Re: [phplib] The change to PEAR... From: Alexander Aulbach (ssilk <email protected>)
Date: 01/28/01

On Thu, 18 Jan 2001, nathan r. hruby wrote:
}maintainership of phpSlash :) And I think some of the problems are not
}fixable by one individual, they are larger issues that need to be
}addressed by the php community as a whole [ie: module desgin (dealt with
}to a certian degree), keeping in synch with PEAR (for both module
}devlopers and app devlopers and end users)], as well as the designers of
}PHP [ie: namespaces]
}
}I think there are a lot of unanswered questions and a lot of thought
}that still needs to go into PEAR before it can be a truly invaluable
}resource, and I worry that putting phplib into PEAR now will have
}detrimental effects as PEAR hits the wall with these questions and
}instead of properly devloping PEAR into its own thing it'll become a
}patchwork of diffrent standards and implementations. I'd like to see
}this merger of phpLib into PEAR as a good reason to work on these
}questions and issues so they don't bite all of us in the ass later.

As said before, PHPLIB's main point is, that we have some classes, mixing
them together and if they don't fit our needs, we write an enlargment for
it, a new function here, some new vars there and my small and beautiful
application is ready.

This works, and it will work in future like this. But what will happen, if
you take some of those builded applications and put them on a virtual
server - but they should for example use the same session-data
(Application A and B use the same variable $hugo for completly different
things). I could really say, that this kind of creating web applications
isn't state of the art anymore (too comlex, too expensive, no long time
income), but there is currently no other way to do it.

So I thought about, what must be done, to reduce all of this afford? How
can we make it easier, to make web appications. How can we reduce the
overhead? How can we make a toolbox, where we could grasp into, take a
hamer, and knock it over another application? I don't know. This is too
far, too complicated, too expensive and nobody knows, if this is really
that, what we need.

Then I came to the conclusion, that we need something, that can be used as
a light at the end of a tunnel. What is really missing, is an aim or
direction.

An aim, where all this development should end and how it should be
reached. This aim should be changed several times, as long as it really
fit's our needs.

Currently I don't see any aim. I have for example other aims than
Christian or you. So the aim of PHPLIB is looking like a pancake:

           ____________
          / \__
        / ^ \
       | |
       | < PHPLIB > |
       | |
        \ v /
         \__ /
            \___________/
          

PHPLIB is growing and growing, until it is hot under it and the dough is
soft.

[BTW: the same problem is in my opinion with PHP: everyone put dough on
it]

We need an aim like this, to bring the cake into a "better" form:

        PHPLIB ------------------------> HRNZ

But this aim is too big, much better:

        PHPLIB -> PEAR -> HUGO -> BLA -> HRNZ
                  ^ ^ ^ nice wish
                  | | less concrete
                  concrete aim

Everyone who has studied software development, knows how to do this.

Let's define what we want. This is difficult enough, cause everone has
other needs. And then let us define, how we can reach it in very small
steps and in a way, that we come not into a dead end. 

-- 

SSilk - Alexander Aulbach - Herbipolis/Frankonia Minoris

---------------------------------------------------------------------
To unsubscribe, e-mail: phplib-unsubscribe <email protected>
For additional commands, e-mail: phplib-help <email protected>