Is eZ Publish 4.0 like Microsoft Vista?
Disclaimers usually go at the bottom of a post but because this one is quite lengthy I've decided to put it at the top. Mostly the stuff I post here is of a technical nature, but this time I'm posting an option piece, and as such it is full of wild speculation, personal thoughts, suggestions and of course opinion, all of which may have no basis in any reality other than my own.You are here - the current state of play
I've been using eZPublish 4.0 since the beginning of the year. It's not quite what I envisioned when the rumbling of version 4.0 first started, oh so long ago, actually it's nothing like what I imagined.
Hence the title of this post. Initially Vista (Longhorn) was slated as a "minor step" between Windows XP (Whistler) and Blackcomb. However it gradually took on many of the features slated for Blackcomb. People got excited, but it's release was delayed several times, the developers were given other priorities and finally development started afresh based on Windows 2003 server and a limited feature set.[1] When finally released the general consensus seems to be one of disappointment or at least significantly underwhelmed.
Over time there has been quite a bit of excitement of what a new version of our favourite CMS might provide. Given the quantum leap in functionality between eZ Publish 2.0 and 3.0 who knew what unimaginable ecstasy version 4.0 would bring! However time marched on, PHP4 end of life was announced and eZ Publish 4.0 became a port of eZ 3 to PHP5 (continuing the community effort by Paul and Kristof).
Now this is not a bad thing, on the contrary it's the best thing that happened to eZ Publish in recent times (including the Online Editor in the distribution would be a close second).
At the local PHP meetup I would get strange looks because I was still using PHP4. I could get people excited about eZ Publish but people soon lost interest when told that it required PHP4 and a specific version at that. No matter how good the product sounded it's was hard to get people interested about something that had "special requirements". "None of the other PHP CMS have those requirements" was the standard response. It was impossible to get a foot in the door. Hopefully now with PHP5 support more people will be willing to try eZ Publish.
PHP5 support also makes it easier to get eZ Publish into corporate environments. RedHat seems to be the Linux distribution of choice in the corporate world (at least here in Australia) and it's lack of support for php4.4 has meant that in situations where we were the "application provider" we either had to convince the systems people to install the newer unsupported (by RedHat) version of PHP or do it ourselves. Of course this just created a bunch of issues around support (of the custom PHP) and called into question the choice of eZ Publish as a CMS platform.
Oh yeah, eZ runs faster under PHP5 as well!
Three cheers for eZ4.0 on PHP5!
There are a bunch of subsystems in the current version of eZ Publish that are tightly coupled with the main content module that makes them pretty difficult to reuse in custom modules. Two that come to mind are the Image system and the Rich Text/XML Block Datatype. It seems such a waste that these subsystems can't easily be reused outside of the content module.
From what I've seen of the eZ Components they will solve these type of issues. They will form the building blocks on which future versions of eZ publish are built. From what I can tell current version of eZ doesn't utilise components at all, although some of the components features have made their way into the eZ publish libraries. (Of course running under PHP5 allows the use of components in extensions).
[1] Paraphrased from: http://en.wikipedia.org/wiki/Development_of_Windows_Vista
You are here - the current state of play
I've been using eZPublish 4.0 since the beginning of the year. It's not quite what I envisioned when the rumbling of version 4.0 first started, oh so long ago, actually it's nothing like what I imagined.Hence the title of this post. Initially Vista (Longhorn) was slated as a "minor step" between Windows XP (Whistler) and Blackcomb. However it gradually took on many of the features slated for Blackcomb. People got excited, but it's release was delayed several times, the developers were given other priorities and finally development started afresh based on Windows 2003 server and a limited feature set.[1] When finally released the general consensus seems to be one of disappointment or at least significantly underwhelmed.
Over time there has been quite a bit of excitement of what a new version of our favourite CMS might provide. Given the quantum leap in functionality between eZ Publish 2.0 and 3.0 who knew what unimaginable ecstasy version 4.0 would bring! However time marched on, PHP4 end of life was announced and eZ Publish 4.0 became a port of eZ 3 to PHP5 (continuing the community effort by Paul and Kristof).
Now this is not a bad thing, on the contrary it's the best thing that happened to eZ Publish in recent times (including the Online Editor in the distribution would be a close second).
