0.9.7 Manager templating

@kudolink : you can get all necessary information about donwloading and installing 0.9.7 SVN here : http://svn.modxcms.com/docs/display/MODx097/Development+and+Testing+Notes

Now, back on the discussion at hand.
Lots of feedback here !

I certainly can understand your thought process and I am sure many people will raise those questions when we go into public alpha. Heck, I might have raised them if I had not kept a close eye on 0.9.7 development.

I think Ryan just raised some key issue on why PHx might not be the best tool for the job. Don’t get me wrong : I love PHx and I use it extensively (@ryan: I am very interrested to know how PHx killed one of your websites...). I am pretty excited that 0.9.7 will make PHx filters and conditionnal statements a native feature of the MODx parser. Conflict with other plugins or snippets such as eForm or Jot has shown that using PHx is not a trivial choice and sometimes requires some hacking and that’s never a good thing in the long run. Now let’s not shoot this great addon : as I have said, I use it extensively. I am just glad that 0.9.7 parser will allow for similar features, since it’ll avoid those conflict issues.

I completely trust our dev team to make the technical choices for the new branch, and that includes many of those aspects which go beyond my skill scope. MODx has always been about flexibility, and you’ll see that with 0.9.7+ it’ll get even more flexible than it already is (now my head spins at the concepts behind the new branch but it also did when I first came to MODx : get a firm grip, hang on and you’ll love the ride )

I realize we can’t count on people giving us a vote of confidence offhand and I think pushing us towards more clarity on this item is a good thing. Many question will arise, and the process of explaining and advocating the key choices behind the new branch will also help us be more efficient marketing it. Again, thanks for pushing us !

Back on the manager templating.
Even if I am no dev your post helped me get new arguments :

• Most web apps have no backend templating whatsoever : MODx will be one of the few opensource CMF to offer easy templating of its backend. Check out most CMS/CMF and you’ll see that templating is strictly a frontend feature. Backends have very poor separation between logic / presentation, they’re a mess to customize. As a freelance webdesigner, I can certainly see how having a customizable manager can greatly improve my selling pitch : how about a tailored manager to exactly fit my client needs (and/or skills) ?... wait a minute, that should be great !!!
• Frontend templating and backend templating don’t have the same requirements : I think if frontend template engines were able to handle backend requirements, they would be used for backend. I don’t know of any web app which uses its own frontend template engine to power its backend. But I do know web apps which use Smarty in the backend (activeCollab is an example, and that’s where I started learning a bit about Smarty). Frontend is mainly about content and displaying, Backend is about processing and logic. I am probably oversimplifying and I am unable to dissect it but that how I intuitively see it.

With Smarty, I am confident we’ll be able to do things which we tried to do with the ManagerManager plugin for instance, and it will greatly help making the UI better.

About having a purely CSS manager with minimal JS, funny you’d mention it... Jay has raised the idea to gather a team to do just that and even more (add a focus on accessibility on top of what you had in mind).

Quote from: davidm at Apr 21, 2008, 08:34 PM

About having a purely CSS manager with minimal JS, funny you’d mention it... Jay has raised the idea to gather a team to do just that and even more (add a focus on accessibility on top of what you had in mind).

Yes. I think that there is a place for a wholly CSS/HTML interface. Jason (OpenGeek) has pointed out that it would not be impossible but is not the focus of the current team. Those who are skilled UI designers and know php would be more than able to write out the interactions using whatever engine/templating system.

currently I can select the manager template from 3 different templates.
If smarty makes it 4 that’s a good thing.
the more the merrier!

Quote from: rainman at Apr 22, 2008, 01:25 AM

currently I can select the manager template from 3 different templates.
If smarty makes it 4 that’s a good thing.
the more the merrier!

The thing is that currently the manager templates are not fully modifiable. They are essentially skins as the legacy etomite stuff has alot of the presentational output entwined in the core. In 0.9.7+ the layout, position, element struction and page structure will be completely (I think) isolated from the core. Essentially, the only connection to the core will be via the Smarty engine over EXTJS (unless you want to roll your own) which is why this is such a dramatic change from previous versions.

