Comments for Ramblings of a Code Monger http://cameronc.wordpress.com The weblog of some guy named Cameron Cox Fri, 01 Aug 2008 03:13:22 +0000 http://wordpress.com/ hourly 1 Comment on The Zend Framework sucks, and why I’m moving to Rails. by moisadoru http://cameronc.wordpress.com/2007/12/15/the-zend-framework-sucks-and-why-im-moving-to-rails/#comment-260 moisadoru Fri, 01 Aug 2008 03:13:22 +0000 http://cameronc.wordpress.com/2007/12/15/the-zend-framework-sucks-and-why-im-moving-to-rails/#comment-260 Wordpess is written in php; just if the language itself allows people to write bad code does not mean it is bad itself. I prefer this over the strictness of some language or framework that forces me to do the things *their* way. And sometimes I don;t realy need mvc or other super trendy patterns, I just let the creativity fly. You should not hate things that much, you'll just get old sooner. Wordpess is written in php; just if the language itself allows people to write bad code does not mean it is bad itself. I prefer this over the strictness of some language or framework that forces me to do the things *their* way. And sometimes I don;t realy need mvc or other super trendy patterns, I just let the creativity fly. You should not hate things that much, you’ll just get old sooner.

]]> Comment on The Zend Framework sucks, and why I’m moving to Rails. by Stas http://cameronc.wordpress.com/2007/12/15/the-zend-framework-sucks-and-why-im-moving-to-rails/#comment-6 Stas Sun, 23 Dec 2007 16:44:59 +0000 http://cameronc.wordpress.com/2007/12/15/the-zend-framework-sucks-and-why-im-moving-to-rails/#comment-6 Cameron, First of all, people have put a great deal of effort to build the Zend Framework, so at least show some respect. Second, if you have criticism, that's fine. That's even great: constructive criticism helps building better stuff. But the keyword is CONSTRUCTIVE! It means no "f"-words, but just to speak to the point: don't like somethig? - propose an ALTERNATIVE - saying something is a total crap doesn't help much. And third: I use Zend Framework, and I enjoy it a lot, as well as many others do. Maybe it's not perfect (and what is?), but it's VERY useful: it cuts the development times and helps you to produce a cleaner code. It's extremely customizible and non-"invasive" at all due to the "Use At Will" approach. I thank the Zend Framework developers a lot for producing such a great framework. Stas Cameron,

First of all, people have put a great deal of effort to build the Zend Framework, so at least show some respect.

Second, if you have criticism, that’s fine. That’s even great: constructive criticism helps building better stuff. But the keyword is CONSTRUCTIVE! It means no “f”-words, but just to speak to the point: don’t like somethig? – propose an ALTERNATIVE – saying something is a total crap doesn’t help much.

And third: I use Zend Framework, and I enjoy it a lot, as well as many others do. Maybe it’s not perfect (and what is?), but it’s VERY useful: it cuts the development times and helps you to produce a cleaner code. It’s extremely customizible and non-”invasive” at all due to the “Use At Will” approach. I thank the Zend Framework developers a lot for producing such a great framework.

Stas

]]> Comment on The Zend Framework sucks, and why I’m moving to Rails. by http://thebigec.com/eric/ http://cameronc.wordpress.com/2007/12/15/the-zend-framework-sucks-and-why-im-moving-to-rails/#comment-5 http://thebigec.com/eric/ Sat, 22 Dec 2007 18:38:44 +0000 http://cameronc.wordpress.com/2007/12/15/the-zend-framework-sucks-and-why-im-moving-to-rails/#comment-5 Wil, I don't think he's wrong about everything.. I don't want to say "Quality" is bad, but consistency is definitely an issue. A quick example at this... look at all of the Zend classes you can extend, such as Filter / Validate, View, Cache (Actually, can't do anything with cache) by adding in your own name spaced classes. All of these classes use extremly different methods to register namespaces & paths of code it can use. Zend_Cache can't even USE any custom code. It gets to be a little bit of a pain in the ass to be honest with you. In the case of Filter/Validate, it would seem you'd want to make these additions static, to apply to your entire application. But that's not how it works, you need to register them for every single instance you make (Or you just need to manually initialize the classes you want to use, which doesn't work as nicely). Initially, I thought the 1.5 release sounded like a bogus marketing attempt to make the framework more attractive to "corporate" types... You say it's not, but when you look back and the hole Filter thing became an issue, it was marketing then that was driving the decisions. How do we know they aren't now? Another big issue I think exists within the framework is open communication. By watching, it appears the Zend team just develops stuff and puts it in, or doesn't use the proposal system right. Thomas's proposals appear to be a case of preferential treatment (half complete proposals being accepted), and there was that entire OpenID issue (Zend working on a component behind the scene, and someone from the community doing the same thing, causing overlap because of miscommunication on Zend's part). But anyways, Wil it appears your getting everything together. I'm glad you guys got the Form proposal in, and it's cool your taking time to work on CLI tools (something I think is very much needed). Cleaning up the wiki was a stellar idea too ;) Regards, Eric Wil,

I don’t think he’s wrong about everything.. I don’t want to say “Quality” is bad, but consistency is definitely an issue.

