If you *had* to choose another CMS for a project, what would it be?

Quote from: brianlarson at Jan 07, 2011, 04:28 PM

I’ve never heard of ProcessWire. Thanks for introducing it to me! Looks intriguing and it looks like you just need to be somewhat comfortable in PHP with the templating ( http://processwire.com/api/templates/).

Hope I’m not too far off topic, but found this thread via my GA referrers. I’m the developer of ProcessWire, and was happy to see it mentioned here. While I’m not that familiar with ModX, I’m interested in learning more. Regarding comfort level with PHP: You don’t need to be all that comfortable with PHP to use the ProcessWire API (though of course it doesn’t hurt). The API was largely patterned off of jQuery in that it’s meant to be usable without knowing PHP, in the same way that one doesn’t really need to know Javascript to get things done with jQuery. Most of the selection and traversal methods also work just like jQuery. Likewise, the API provides you a DOM-like view to your site’s pages where it’s easy to grab any page(s) based on where they are, what any custom fields contain, what type (template) it’s based on, or any combination of these. This is done with selectors (like jQuery attribute selectors), which represent a large part of the API and have nothing to do with PHP. I thought that this approach to the API was preferable to a proprietary tag-based language, at least in this case.

Have you pretty much solely switched to ProcessWire? If so, what were the biggest tipping points (whether it’s from a primarily a designer standpoint or a developer standpoint or both)? I’d love to see a quick comparison if you’re inclined and have a few minutes.

I’m also interested in hearing GoustiFruit’s thoughts on this too. I appreciate the ProcessWire mention. I don’t have experience with modx other than trying out the Revolution demo recently, and am glad to see there is some common ground. ProcessWire is new, and in development, but getting closer to a full release (though you can DL the developer preview from GitHub). It’s not meant to replace any other product so much as it is to be a complimentary tool to add to your arsenal.

Ryan

I don’t have that much experience with modX, Processwire, or other CMS/CMF or even PHP; I don’t even have a real site done with any of these yet But I’m planning to, and I’m learning bit after bit. And so I settled on modX a few months ago, as a learning engine, after reading all the good reviews and browsing these forums for weeks. So I started developing my first website using modX; first with Evo then with Revo as I discovered modX at about the time Revo was released. But anytime I wanted to do *anything* in modX, I had to learn a new markup, how to use a new snippet (and first had to find the correct documentation; most of the time I had better luck by searching for specific tutorials on the net; ex. WayFinder, Ditto - then GetResources, ...). Same thing when I tried to connect to the DB to access my own tables, I started with Evo and it was quite easy with plain queries, then in Revo I couldn’t find a way until I was told it was better to use XPDO/PDO, which I tried to learn for a couple weeks, before someone told me it was also possible but not advised to do it the old way. Also I was very frustrated by the slow interface in modX manager, and by the fact that - for example - I had to create sub-resources to allow the inclusion of multiple (ie. dynamic number) images or urls. And there is that "problem" with the templates having to be stored in the DB, requiring incessant copying/pasting from my favorite editor to modX UI (yes after some time I was using the hack to load them from file, but that kept me thinking there was something wrong in its conception). Also one thing I kept reading when discovering modX is that it was one of the only, if not the only, framework that allowed managing multiple sites from one single installation; I gave it a try and was submerged by the complexity of the process, having to set up a lot of configuration options, having to play with "contexts" (still have to understand this fully), etc. It’s certainly something very powerful, but it’s also something that made me realize I didn’t need all of this complexity, that I needed something much easier.
So I was still learning modX but started to search for alternatives; tried a few and stumbled upon Processwire. After reading the presentation and watching the little video introduction, I thought it could be the one. Gave it a try, and could recreate the website I was working on with modX for weeks in just a few lazy days. And all immediately made sense. It was like I just found the missing piece of the puzzle But I must say that modX prepared the path for me because Processwire allows me to work about the same way, I managed to use some concepts into Processwire because I already learned them from modX.
So just to list a few positive points in Processwire against modX:
- no need to learn new markups: [(, [[, [+, [*, [!, ... Always brings confusion when I’m coding in templates, snippets, chunks, ... I end up trying each combination until it finally works !
- multiple "template variables": easy selection of multiple images for one single page for example.
- I can code "snippets" equivalent with a few lines of code, removing all the bloat. For example there is not need for something like GetResources, I can easily find any page and access its variables thanks to the simple API.
- I can make "chunks" equivalent too. I put "snippets" and "chunks" in their own folders, so can easily edit them in my editor of choice.
- fast responsive manager: I can browse hundreds/thousands of pages with no slowdown; in modX, before the resources tree reloads, you can forget what you wanted to do !
And... that it’s for now. Still learning.

