Development updates?

So bugs.modx.com is no longer being used, in favour of github which I'm sure has been in use for some time. But in spite of the implication that the modx team are now using github more explicitly to track/report issues, it seems quick clearly that it in no way replaces the most interesting functionality of the old system: an indication of progress towards a new release.

I have a couple of projects which I've had on hold for a while now, only really because I thought 2.3 was just around the corner. Both projects are now becoming more urgent so I imagine I'll have to press on with 2.2.10 and simply upgrade later on.

To be clear, this is in no way a complaint. I sympathise with the developers of good, open source software, the time and dedication it requires, and often the time they must spend doing more financially urgent things before committing yet more voluntary time to core development. In fact I empathise because I'm in a very similar situation myself. So I don't mean to rush or heckle. The reason I'm posting now is more fundamental, and in an attempt to improve things for the community, if that's not unreasonable.

The last time I contemplated posting on this forum to ask about development progress was because I was concerned that bugs.modx.com wasn't actually being used very well. So I suppose it shouldn't come as any surprise that it has been dumped entirely now. What was most irritating (perhaps too strong a word, I mean no offense by it) was that while issues did seem to be tracked using that system, not *all* issues or new features made an appearance. But more importantly, they seemed not to be properly 'completed' (ticked off the list) in a timely fashion. Instead, I'd wait a while for some indication of activity, there'd be none, and then all of a sudden - a release and everything's done!

Of course, it's not the perogative of the developers to indicate any measure of progress. So I don't hold them accountable for doing so. I simple wonder if this specific area of development could be improved, for the betterment of the community as a whole.

Before I conclude, perhaps it's worth asking if I'm missing something. Is there some other channel through which progress is being reported? Am I simply ignorant? I'm excited for modx. I love the way it works and I sincerely hope that some of the most obvious weaknesses will soon be resolved (which for me at the moment is really just it's seemingly deeply engrained lack of support for any non-webkit browsers, plus it's inherent inability to sufficiently approximate WordPress categories for effective blogging). I enjoy[ed] watching the progress and to some degree I still can, but right now I'm looking at these forums and github and it looks a lot like there's next to nothing progressing with core development.

I should really reiterate that I don't mean this to sound like a complaint about the apparent lack of progress. To be clear, if little progress has indeed been made I'm with doubt that the developers have very good reasons. My plea is for an improved flow of information about any progress which *might* be happening. Perhaps something to give a hint as to how far off 2.3 really is, and whether it's worth waiting a few more days. To the best of knowledge there hasn't been a beta release, let alone an RC for 2.3, so I don't expect to see the final release for weeks, minimum.

So, do the developers discuss progress in a public channel I'm not aware of? Does github have some progress indicator I haven't noticed? Is core development taking a short break? Will 2.3 appear within a couple of months? Do the developers feel the core has reached a high enough quality to drive the mass critical to support their paid cloud option and make it all worthwhile, or has profit latency forced them to redirect their efforts prematurely?

All answers are valid, and no judgement is passed. Thanks to the devs for all their hard work and ingenuity to date, and for the existing core which I find to be inspirational.

Hey Dave, thanks for bringing up a very important topic. You're entirely right with your feelings. You're definitely heard and your feelings basically track with ours. We'll get back to you here shortly, but I wanted to quickly say thanks for taking the time to write such a well-reasoned, detailed post.

Bump!
I also have a few projects on standby mode. Would love to hear more or hear when the planned Google Chat is happening. [ed. note: eladnova last edited this post 10 years, 3 months ago.]

An update to this conversation, we're in the last stages of migrating the issues from bugs.modx.com to GitHub. Creating Milestones for each of the projects requires those issues being in our GitHub projects. We will work to improve the flow of information about ongoing efforts. It is very important for the health of the project as a whole as well as for the continued confidence people have in MODX.

I agree with Dave: I don't care how many cool kids are using Github, it is not a full-featured bug-tracker. It has VERY few useful features when it comes to issue statuses, versions, etc. and I was extremely disappointed to see Redmine abandoned in favor of Github (don't worry Dave -- you were too polite to complain, so I'll take on that role with my megaphone).