A quick example at this… look at all of the Zend classes you can extend, such as Filter / Validate, View, Cache (Actually, can’t do anything with cache) by adding in your own name spaced classes. All of these classes use extremly different methods to register namespaces & paths of code it can use. Zend_Cache can’t even USE any custom code. It gets to be a little bit of a pain in the ass to be honest with you.

In the case of Filter/Validate, it would seem you’d want to make these additions static, to apply to your entire application. But that’s not how it works, you need to register them for every single instance you make (Or you just need to manually initialize the classes you want to use, which doesn’t work as nicely).

Initially, I thought the 1.5 release sounded like a bogus marketing attempt to make the framework more attractive to “corporate” types… You say it’s not, but when you look back and the hole Filter thing became an issue, it was marketing then that was driving the decisions. How do we know they aren’t now?

Another big issue I think exists within the framework is open communication. By watching, it appears the Zend team just develops stuff and puts it in, or doesn’t use the proposal system right. Thomas’s proposals appear to be a case of preferential treatment (half complete proposals being accepted), and there was that entire OpenID issue (Zend working on a component behind the scene, and someone from the community doing the same thing, causing overlap because of miscommunication on Zend’s part).

But anyways, Wil it appears your getting everything together. I’m glad you guys got the Form proposal in, and it’s cool your taking time to work on CLI tools (something I think is very much needed). Cleaning up the wiki was a stellar idea too ;)