GoustiFruit, thanks for your thoughts on this. I’m glad to hear you like ProcessWire, and from the things you mentioned, it sounds like you really understand it well too. It’s a new project and still evolving, and I appreciate feedback and suggestions, and anyone that wants to participate in the project. Stop by the ProcessWire forum sometime to talk more (processwire.com/talk/). And let me know if I can be of any help as you explore ProcessWire. The documentation is coming along well, but there’s still lots more to cover, so I encourage everyone using it to email me and and/or post in the forum with questions as they come up.

I don’t know enough about ModX to understand the context in some of the items you mentioned, but I look forward to learning more about it. I think that every product has it’s highs and lows ... and the definition of each depends on the needs on an individual project, the designer’s/developer’s needs and development style. What are lows to one person (or project) are highs to another. Like you, I’m not a fan of tag-scripting and proprietary markup tags in a CMS API, but I’ve talked with enough people (especially expression engine users) to make me think that there are good reasons for it too. Namely, providing greater accessibility to designers with a syntax that they "perceive" is simpler, because PHP is perceived as complex by some ... even in contexts where I’d argue it’s not. Likewise, I’m also not wild about templates or snippet-type things that you have to edit in a CMS, because I’m very attached to my development process and editor (VIM!). Using the CMS to edit these things seems like the wrong place and slows me down. But that’s just an opinion... I’ve talked to others that think this type of editing in the CMS is key because the flexibility in modifying deployed sites (should they be doing this??) is more valuable than the editing platform to them, among other reasons.

To reply to the subject of this thread: If I couldn’t use the 2 CMSs referenced in this message, then I would use Drupal 7 or Expression Engine 2.

Ryan, your work looks fantastic. Kudos! It looks like learning templating for a designer like myself that knows *enough* PHP wouldn’t be super difficult. I look forward to trying out ProcessWire. Maybe that’s the third one I need to add to my arsenal (after I add the dang 2nd which is WordPress... mainly due to demand and popularity).

I think you’re definitely on to something with designers being drawn to MODx. In all honesty learning the MODx tags and basic snippets was quite easy for me and the community is SUPER helpful and friendly. I do think it’s more than "perception" though. For instance, MODx’s navigation snippet WayFinder (http://wiki.modxcms.com/index.php/Wayfinder) is very powerful and probably much easier to use for the non-dev types. Another thing to note: Include directives can be used (though I haven’t tried it) so you can do dev from your favorite IDE and upload via FTP instead of being forced to use MODx Manager while developing. I totally realize that you’re new to MODx (but NOT to CMSs obviously!...ProcessWire looks awesome) so I’m not writing this to "defend" it by any means. I just wanted to speak to the design integration issue and proprietary CMS tags/language for development.

P.S. Are all the source files available for the construction of the ProcessWire demo site? It’s a REALLY nice demo site and I’m deconstructing it both in the admin area as well as PHP/HTML files would probably be super helpful in learning about it.

P.P.S. Correct me if I’m wrong on this but... do so few awesome CMSs lack a publishing/drafting/versioning workflow of sorts because it just happens to be SUPER complicated to develop? That was one thing that really turned me on when checking out SilverStripe.

P.P.P.S. Since this is about MODx and other CMSs I wanted to throw out a steamy question... why, in your opinion, does Joomla! seem to have the worst reputation in the open source CMS community? Is it just a tool that’s plug-and-play for complete non-nerds?

Quote from: brianlarson at Jan 10, 2011, 02:58 AM

P.P.P.S. Since this is about MODx and other CMSs I wanted to throw out a steamy question... why, in your opinion, does Joomla! seem to have the worst reputation in the open source CMS community? Is it just a tool that’s plug-and-play for complete non-nerds?

Only 1 valid answer: test that out (create custom template, custom add on, and make an upgrade when it’s available).

I have a freshly sharpened pencil over there. Something tells me that I may be better shoving that gently unto my eyeball. I’ve tried to use the admin for an old site that was previously done in Joomla! and that was enough to disgust me just from the point of view of the client (that site has since been redone in Evo at alumacraft.com). Shameless plug.