The other issues you touched on are potentially huge:

1. the non-webkit browser stuff brings up the fact that the MODX manager is fatally coupled to ExtJS. All the beautiful abstraction of the front-end designer/coder experience is thrown to the wind in the manager -- the more of it I see, the more frustrated I become. On the front end, you can do anything -- freedom! In the manager? It's a nazi-esque death march -- "VE VILL MAKE YOU USE OUR EXT JS OR VEE VILL KILL YOU." No freedom at all. Things that should have been abstracted are hard-coded, and stuff that should have been hard-coded are abstracted to a degree that could be called fanatical. I'm fully suggesting that the MODX core should be *completely* rewritten, Ext JS abandoned, and we move towards more modern dev practices (e.g. composer, flexible routing, etc.)

2. Slow development: I don't think there are enough hands on deck here, so dev progress is slow. Because some of the core stuff is so convoluted and real proper documentation is rare, the problem compounds because willing volunteers can't wrap their head around how to help out. Some of the hiccups with the MODX cloud etc. I think have buttonholed key resources in the basement fixing leaky plumbing instead of doing core development.

1. the non-webkit browser stuff brings up the fact that the MODX manager is fatally coupled to ExtJS. All the beautiful abstraction of the front-end designer/coder experience is thrown to the wind in the manager -- the more of it I see, the more frustrated I become. On the front end, you can do anything -- freedom! In the manager? It's a nazi-esque death march -- "VE VILL MAKE YOU USE OUR EXT JS OR VEE VILL KILL YOU." No freedom at all. Things that should have been abstracted are hard-coded, and stuff that should have been hard-coded are abstracted to a degree that could be called fanatical. I'm fully suggesting that the MODX core should be *completely* rewritten, Ext JS abandoned, and we move towards more modern dev practices (e.g. composer, flexible routing, etc.)

Its not that difficult to build your very own modular manager, or parts of it completly as frontend-pages, completly without using extjs. All or most of all manager-processors can just be reused without any changes.
I don't think the core itself does need to be rewritten at all.

Quote from: Everettg_99 at Jan 23, 2014, 06:46 PM

Things that should have been abstracted are hard-coded, and stuff that should have been hard-coded are abstracted to a degree that could be called fanatical.

This made me laugh out loud.

That said, I think you're really talking about the Manager UI, not the core. There's a reason that the core directory and the manager directory are not combined.

I don't think anyone currently working on MODX (including everyone from beginning MODX extra developers to the Chief Architect) is particularly crazy about the Manager UI or the current state of Package Management, but you're definitely not forced to use ExtJS for anything except modifying the existing Manager forms (and you could always replace them rather than modifying them).

SiteCheck, NewsPublisher, and MyComponent are cases in point. None of them use ExtJS at all, so you're definitely not forced to use it. NewsPublisher provides a fairly complete version of the Create/Edit Resource panel without using *any* ExtJS. MyComponent provides a Manager interface for package creation through a snippet, but it could just as well be a CMP (again with no ExtJS). SiteCheck is a CMP you use for testing the integrity of selected parts of your site, but it uses no ExtJS at all. I should also mention Notify, which appears to be a CMP launched in the Create/Edit Resource panel in the Manager but is really done in the front end, again with no ExJS.

It would be entirely possible to rewrite the entire Manager using some other JS library like Ember or JQuery, or by using plain HTML/CSS. You could do this without altering *anything* in the core directory and I often think about doing it just as a mental exercise while fighting insomnia.

Think of the Manager as just a series of HTML forms. Various actions are performed when they're submitted. You could replace any of them with your own form just by changing the location of the action file in Manager Actions. ExtJS was chosen (I think) because at the time, it provided sophisticated grid processing, but it's a credit to our MODX chief architect that it and the Manager are completely independent of the MODX core.

A fair point no doubt, Bruno, and Bob, but ultimately for the good of the project, and it's community, it should be acknowledged that Modx is and will be judged on what it provides in the box.

The irony of course is that the functionality and logic of the manager is one of the primary strengths of the system. It's even mentioned in one of the testimonials on the modx.com website. It's just unfortunate that the form and performance of the manager is one of the primary let-downs.

But as has already been mentioned, perhaps the core is just fine - but that doesn't change the fact that the default manager clearly needs an overhaul for the sake of both performance and compatibility.

Aside from my gripe about modx's inability to approximate a WordPress blog, one of the reasons I always second guess my decision to work with modx on any given project, including one I'm about to start, is that I worry that I may have to tell the client to switch browsers just because some of the dropdown menus and RTE popups will probably throw a wobbly from time to time and they'll need me to intervene with little workarounds to get some extremely basic stuff done.

