[UPHPU] PHP Framework Project Launch - Coders welcome to attend

Mac Newbold mac at macnewbold.com
Thu Jun 2 12:55:15 MDT 2005


Today at 12:21pm, John David Anderson said:

>> I'm interested to see what others think about these. I honestly think that 
>> only BinaryCloud is even worth taking a second look at, for my purposes. 
>> It's the only one that appears simple enough and flexible enough and general 
>> enough to accomplish the widely varied things I need to do on the sites I 
>> build.
>
> I agree - I wanted to post this quick list to show that there are things like 
> this out there, and hopefully we can learn from their mistakes and successes. 
> I agree with you on you commentary of each.

Wow! That's a surprise. I thought I was for sure going to offend people 
who really like the frameworks. But learning from them is _exactly_ what 
we need to do. Some of them (Seagull I think) claim to have done the same 
thing, but they look more like a 1-off than a new and different thing.

>> After going through this list, I realized how to describe one of the key 
>> things I'm after: a framework should make it much faster to develop a 
>> quality site.
>
> I vote we make this the mission statement for the project. :o) Another purpose 
> of a framework is to make maintenance easier.

Yes, maintenance is important to me too. One thing that would be great is 
if it could be relatively easy to incorporate upgrades to the base 
framework stuff without having to redo or re-merge a bunch of the 
customizations done on the site. But that's probably secondary to the 
faster and easier goals, for me anyway.

>> Some of these are so big, so complex, and would add so much development time 
>> overhead that the time savings might not ever appear, even after I've done a 
>> few sites with it. When I think of learning curve, I think of something that 
>> may have some slope at the beginning, but after I've done two or three 
>> projects, should be basically flat. The "development time overhead/savings" 
>> of a framework has a curve too... on the first few projects, it's higher, as 
>> the learning curve is steep, but by the time the learning curve flattens 
>> out, there should be a dramatic and consistent time savings from using the 
>> framework. With a lot of these systems, I'm far from convinced that this is 
>> the case.
>
> Agreed. I've tried starting up personal projects using these - just to learn 
> them - and haven't been totally convinced that they help out.
>
> What we may want to look at is an approach more similar to PEAR - a set of 
> classes you can use if you want quick functionality.... It seems like many of 
> these frameworks approach the task, trying to encapsulate the whole 
> application under it's umbrella.

I think our framework should be using stuff like pear where appropriate, 
but what I had in mind is more like this other suggestion:

> ...Or maybe more like a set of PHP scripts that ask for what you plan to do. 
> Run them, and they take your design and give you a starting bare-bones site 
> that you can then move on to customize.

Even a little install script could do that, given some simple logic about 
dependencies between the pieces. That would give us the ability to make 
different versions of different pieces that weren't intended to be used 
together, but as alternatives. For example, we could make some basic 
layouts and let the user choose between things like header-body-footer, 
header-sidebar-body-footer, three-column static, three-column fluid 
(w/tables instead of css floated divs), and so forth, for layout choices 
and we'd just put in the right files for them.

Mac

--
Mac Newbold		MNE - Mac Newbold Enterprises, LLC
mac at macnewbold.com	http://www.macnewbold.com/



More information about the UPHPU mailing list