A good discussion ensures after Mike asks a simple question: for news organizations, which is better, a blog CMS or an enterprise CMS? My answer? They both suck.
I’m not sure I agree that this discussion has anything to do with software, really. Saying things like WP or EE aren’t extensible isn’t really true — they’re both easily changed through custom programming. Sure, you may have to follow pre-established conventions to avoid repeating work, but you don’t have to. And you’d have to make the same choices with Ellington (i.e. to “roll your own” solution or “go with the flow”). If a feature needs to be implemented, someone will have to deisgn, develop and deploy it — regardless of the technology or platform.
Really, I think the true discussion is about when will newspapers and media outlets hire uber-talented web programmers as an in-house teams, and give them a very pivotal role in the news workflow (like your situation at the Kansas paper). Whatever tech route you choose, you still need a talented team to make it all make sense to the readers.
And you’d have to make the same choices with Ellington (i.e. to “roll your own” solution or “go with the flow”). If a feature needs to be implemented, someone will have to deisgn, develop and deploy it — regardless of the technology or platform.
Yeah. I think the difference, though, is that something like Ellington is designed form the ground up as a platform to be extended. You can add to it, with zero limitations, without ever touching Ellington’s source code. With Wordpress, you have a plug-in architecture, which allows for limited extensibility, but if you want to go beyond that, you’re going to have to modify Wordpress directly, which is never pretty (and it makes maintenance a nightmare, because you have to roll your changes into any upgrades).
I’m careful not to talk about EE much, because I really have never used it and don’t know enough about it to make an assessment. But, my impression is that it, too, has limits to what you can do without hacking the EE code itself.
The main thing I want, especially for a serious news site, is a platform I can build upon, rather than just a product I can use. I don’t think Wordpress qualifies. EE might, I’m not sure. Custom solutions definitely do.
I still don’t quite see this distinction — WP doesn’t require that you use a limited, crippled “plugin API” to develop for the platform. Really, you’re just writing functions and routines that extend the core system, using the established programming paradigms of the platform — just as you would in EE, Django, Rails, etc. You don’t have to hack at the core files, and you shouldn’t (much like you shouldn’t modify the core framework files in Django, so you can upgrade them easily).
I’ll agree 100% that WP isn’t the first choice I’d run to when developing for a production level news site that had a multitude of content contexts. Still, for many “news like” applications (e.g. magazines, small papers, etc) it will be at the top my short list.
Any chance you’ve seen the AllThingsD site? It’s an extensive “magazine-y” type news site that’s powered by WP and WPMU. I think it does a great job of showing how well WP can handle a wide variety of content and not appear very ‘bloggy’ or constrained by the publishing technology.
In my opinion, WP is not a ramshackle, hacky, blog-only system (as many in the industry see it). Rather, it’s a mature, solid, and useful tool for many applications, and it can be extended in a reliable, sensible way. It’s not perfect, but then I really don’t think any software platform will ever take that honor :)
Well I’ve used several of these systems, and I can speak for a portion of them.
First off, throwing in Rails & DJango versus EE& Wordpress is not an apt comparison. Rails and Django are not publishing platforms, just like .Net is not a publishing platform. Their purpose is to facilitate easy development and really shouldn’t be included in this conversation about Wordpress.
This isn’t to say you couldn’t build a Wordpress-esque application with Rails & DJango, but also you could build a system that doesn’t even touch content with Rails & Django: It could merely run server processes, or take registration keys for software. With Wordpress and EE it’s more like a Home Theater in A Box System, but Rails and Django is like the motherboard that could possibly be used to make a great receiver, or breakfast.
Wordpress itself is built specifically with ease of setup in mind. This is why it’s a great system. It can be modified extensively to fit a purpose, but it’s not what’s at the core of the system. Some very smart people have done some smart things with it, but it is still a blogging platform at its core.