Sorry to get off topic, folks. I’m confident Joomla! doesn’t belong in this discussion...?

Quote from: brianlarson at Jan 10, 2011, 02:58 AM

Ryan, your work looks fantastic. Kudos! It looks like learning templating for a designer like myself that knows *enough* PHP wouldn’t be super difficult. I look forward to trying out ProcessWire. Maybe that’s the third one I need to add to my arsenal (after I add the dang 2nd which is WordPress... mainly due to demand and popularity).

Brian, thanks for checking it out. I think you’ll like the templating system, but let me know. I’m also a designer by trade and like to think the template system would appeal to other designers, even if it isn’t tag-based. Though I should mention that previous versions [Dictator CMS, 2003–2007] were fully tag based, and ProcessWire 1 [2007–2009]) was combination "tagscript" and PHP API based. I left tagscript out of ProcessWire 2 because I no longer used it myself. I know there are enough people that still like tag-based syntax that I may build it as an plugin module for those that prefer it at some point. Here’s a little more explanation: http://processwire.com/api/why-php-syntax/

In all honesty learning the MODx tags and basic snippets was quite easy for me and the community is SUPER helpful and friendly. I do think it’s more than "perception" though. For instance, MODx’s navigation snippet WayFinder is very powerful and probably much easier to use for the non-dev types.

I think that makes sense, and Wayfinder looks interesting. If it’s handy, could you post a simple Wayfinder tag example and tell me what output it would generate? Btw, I am also very impressed by the MODx community and team.

Regardless of the scope and complexity of the output, I’m guessing that MODx would take a tag like {some_tag name="value"} and convert it to PHP like some_tag("name=value"); At least, this would be the case in ProcessWire (it may be different in ModX). But the syntax differences ultimately come down to whether you prefer brackets or parenthesis and where you like to place the characters. The advantage of the PHP syntax is that it’s non proprietary and doesn’t involve the overhead of the tag parsing engine. This is an oversimplification to some extent, but it’s one reason why I decided to stick to PHP-based syntax, at least for now.

P.S. Are all the source files available for the construction of the ProcessWire demo site? It’s a REALLY nice demo site and I’m deconstructing it both in the admin area as well as PHP/HTML files would probably be super helpful in learning about it.

This is available. Drop me a note at ryancramer.com/contact/ and I’ll email you the link (there are some notes that go along with it that accompany the email).

Correct me if I’m wrong on this but... do so few awesome CMSs lack a publishing/drafting/versioning workflow of sorts because it just happens to be SUPER complicated to develop? That was one thing that really turned me on when checking out SilverStripe.

I would say it’s not particularly complicated to develop, but it does increase the client’s perceived complexity of the CMS quite a bit. ProcessWire 1 had page drafts, versioning, comparing and workflow functions. A few clients thrived on them, but most never used them. Confusion over the terminology and capability drove more than half of my tech support for the product. Maybe my execution was off, but it was certainly simpler than most others I’ve used.

ProcessWire 2 had drafts in the core, and versioning was in progress. But while training a client using ProcessWire 1, I decided to remove it shortly before releasing PW2. I want to keep the core as simple as possible, and keep anything that might be optional in modules. So the plan is to bring back all of these functions in an add-on module, so that it won’t be part of the core and confuse the majority of people who don’t want it.

Btw, I was impressed with this feature of SilverStripe as well. They have taken the capability further than I’d seen. On the other hand, I couldn’t see using SilverStripe in my own projects as I must be too set in my ways about other aspects of it.

Since this is about MODx and other CMSs I wanted to throw out a steamy question... why, in your opinion, does Joomla! seem to have the worst reputation in the open source CMS community? Is it just a tool that’s plug-and-play for complete non-nerds?

I don’t know much about Joomla. It’s not for lack of trying… I have tried various versions, including back when Joomla was Mambo (before they split). Most recently I tried it in November. Every time, I can’t seem to make it more then 15 mins before deciding that it’s just not for me. Nothing against Joomla, as it may be a very good product. But the strategy/philosophy in how it works just seems to be very different from my preferences.

Well I’m a little late to the party but here’s my $0.02 worth...