The old MODx manager was essentially like CSS Zen Garden you’re dealt the output and have no conrol other than presentation (except in this case there is legacy 90’s-esque presentational HTML in some places). The new MODx will have the capacity for you to remove elements, change document structure and more (beneath the presentation) without setting foot into the core.

First of all, I don’t want to look contentious, probably is only my translation and English skills. I hope this debate looks like a sitting-on-the-sofa-with-a-beer thing than a flame (which, by the way, has never been intended to be).

@Ryan
I know my words cannot be considered to change the development course for 0.97, far from believing this and is not the point.
The understanding of your views is ways more important, because I’m deployng sites with MODx now but in the future it has to grant me some kind of consistency in its life, I don’t want to find me changing all my sites to another CMS from one day to another because you (plural) took a road that’s not what I need and I put some questions because I want to understand where are you leading us.
And this includes how do you think to face situations. I only wanted to understand where are we going, not more, not less.
And no, I don’t want to get rid of PHP... I asked if you want to (that was a paradox, I’m sure you not).

On the other side, I’m sure thing like PHx or other things we use have flaws and similar, but I see this as a reason to keep working, and enhancing, and creating. Not all the bad things needs to be stripped away.
MODx as it is now has feeded a very creative community, we have alternatives to choose when we face a need, we can count in people creating/testing/enhancing snippets, and modules, and plugins. PHx is not good as Smarty? We can improve it, shaping it around MODx (it’s only an example). We’re all here to work.

Obviously the manager JS is not the same as the frontend, I know that, I didn’t understand what you mean. I was talking of the manager and the possibility to use another JS framework IN the manager.
The only thing I can tell is that Ext can’t be so essential. It’s one of those thing in the web I find more cool than useful, sincerely when you say "you can customize the manager" I hope Ext will be strippable, but I don’t think so.
I fear you’re rising the learning curve without meeting a need. And you know this will have an impact on the community.
Have you taken into account, speaking of time, how much support it will need after all of this implementations before the community can effectively pick up the enhancements and start using them to build something better? It will be simpler or will require more technical skills?

But I want to trust you, because MODx is very good and I hope who have done such a good work can do even better.


@davidm
I’ve always been very near to your words in every post I’ve read by you, and so this time.
But let’s specify flexibility: for me, flexible is what gives a simple base to build on top quickly. That’s why I talk of a simple CSS manager, which a (optional) processor for ajax calls eventually. That is "flexible" in my view.
This is why I argumented that I’d see Smarty as a external plugin, or I can understand it if you tell me that is part of the "default manager template", and not the "visual presentation layer" of the manager. That mean I’ve to face Smarty, and I don’t understand why when we’re yet able to display a web-page with what we have, and more significantly we yet know and can control.
(I think this time I’ve explained the concept very near to what I mean)
I can only, as I’ve told to Ryan, trust you as you’re a freelance like me and you understand my needs so if you (someone in the process, which has a better view than me on the actual state and the thoughts that bring the development where it is) say it’s good I can’t do any better than trusting you.
I’ll pm you for the 0.97 branch.


@smashingred
There are many reason to have a whole CSS/HTML gui, I completely agree with you. I think about portability (I made sites which you can surf on a phone: why not manage?), the yet discussed customization, to have a very light (no or less JS to download - with Ext is simple to go over 300kb pages!) alternative, or accessibility.
Take me into account if you want to work on this and lack workforce, ok? I promise I’ll be way less polemic as I look like now


Thx for the replies guys.

Personally, I want an iPhone interface to the manager.

Kudolink, here’s an example for you: take a site that has more than 1500 pages running a hacked together development proof of concept version of MODx 091 + security patches. Run it through the conversion process, and approximately 20 seconds and a 10MB log file later you’ve got said same site running in 097. Granted, a couple of things broke, but those were things that needed to be fixed and made into "proper" solutions even in the original version and that took minutes, not hours to fix. If you wish to know more about what site I’m talking about please IM me, or better yet grab your headphones and set up Skype and we’ll have a conversation. I’ll treat you to my special redneck-version of English.