Expression Engine, I have found, is much more versatile, not only in handling modifications but in giving designers freedom. The system was built around whatever you design, not the same with Wordpress where you have to build around the system for your design. That is the fundamental difference between the two, and because of that Expression Engine has a higher learning curve upfront. It is not as easy to use right upfront as Wordpress is.
But Wordpress is not ramshackle, nor hacky, I will agree on that. It is a great tool, but it is also the one you have to modify the most to follow your will. It may no longer be a platform solely for blogging, but that was its original intent, and even though it has evolved, it hasn’t quite stripped away that baggage.
Well, first, allow me to admit that I haven’t used Wordpress in years, and it’s quite possible (even probable) it’s advanced beyond the tool I remember. But, as it pertains to a news organization’s CMS, I view its limitations primarily in its ability to handle structured data for unique content types. I think this is something that is hard for people who have never worked in journalism to understand. The All Things D site, while a great site, is really nothing like a daily newspaper’s site. It’s the unique requirements that news organizations have that are where I think Wordpress falls down.
Consider this example: when I worked in Lawrence, one of our most popular sections was obituaries. Ideally, users would like to be able to search obits in many ways. For example, it would be nice to be able to search for obits in the past week of people who live in a particular town. Once you’ve found an interesting obit, it would be nice to have it tied to other news data. For example, if I find an obit of a person, I would also like to be able to easily find the birth announcement of his daughter. Or the engagement announcement for him and his wife.
The only way to accomplish this sort of thing is to have obits as a unique object type in the database, with unique properties and unique relationships to other object types. I don’t think Wordpress would allow this (correct me if I’m wrong). I think, using Wordpress, you would pretty much treat obits the same way you would any other blog — each obit would have a title and a blog of text, and not much more. That works, sure — but it’s definitely not the sort of think cutting edge news organizations want to do.
There may well be ways to handle this stuff in Wordpress (or similar), but I don’t know what it is, and I have a hard time believing you could do it without serious hacking. I could definitely be wrong, though.
@Jeff and Kenny:
Okay, time for some concessions — WP is not the default tool for a site that requires a variety of unique data types, each requiring relational ties to other unique data sets. In that case (the news site), it is of course best to have something infinitely flexible, and built with the intent of custom programming. Agreed in full.
However, my whole point in even bringing up this argument is just to point out that WP is a very capable, flexible, and extendable system. It can easily handle a variety of “news like sites” and present, organize, and manage a good variety of content types. I say this because I’ve handled CMS development using WP on two very custom sites, and the system works incredibly well. It lets someone like me, a design technologist, create a reasonably robust management system on tight budgets and timelines. Would a custom, hand-built system have worked better — possibly. But I can guarantee that it would take a programmer more time to design, develop, test, deploy and refine such a solution.
Wordpress get’s an incredibly bad rap — one that, in my mind is not deserved. Anyway, thanks for the nice discussion… I think I’ll blog about this soon :)
EE allows you to build up arbitrary data structures from a set of primitive data types and then using a macro looking template language, list, view, search, paginate, add, edit, update and finally delete entries.
EE routines and modules are geared to blogging of course, but it is very flexible at doing other things. How EE manages to do this inside the database, while clever in its own way, its not pretty.
Once you exceed about three data structures that need relationships between them to work, then you can expect your average blog cms to start to crack under the pressure, esp if it needs to deal with large amounts of data. Gluing and enforcing relationships in template logic, good luck with that..
@Chris: Even though the data structures aren’t pretty, you can put query code into EE and do your own queries. You can also throw PHP in there to handle more complex interactions. It’s only for the hardcore, sure, but its built into the EE system. Plus, its excellent, extensive caching system takes away a lot of the heavy costs of those complex queries it runs.
From my understanding as well EE 2.0 is supposed to handle those relationships better… and it comes with jQuery.