There's a lot of hate for IE around. But its reputation is horribly disproportionate. I like a clean and slick browser environment and I have never seen a need for any addons. Addons might be one of Firefox's strengths, but as I don't need 'em I find IE10/11 to be a superior/lighter/faster experience.

To a large degree I've resisted the needless change of browser because I don't buy the reasons many people give. I don't need to use a different browser just to feel like I'm in control, I don't like to switch to software I either don't need or don't see a clear advantage to, I'm a very happy PC user who doesn't understand the apparent draw OSx has on people (my experience is that Apple's hardware is second to none but their software is consistently second rate), and I don't hate on Microsoft for no particular reason.

Ultimately, I suppose I have to admit that while IE has the lions share of the market these days (still, surprisingly, and despite Google's attempt to force their way onto everyone's machines like a virus), amongst developers and advanced users I have little doubt there are more Chrome and FF users.

Never the less, people who USE the websites I develope are NOT often advanced users, and they DO mostly use IE. So I have to convince them to change to FF or Chrome from the first time they report that something isn't doing what its told.

I have read that 2.3 will have a manager UI upgrade, though. Are we factoring that into this discussion? It doesn't seem like it. Can we get some kind of hint of what to expect? Is all this already moot?

including one I'm about to start, is that I worry that I may have to tell the client to switch browsers

Not sure why you would need to do that, the company I work for has built over 200 websites in MODX over the last few years ( both evolution and Revolution ) and we haven't had to tell anyone to change their browser )

just unfortunate that the form and performance of the manager is one of the primary let-downs.

I do agree, the manager need work and speeding up, the main issue is the document tree which refreshes on every save, so a site that has hundreds of documents like mine is a bit of a pain.

my experience is that Apple's hardware is second to none but their software is consistently second rate

Second rate software ?, really cant believe that has come from a windows user lol, OSX is one of the most secure out there, I've had a imac and mac book pro for 5 years with no problems at all, sisters WINDOWS computer is always full of viruses, but I suppose thats just down to personal taste and wanting to have virus software installed as default ( none on my machines ;o) )

Never the less, people who USE the websites I develope are NOT often advanced users, and they DO mostly use IE. So I have to convince them to change to FF or Chrome from the first time they report that something isn't doing what its told.

Can you give a example, is this manager side or public website side ? As I've already said we have no problems with IE in the manager interface.

Just to get back on topic, I think Jason and the team are doing a fantastic job developing MODX, god I remember the days of Etomite and you really dont want to go back there lol

2.3 is going to bring some new surprises for us and really change the MODX game like Revolution did over Evolution but there's still going to be people out there that don't like change, all the team can do is make a derision on what frameworks to use and what is best at the time, then build on that.

For me since I discovered MODX it has been and always will be my number 1 choice.

The comments/feelings posted here are valid, understandable and actually very valuable to the MODX core team. So thank you.

It seems there's some concern and trepidation so maybe the following can help alleviate that:

• 2.3 is coming later than expected. The reasons have been speculated on herein, and not inaccurately, but let me add that one of the primary reasons was the shift in priorities before the new year. We've been working very hard on *critical* infrastructure issues that simply couldn't stand to wait any longer. So, as you will hopefully see vast improvements in how services like Extras perform in the coming days and weeks, you will also hopefully forgive us for putting 2.3 on the back-burner temporarily to achieve that.
• 2.3 does include a LOT of Manager UI improvements. While it is still coupled with ExtJS, it has been "re-imagined": the IA has changed dramatically, the look and feel as well as performance have benefited from a LOT of attention. Dedicated community members have joined us in this effort and so it's with the utmost pride – and urgency – that we strive towards its release. It's coming soon.
• As for Github, Dave brought up a point: bugs.modx.com never really truly delivered on the promise of fully transparent progress tracking. The fact is, Redmine added yet another layer of complexity to an already bloated infrastructure, and since we already use Github, it presented a very attractive alternative to Redmine, for tracking issues. Only time will tell if this was a "good" move (I personally think it is), but one thing is certain right now: as soon as we decommission the tech stack running Redmine, we'll benefit from it. So hopefully you can get behind us on that one.

I can't tell you enough how grateful I am personally for your commitment to, and involvement with, the MODX platform. Choosing a CMS is a daunting task, and every day you put your livelihoods at stake when you make that choice. We appreciate it, and we really do spend every waking hour, of every working day, trying our best to make MODX the best decision you ever make. Thanks for your support.