Regards,
Eric

]]> Comment on The Zend Framework sucks, and why I’m moving to Rails. by Wil Sinclair http://cameronc.wordpress.com/2007/12/15/the-zend-framework-sucks-and-why-im-moving-to-rails/#comment-4 Wil Sinclair Sat, 22 Dec 2007 06:32:25 +0000 http://cameronc.wordpress.com/2007/12/15/the-zend-framework-sucks-and-why-im-moving-to-rails/#comment-4 Hey Cameron, I'm the new lead for the framework team at Zend. I wasn't around for the 1.0 release, but I hope you see a lot this stuff addressed in the 1.5 release. In any case, I'm really glad I came across your blog. We always need more feedback, especially as a corporate-backed open source project. It would be stupid of us to put ZF out there as an open-source project in name and license only; we can always use more two-way communication with the community. And we especially need to pay attention when the words 'Zend Framework' and 'sucks' are used so often in such proximity. ;-) That said, I disagree with you on most of what you've said about ZF. :-P I come from a background in Java/JEE and Ruby/RoR. I have to say that there are many things I appreciate about both of them, and I firmly believe that these things are not necessarily mutually exclusive. Java obviously has more libraries available, but beyond (or largely due to) that fact, Java is an excellent language for integrating with almost any remote or legacy system. But simplicity is not built in to Java or JEE. This is where RoR really comes in to its own. Ruby as a language has a set of features that make most developers drool, and RoR has used it to create a great framework that amazes with its grace. All the same, RoR is not the tool I would pick for enterprise integration. (Why do I care about integration so much? Because in my experience most of the value of enterprise applications is tied up in their ability to integrate with other systems, and- yes- we need to pay attention to enterprise customers because they largely pay the bills.) So, we're trying to do something a bit new with Zend Framework. We want to build a system that's as easy to use as RoR, but can handle all those freaky use cases that come up in the enterprise. Obviously, we're not fully there yet on either count. But I think we're still on the right track to deliver on both of these goals; we've chosen to build up to the most common use cases instead of building down from them. That means it's going to take us a bit longer to get there, but in the end we should have a framework that can easily fill the gap between Java/JEE and Ruby/RoR with the best of both worlds. I can't really express an opinion on the 1.0 release- I didn't make the decision, and I wasn't part of the community at that point. I can tell you that we've spent a lot of time listening to users in deciding what goes in to 1.5. We've got a powerful forms implementation with a few nice AJAX perks, we've got a CLI tool that should really reset expectations on what can be done at the command line, and all kinds of PDF, Zend_Lucene, Identity 2.0/LDAP auth features, in addition to a whole lot of community-driven improvements and even more web services. I'd personally love to hear your feedback on my own CLI proposal: <a href="http://framework.zend.com/wiki/x/wK." rel="nofollow">http://framework.zend.com/wiki/x/wK.</a> BTW, I don't think we 'steamroll' anything based on marketing (although BC is important to us- we do want to bring the latest features to everyone without their having to rewrite their apps, after all), but if you ever change your mind about contributing to Zend Framework, just send me a mail with a link to your issue or proposal. I'll make sure we take a good look at it in the next round of reviews. That said, no preferential treatment for any proposals; some are better off 'steamrolled', as you put it. ;-) Thanks, and best of luck with whatever framework you're using. ,Wil Hey Cameron, I’m the new lead for the framework team at Zend. I wasn’t around for the 1.0 release, but I hope you see a lot this stuff addressed in the 1.5 release.
In any case, I’m really glad I came across your blog. We always need more feedback, especially as a corporate-backed open source project. It would be stupid of us to put ZF out there as an open-source project in name and license only; we can always use more two-way communication with the community. And we especially need to pay attention when the words ‘Zend Framework’ and ’sucks’ are used so often in such proximity. ;-)
That said, I disagree with you on most of what you’ve said about ZF. :-P I come from a background in Java/JEE and Ruby/RoR. I have to say that there are many things I appreciate about both of them, and I firmly believe that these things are not necessarily mutually exclusive.
Java obviously has more libraries available, but beyond (or largely due to) that fact, Java is an excellent language for integrating with almost any remote or legacy system. But simplicity is not built in to Java or JEE.
This is where RoR really comes in to its own. Ruby as a language has a set of features that make most developers drool, and RoR has used it to create a great framework that amazes with its grace. All the same, RoR is not the tool I would pick for enterprise integration.
(Why do I care about integration so much? Because in my experience most of the value of enterprise applications is tied up in their ability to integrate with other systems, and- yes- we need to pay attention to enterprise customers because they largely pay the bills.)
So, we’re trying to do something a bit new with Zend Framework. We want to build a system that’s as easy to use as RoR, but can handle all those freaky use cases that come up in the enterprise. Obviously, we’re not fully there yet on either count. But I think we’re still on the right track to deliver on both of these goals; we’ve chosen to build up to the most common use cases instead of building down from them. That means it’s going to take us a bit longer to get there, but in the end we should have a framework that can easily fill the gap between Java/JEE and Ruby/RoR with the best of both worlds.
I can’t really express an opinion on the 1.0 release- I didn’t make the decision, and I wasn’t part of the community at that point. I can tell you that we’ve spent a lot of time listening to users in deciding what goes in to 1.5. We’ve got a powerful forms implementation with a few nice AJAX perks, we’ve got a CLI tool that should really reset expectations on what can be done at the command line, and all kinds of PDF, Zend_Lucene, Identity 2.0/LDAP auth features, in addition to a whole lot of community-driven improvements and even more web services. I’d personally love to hear your feedback on my own CLI proposal: http://framework.zend.com/wiki/x/wK.
BTW, I don’t think we ’steamroll’ anything based on marketing (although BC is important to us- we do want to bring the latest features to everyone without their having to rewrite their apps, after all), but if you ever change your mind about contributing to Zend Framework, just send me a mail with a link to your issue or proposal. I’ll make sure we take a good look at it in the next round of reviews. That said, no preferential treatment for any proposals; some are better off ’steamrolled’, as you put it. ;-)

Thanks, and best of luck with whatever framework you’re using.
,Wil

]]> Comment on The Zend Framework sucks, and why I’m moving to Rails. by Ninja-tje http://cameronc.wordpress.com/2007/12/15/the-zend-framework-sucks-and-why-im-moving-to-rails/#comment-3 Ninja-tje Tue, 18 Dec 2007 15:06:04 +0000 http://cameronc.wordpress.com/2007/12/15/the-zend-framework-sucks-and-why-im-moving-to-rails/#comment-3 Ja, this makes a lot of sense. We were all hoping ZF to be the Rails for PHP, but bad planning and hasty decisions made it into nothing more than a library. Why not settle for a good set of conventions? I don't see how it would stand in the way of using ZF as a library. Only the MVC parts are glued together as it is, and could benefit so much from a few simple guidelines. Ja, this makes a lot of sense. We were all hoping ZF to be the Rails for PHP, but bad planning and hasty decisions made it into nothing more than a library.

Why not settle for a good set of conventions? I don’t see how it would stand in the way of using ZF as a library. Only the MVC parts are glued together as it is, and could benefit so much from a few simple guidelines.

]]> Comment on The Zend Framework sucks, and why I’m moving to Rails. by Piraatje http://cameronc.wordpress.com/2007/12/15/the-zend-framework-sucks-and-why-im-moving-to-rails/#comment-2 Piraatje Tue, 18 Dec 2007 14:28:14 +0000 http://cameronc.wordpress.com/2007/12/15/the-zend-framework-sucks-and-why-im-moving-to-rails/#comment-2 Funny (and sad at the same time) part is that you actually have a point Funny (and sad at the same time) part is that you actually have a point

]]>