I’d echo a lot of the comments on Drupal, WordPress (WP) etc. WP is really moving forward now as a more general system and I’d recommend taking a look at Pods CMS for WP to give you a lot more CMS functionality including great relationships, custom data structures, custom admin screens etc. WP 3.1 is due out soon - and with it Pods 2.0 which will up the ante quite a bit on that front. In this respect I think it goes a bit more towards the DIY/framework side of things than the ’rent an apartment’ crowd...

I’ve done a lot of work with Drupal and never been a fan due to the complexity of the admin UI and the work you need to put in for it to be really client friendly, plus the gruesome code generated by default at the front end... Drupal 7 will hopefully fix a lot in both those areas... but the templating still way behind SS/MODx IMO and codebase looks tired by comparison. Not to mention with WP for example there’s a commitment by the core team to backwardly support as many add-ons as possible but with Drupal they seem to change whatever they think needs doing and then let all the add-on folk do the work of accommodating those changes.

I think probably the closest system to MODx is SilverStripe (SS) in terms of being a CMF and CMS combo - but doesn’t really use a tag system - only PHP - you’re essentially writing OO PHP for pretty much all customisation; so better suited to pure PHP devs than CSS/JS folk I think. Revo is great now the tag syntax is more ’unified’ and you can drag-drop with prompts for parameters - so much better than Evo in my opinion. The ability to work with custom and logical database structures with an easy way to establish relationships between data objects is a real strong point for SS (similar to xPDO work that’s come into Revo). Community is also great - but smaller than MODx. The code is very portable so extremely easy to re-use from one project to another and moving entire sites is possibly the easiest in any CMS I know of (which is a strong selling point in my book). I know the versioning and workflow stuff is in the pipeline for MODx but at this stage I think SS has the edge in this respect and generally clients have appreciated the ability to preview new versions of a SS site before updating live material - plus rollback if things go wrong.

I’ve used ExpressionEngine quite a bit but there you’re generally working exclusively with the tag system and if you want to use PHP and SQL then you could find the behind-the-scenes data structures a bit confusing for want of a better word... Plus of course if you just want to try out lots of extensions etc. then many are commercial so you can’t just download and start playing as you can with the FOSS systems like MODx/SS/WP/Drupal etc. Plus I think you’ll find most sites you build require not just a core ExpressionEngine licence but also at least a few commercial add-ons so you’d have to weigh that against the results you could achieve by contributing the same ammount of cash each time to a FOSS project (either core or add-on) or using the cash to hire someone who could help fill in any gaps you have in your own skillset...

A couple of the better articles I’ve come across: http://www.webbower.com/blog/the-great-cms-roundup/ and http://sonspring.com/journal/why-drupal. Well hope that helps...

I’m not a fan of tag-scripting and proprietary markup tags in a CMS API

It took me a while to get used to this, but as I learn more about MODx I’m forming the opinion that it’s not just a matter of perception of ease by non-coders. I think doing things this way has a design logic as well, as it creates a rigid wall between markup and code. This helps to make sure everything stays completely modular. Any MODx snippet or plugin can be easily packaged and distributed and used in many installations, by coders and non-coders alike. Obviously you can do the same thing just using functions and classes and so forth, and just enforce modularity with best-practices, but in MODx you have no choice but to completely separate the code from the markup and package it in a way that makes it reusable. Sometimes not giving people the choice to do things wrong makes for a better design.

Also, there’s no reason to give up your editor or edit code in the CMS, you can just copy/paste or you can keep your code in separate files and distribute them via the package manager (in Revo). With code of any significant complexity, you definitely want to test it in your dev environment before deploying it live.

Having said that, your project definitely sounds interesting and worth checking out -- just wanted to throw in my $.02 in defense of MODx, which is by far the best system I’ve used so far.

Quote from: esnyder at Jan 18, 2011, 11:59 PM

Also, there’s no reason to give up your editor or edit code in the CMS, you can just copy/paste or you can keep your code in separate files and distribute them via the package manager (in Revo). With code of any significant complexity, you definitely want to test it in your dev environment before deploying it live.

I just want to step in for a second here and say that the core dev team is very aware of the desire/need to edit Snippets/Plugins/etc outside the MODX manager. We’ll probably be adding either "Static Elements" (or some similar solution) to solve the problem while still allowing the current implementation of editing in the manager, and that this is one of the top 3 features we’re prioritizing development for in the next few releases.