Hi,

The post by tilly has some valid points about lack of direction (the
organisation of the project is not that bad if you read the mailing list).

Aaron's post says:
* Easy access for Architects, Designers, and Developers to the tools,
  techniques, and guidelines to write enterprise software in Perl.
* Access to knowledge for System Administrators to install and operate
  enterprise systems which include enterprise applications written in Perl.
* Provide IT managers' confidence in hiring enterprise architects who
  specialize in Perl, in supporting enterprise development projects in Perl,
  and in deploying enterprise applications in Perl.

Basically that is access, installation/administration, confidence is
either generated through marketing or quality of the project.

But access to and installation of    WHAT ?

If what is supposed to be enterprise software, then what kind of enterprises,
which tasks, etc. This is no specification at all. It would be fine if
the above three points are the goals of a community and define the tasks
of building an environment. But that doesnt relate to code. Unfortunately
people expect code, as as the post on perlmonks say, they expect "their"
code. To me sufficiently good code would be enough. I dont say all the
small parts that are currently worked on and spoken about (serialization etc)
are unimportant or bad code. But if you have big projects, putting together
some modules is not enough at all. You need a architecture that spans the
whole thing.

There is almost no talk about architecture (in this big scale) on the list.
Maybe all but me know already how it should look like, but I doubt this.
You are not touching the big and hard parts like:
-concurrency
-object models
-distributed applications (running the same program at several places)
-specification/verification of programs

That is whats expensive to do if you are a alone. Everyone can write their
own module serializing data in his own special way.

If we could provide solutions to the parts above we will attract more people.
Because this is what they are struggling with if they have big projects.
And we have something for WHAT (Note that for small projects this is not
needed and a P5EE community should be enough).
And we should work on these parts.


Torvald

Reply via email to