@Matt: Wordpress does get a bad rap. It powers my site now. I really think that Wordpress gets a bad rap because its the popular kid. Which of course, I can relate to ;)
001 // Matt Brown // 06.26.2008 // 5:51 PM
I’m not sure I agree that this discussion has anything to do with software, really. Saying things like WP or EE aren’t extensible isn’t really true — they’re both easily changed through custom programming. Sure, you may have to follow pre-established conventions to avoid repeating work, but you don’t have to. And you’d have to make the same choices with Ellington (i.e. to “roll your own” solution or “go with the flow”). If a feature needs to be implemented, someone will have to deisgn, develop and deploy it — regardless of the technology or platform.
Really, I think the true discussion is about when will newspapers and media outlets hire uber-talented web programmers as an in-house teams, and give them a very pivotal role in the news workflow (like your situation at the Kansas paper). Whatever tech route you choose, you still need a talented team to make it all make sense to the readers.
002 // Jeff Croft // 06.26.2008 // 6:18 PM
Yeah. I think the difference, though, is that something like Ellington is designed form the ground up as a platform to be extended. You can add to it, with zero limitations, without ever touching Ellington’s source code. With Wordpress, you have a plug-in architecture, which allows for limited extensibility, but if you want to go beyond that, you’re going to have to modify Wordpress directly, which is never pretty (and it makes maintenance a nightmare, because you have to roll your changes into any upgrades).
I’m careful not to talk about EE much, because I really have never used it and don’t know enough about it to make an assessment. But, my impression is that it, too, has limits to what you can do without hacking the EE code itself.
The main thing I want, especially for a serious news site, is a platform I can build upon, rather than just a product I can use. I don’t think Wordpress qualifies. EE might, I’m not sure. Custom solutions definitely do.
003 // Matt Brown // 06.29.2008 // 10:50 PM
I still don’t quite see this distinction — WP doesn’t require that you use a limited, crippled “plugin API” to develop for the platform. Really, you’re just writing functions and routines that extend the core system, using the established programming paradigms of the platform — just as you would in EE, Django, Rails, etc. You don’t have to hack at the core files, and you shouldn’t (much like you shouldn’t modify the core framework files in Django, so you can upgrade them easily).
I’ll agree 100% that WP isn’t the first choice I’d run to when developing for a production level news site that had a multitude of content contexts. Still, for many “news like” applications (e.g. magazines, small papers, etc) it will be at the top my short list.
Any chance you’ve seen the AllThingsD site? It’s an extensive “magazine-y” type news site that’s powered by WP and WPMU. I think it does a great job of showing how well WP can handle a wide variety of content and not appear very ‘bloggy’ or constrained by the publishing technology.
In my opinion, WP is not a ramshackle, hacky, blog-only system (as many in the industry see it). Rather, it’s a mature, solid, and useful tool for many applications, and it can be extended in a reliable, sensible way. It’s not perfect, but then I really don’t think any software platform will ever take that honor :)
004 // Kenny Meyers // 06.30.2008 // 7:42 AM
Well I’ve used several of these systems, and I can speak for a portion of them.
First off, throwing in Rails & DJango versus EE & Wordpress is not an apt comparison. Rails and Django are not publishing platforms, just like .Net is not a publishing platform. Their purpose is to facilitate easy development and really shouldn’t be included in this conversation about Wordpress.
This isn’t to say you couldn’t build a Wordpress-esque application with Rails & DJango, but also you could build a system that doesn’t even touch content with Rails & Django: It could merely run server processes, or take registration keys for software. With Wordpress and EE it’s more like a Home Theater in A Box System, but Rails and Django is like the motherboard that could possibly be used to make a great receiver, or breakfast.
Wordpress itself is built specifically with ease of setup in mind. This is why it’s a great system. It can be modified extensively to fit a purpose, but it’s not what’s at the core of the system. Some very smart people have done some smart things with it, but it is still a blogging platform at its core.
Expression Engine, I have found, is much more versatile, not only in handling modifications but in giving designers freedom. The system was built around whatever you design, not the same with Wordpress where you have to build around the system for your design. That is the fundamental difference between the two, and because of that Expression Engine has a higher learning curve upfront. It is not as easy to use right upfront as Wordpress is.