At the local PHP meetup I would get strange looks because I was still using PHP4. I could get people excited about eZ Publish but people soon lost interest when told that it required PHP4 and a specific version at that. No matter how good the product sounded it's was hard to get people interested about something that had "special requirements". "None of the other PHP CMS have those requirements" was the standard response. It was impossible to get a foot in the door. Hopefully now with PHP5 support more people will be willing to try eZ Publish.
PHP5 support also makes it easier to get eZ Publish into corporate environments. RedHat seems to be the Linux distribution of choice in the corporate world (at least here in Australia) and it's lack of support for php4.4 has meant that in situations where we were the "application provider" we either had to convince the systems people to install the newer unsupported (by RedHat) version of PHP or do it ourselves. Of course this just created a bunch of issues around support (of the custom PHP) and called into question the choice of eZ Publish as a CMS platform.
Oh yeah, eZ runs faster under PHP5 as well!
Three cheers for eZ4.0 on PHP5!
eZ Components to the rescue
I spend much of my time working with eZ Publish as a framework by coding extensions that merge existing data sources into a CMS managed environment as well as creating fully fledged applications that run parallel to the CMS. When I learned of the eZ Components project and that it would from the basis of the new eZ Publish version I was very happy.There are a bunch of subsystems in the current version of eZ Publish that are tightly coupled with the main content module that makes them pretty difficult to reuse in custom modules. Two that come to mind are the Image system and the Rich Text/XML Block Datatype. It seems such a waste that these subsystems can't easily be reused outside of the content module.
From what I've seen of the eZ Components they will solve these type of issues. They will form the building blocks on which future versions of eZ publish are built. From what I can tell current version of eZ doesn't utilise components at all, although some of the components features have made their way into the eZ publish libraries. (Of course running under PHP5 allows the use of components in extensions).
New versions
Yesterday I was alerted to an update to the eZ Publish roadmap by a posting from[1] Paraphrased from: http://en.wikipedia.org/wiki/Development_of_Windows_Vista
Hi Bruce,
ReplyDeleteallthough I understand your wish for a total rewrite to avoid the hideous mixed-up monster, I really believe that the step-by-step strategy has a brighter future seen from a commercial perspective.
Off course, in this way it resembles Vista which as all M$ $tuff comes to life to make money.
But there is nothing wrong in making money, especially when done the way those eZ Folks do it.
The step-by-step strategy adds to the already existing image of eZ Systems as being a proffesional credible provider.
You're right Bruce! Summer is allways winter for someone else and should be avoided at any cost :-)
Felix
Hi Felix
ReplyDeleteThanks for the comment! You bring up a point that I didn't touch on in the post, money and the commercial aspect of eZ. I certainly don't begrudge eZ making money, and as a user of their software and someone who makes a living working with it I hope they make as much as possible so they can continue doing what they do!
I'm not sure I agree with your point about the step-by-step approach enhancing the image of eZ Systems. In the end though, the proof will be in the result. As I pointed out in the post I don't see the merging as a trivial task and suspect it will have implications. I suspect that eZ will be able to pull it off, the question I'm putting forward is whether it's the best long term solution.
With my developer hat on I know that it's a lot more fun starting from scratch armed with a great new library and the lessons of the last project.
BTW - I after looking more closely at the template component I suspect that existing templates will be OK. I don't see the old {section} tag there so this will mean that the older legacy templates will require a clean up.
Cheers
Bruce
Hi Bruce,
ReplyDeleteyou're right off course. There most certainly will be trouble in paradise as the new pieces are put into the puzzle.
When it comes to the new template engine, I'm really looking forward to it. It will boost speed, and that being added to the php5-effect (especially the XML-parser) will be a doble-boost for ezp-performance.
I image there's been quite a few arguments over at the eZ Bunker (aka eZ HQ) concerning the 4.0-approach. One can only hope that they made the right decission and, even more important, that they wil do their very best to make the transission as smooth as possible.
Felix
Hello Bruce,
ReplyDeleteVery interesting perspective. I have to say that I like this idea about feature freeze mode + eZ Components-based rebuild.
Since I do care, I posted a more lengthy reply in my own blog. Let's spread it, maybe it leads to some interesting discussion.
Greetings,
Piotrek