be a "who is prouder of his framework" (with a REAL BIG wink ;) - don´t take that personally please :) )

About the data - I think if we really join forces, it´s not important - we don´t have to specialise on a certain type of data. We could still do "our" tests/checks about compositions, or recommended backend-settings or whatever we want - it´s a framework-thing. I don´t care if my data is displayed by one or the other framework, because it´s the data that is relevant to the users.
But I think @achimrolle will be sad about his piecharts and barcharts and fancy charts. I will be as well - because they are beauuuutifuuuuuuul ;)

If I remember correctly, @BWunderlich wanted to provide a certain kind of data whith this project and the framework which we´ve done is more or less sidework to display that data. But - I still want to see the others opinion about this ( @achimrolle @BWunderlich @trickzZ )

For me - an IRC Session or a hangout or something would be nice to talk about that.

To reinsure you, I like too how you implemented the charts, just keep it :-) I think we can adapt our own chart implementation if it doesn't bother @edannenberg

Yep wait the opinion of your team and then if it's ok, let's talk about that in IRC or hangout.

I think it would be fine to match both together.

Please talk to @kortwotze about the date for IRC Session.

Sorry i missclicked the button.

@diglin I think - if we join - we should create/use one framework for all checks and data-display. So, if you like our approach, what about adapting piechart.js into the resulting framework?

Just had a quick look at devexpress chartjs, pretty similiar to what we used (chartjs.org). However it seems to support legends which is a big plus. On the downside it has a commercial license:

"You must emphasize to end-users of your non-commercial product, application and website that PhoneJS and ChartJS are not free for commercial use and you must provide a link back to this website so your users clearly understand the terms of our license."


How about tomorrow evening or saturday afternoon for an irc session?

Well, its not free for commercial use - and we don´t provide any commercial use with this extension. Anyway, @achimrolle decided to use it ;)

My weekend is pretty packed so far, but my opinion on this should be pretty clear by my comments here, so I am not that desperate to take part in that. ;) I still want to see, whats coming out of such a meeting. :)

What I want is an opinion of my team-members on that matter and my voice would go for an integration where none of both teams will loose too much of their own work. Because we, same as you, put a lot of heart in this little thing ( even if its just a two-days work ) and it´s generally kind of sad to see it go away into another framework.

On the other hand, the fact that we shouldn't force anybody to choose between our extensions, is undeniable a thing. And I think it would be totally awesome to create THE monitoring/status tool for anyone. Install, click "dashboard" and immediately see, what your shop is about, you know?

I think we would walk into a grey area though if commercial magento modules would want to ship with a plugin/widget. Just something we need to check, legends for charts is something i really miss in chartjs.org so i'm all for a switch if we don't need to plaster their link all over the place.


I think we are on same page overall, splitting our efforts into 2 projects that essentially do the same thing makes little sense. I understand your feelings regarding 'lost work', how about we sit together and take the lessons learned from both projects and do a amazing v2.0 with all the goodies from both?

Sorry for being not as active right now. Lots to do this weekend. I am still doing some stuff for Healthcheck to improve it and therefore to improve a possible v2.0. :)

Sadly, none of my mates is saying anything here - @BWunderlich told me, he is super busy at the moment. But @achimrolle should say something. :)

Hi, no worry. I'm also quite busy. I would like to try the integration this week end. Just by taking the existing code of both extension and merge it. Then we can do the improvements. I hope I can do something today into a new branch of the MageMonitoring.

Thats fine for me - let me know, what you achieved or if you have any questions. :)

Thanks for the update. That is very nice. I just want to say, I have severe problems at the moment concerning my job and my apprenticeship - I am not sure how and when I can contribute to this again. :(

Hopefully at the next hackathon that I attend, maybe in a month or so.

Hi @kortwotze, no problem, that's the reason I have also seen the reason to merge both extensions, to keep people engaged.

Good luck for your job