But Wordpress is not ramshackle, nor hacky, I will agree on that. It is a great tool, but it is also the one you have to modify the most to follow your will. It may no longer be a platform solely for blogging, but that was its original intent, and even though it has evolved, it hasn’t quite stripped away that baggage.
005 // Jeff Croft // 06.30.2008 // 8:28 AM
Well, first, allow me to admit that I haven’t used Wordpress in years, and it’s quite possible (even probable) it’s advanced beyond the tool I remember. But, as it pertains to a news organization’s CMS, I view its limitations primarily in its ability to handle structured data for unique content types. I think this is something that is hard for people who have never worked in journalism to understand. The All Things D site, while a great site, is really nothing like a daily newspaper’s site. It’s the unique requirements that news organizations have that are where I think Wordpress falls down.
Consider this example: when I worked in Lawrence, one of our most popular sections was obituaries. Ideally, users would like to be able to search obits in many ways. For example, it would be nice to be able to search for obits in the past week of people who live in a particular town. Once you’ve found an interesting obit, it would be nice to have it tied to other news data. For example, if I find an obit of a person, I would also like to be able to easily find the birth announcement of his daughter. Or the engagement announcement for him and his wife.
The only way to accomplish this sort of thing is to have obits as a unique object type in the database, with unique properties and unique relationships to other object types. I don’t think Wordpress would allow this (correct me if I’m wrong). I think, using Wordpress, you would pretty much treat obits the same way you would any other blog — each obit would have a title and a blog of text, and not much more. That works, sure — but it’s definitely not the sort of think cutting edge news organizations want to do.
Some more examples:
How do you store election results in Wordpress?
Or movie listings?
Or weather data?
There may well be ways to handle this stuff in Wordpress (or similar), but I don’t know what it is, and I have a hard time believing you could do it without serious hacking. I could definitely be wrong, though.
006 // Matt Brown // 06.30.2008 // 10:33 PM
@Jeff and Kenny: Okay, time for some concessions — WP is not the default tool for a site that requires a variety of unique data types, each requiring relational ties to other unique data sets. In that case (the news site), it is of course best to have something infinitely flexible, and built with the intent of custom programming. Agreed in full.
However, my whole point in even bringing up this argument is just to point out that WP is a very capable, flexible, and extendable system. It can easily handle a variety of “news like sites” and present, organize, and manage a good variety of content types. I say this because I’ve handled CMS development using WP on two very custom sites, and the system works incredibly well. It lets someone like me, a design technologist, create a reasonably robust management system on tight budgets and timelines. Would a custom, hand-built system have worked better — possibly. But I can guarantee that it would take a programmer more time to design, develop, test, deploy and refine such a solution.
Wordpress get’s an incredibly bad rap — one that, in my mind is not deserved. Anyway, thanks for the nice discussion… I think I’ll blog about this soon :)
007 // Chris Blown // 07.01.2008 // 5:10 AM
EE allows you to build up arbitrary data structures from a set of primitive data types and then using a macro looking template language, list, view, search, paginate, add, edit, update and finally delete entries.
EE routines and modules are geared to blogging of course, but it is very flexible at doing other things. How EE manages to do this inside the database, while clever in its own way, its not pretty.
Once you exceed about three data structures that need relationships between them to work, then you can expect your average blog cms to start to crack under the pressure, esp if it needs to deal with large amounts of data. Gluing and enforcing relationships in template logic, good luck with that..
008 // Kenny // 07.01.2008 // 7:32 AM
@Chris: Even though the data structures aren’t pretty, you can put query code into EE and do your own queries. You can also throw PHP in there to handle more complex interactions. It’s only for the hardcore, sure, but its built into the EE system. Plus, its excellent, extensive caching system takes away a lot of the heavy costs of those complex queries it runs.
From my understanding as well EE 2.0 is supposed to handle those relationships better… and it comes with jQuery.
@Matt: Wordpress does get a bad rap. It powers my site now. I really think that Wordpress gets a bad rap because its the popular kid. Which of course, I can relate to ;)