Two things on which we refuse to compromise include maintaining flexibility (i.e., continuing in the same spirit that MODx is know for) and on releasing things with major breaking changes between versions without adequate and dire warnings. This last requirement has probably contributed to a 6-month delay in 097’s public preview. That’s not to say that the conversion utilities will be released simultaneously with the preview, but they’re an important part of the overall plan to make sure we keep folks happy!

I don’t understand what’s this all about..

In current MODx branch, it’s not possible to template the back-end at all.. logic and presentation are mixed in various files. You have to use systems like ManagerManager, CSS changes (for small modifications) or hack the core if you want to change something in the back-end.

In 097 the back-end is fully templateable by default via Smarty. So it’s a big progress.

And in 097, it’s possible for us as an community to develop new back-end interface(s) that use for example MODx templating engine and just HTML and CSS withtout any JS..

What’s the problem?

It seems I hit a nerve! 


Your hidden message was: MODx 0.9.6.1 addons are not compatible with 0.9.7. I need a confirmation!
It doesn’t make sense developing software that modifies the core. Nobody trusts software that modifies the entire system, when it is beeing developed by a one man army.

However my problem is not the templating engine, please don’t flame bait with smarty pro/contra when you talk about hypthetical things like it’s integration in the future.
I have no opinion about smarty in combination with MODx, as long as it is not seamlessly integrated into MODx Framework. (I like both systems, both have pro/contra)
But let’s focus on the core itself. Now you see that a cmf should be flexible enough to manage itself.
Means code can control the view of a back- or frontend (and maybe other things too like config etc). Seperation between core and user.

Can the devs please provide information about the next modx version like release year and new features, or structural changes.
You can post a link here, that would be very nice.


(Something that is very important IMHO is that an accurate user-interface needs high fault tolerance and weak artificial intelligence, like autosuggest or -correct et cetera.
Just little examples: "invalid html" ; "wrong config" ..)

. . . . . . . . . . . .
a dot is a dot.
. . . . . . . . . . . .

Xsss4hell, the best places to get lots of info on 0.9.7 are the MODx Next and xPDO

Quote from: Xsss4hell at Apr 27, 2008, 09:39 PM

It seems I hit a nerve!

First, you are taking credit for kudolink striking a nerve, though he didn’t strike a nerve IMHO as much as bring up very valid questions that probably need to be moved to a unique topic. This discussion hijacked your interesting original topic for sure, but I don’t see how any of this relates to your original post here, and it certainly wasn’t flame baiting.

Second, can you please refrain from using increased text size in your posts for no apparent reason. It makes me not want to read your posts at all (i.e. I guess you have stricken a nerve after all).

Quote from: Xsss4hell at Apr 27, 2008, 09:39 PM

Your hidden message was: MODx 0.9.6.1 addons are not compatible with 0.9.7. I need a confirmation!

There was no hidden message. This has been discussed for quite some time, and a migration process is being well-defined. You can choose not to migrate based on how many changes might be required in your custom or 3rd-party contributed add-ons and/or other factors, like how important the new features are to you.

Quote from: Xsss4hell at Apr 27, 2008, 09:39 PM

It doesn’t make sense developing software that modifies the core.

I agree, I think...

Quote from: Xsss4hell at Apr 27, 2008, 09:39 PM

Nobody trusts software that modifies the entire system, when it is beeing developed by a one man army.

Many people contribute to MODx (it’s called the MODx Team), so I don’t appreciate the misleading statement.

And I imagine there will be some resistance to the major changes coming; people generally don’t like change. As for trusting the decisions of the MODx Team, well, that’s for the MODx community to decide. Regardless, there is always an adoption curve for anything new, so I’m sure the MODx 0.9.6.x branch will continue to live on for quite a while even once 0.9.7 final releases are available. We will likely even allow continued bug fixing and limited feature development on this branch if someone wants to take on the responsibilities involved in doing so.

As always, you are free to donate your time/contributions to the MODx project in whatever way you feel you can. If you show promise in your contributions here in the forums, in code, or otherwise, you’ll be invited to join the "army" based on the consensus of the MODx Team.