Fruity Zen (PEAR)

[quentinjs]

I have been frequenting the PHP PEAR site more and more frequently and see many modules that could be useful for my own development, but also in development of add-ons for Zen.

I know currently that Zen Cart doesn't use any of these and I am wondering why. The other flip side of the question is... should add-ons or future versions of ZC use PEAR?

This has been a thought troubling me for past couple weeks. I finished of a stable release of a newsletter submission product a couple days ago, which uses a couple PEAR mail related modules, and I had pondered making it work within ZC or allow ZC to hook into it. And also thought about the form module of PEAR and how it could possibly make form creation much simpler as well. Which is fine and dandy for my own pet projects, but my concern is then about releasing them for others to use. They also have AJAX modules that could really enhance many areas of ZC, especially in the Admin area.

What are your thoughts on this, its not totally a ZenCart related topic, I wrote a blog entry on this in my blog at http://www.sarafinchan.com/?p=23 as well.

Hi Ho Hi ho its off to coding I go .....

[kuroi]

I use PEAR for non-Zen Cart work and like it greatly. It offers a helpful, high-quality repository of code to perform common (and often rather boring) tasks that can greatly reduce development time. However I wouldn't recommend it for Zen Cart.

As I understand it, one of Zen Cart's objectives is to be as widely useable as possible. If it were based on PEAR then that would require that users web hosts had PEAR installed. Although this is likely for Linux servers, it may not necessarily be the case for servers running on Windows. And even users on Linux servers may find themselves having to add specific PEAR modules to their environments (where their hosts allow them).

Instead Zen Cart is almost self-contained, and the external requirements seem to be near universal on decent hosting packages. Furthermore, the core functions that might otherwise be provided by PEAR are under the development team's control, thus avoiding the problems that can occur when a package is developed and tested with one set of PEAR modules, but then run by users whose servers are home to a much wider range of versions.

[quentinjs]

:kuroi

What do you feel about PEAR for add-on modules for ZenCart?

How popular is PEAR for hosting services? Is this something that most support, or is its support by hosting companies rather sparse?

[kuroi]

I can envisage Zen Cart add-ons that could be built much more quickly if based on PEAR packages, and indeed, given that we all have limited time to volunteer, might come into existence only by being based on PEAR. However, it's not something that I would do myself, as I try not to have more active add-ons than I can provide some level of support for, and I suspect that anything based on PEAR is likely to become high maintenance. That that would not preclude it being used as a high quality shortcut for client work.

It's actually very difficult to ascertain the extent to which hosts actually support PEAR or a useful installation of it. A basic PEAR installation is bundled with recent PHP installers and will be automatically installed unless a host sets it not to be. So I guess most hosts say yes, they have it. But this really doesn't count.

What they're talking about (and I suspect don't know) is merely a platform for installing the useful stuff. Their customers are completely dependent upon whether the host actually chooses to go ahead and install any of that, and if so what. I've searched the FAQs of four web hosts recently without finding any reference whatsoever to PEAR. When contacted, only one (a Zen Cart certified host) seemed to have any real idea what I was asking about. I suspect that this is a fair indication of how much attention they pay to it.

On the other hand, there is a very good guide in the PEAR manual to installing a local copy in a shared hosting environment and oodles of unofficial pages helping with the same thing. This rather implies a significant need to fill this hole left by the hosting companies.