I am new to P5EE, and interested I helping. I have been looking for a way
to contribute to the community. This is the area of Perl reputation the I
would like to see improve the most.
I agree with every thing that Stephen has said.
I am not sure this is the place to contribute what I would like to offer
just yet. Please help me determine where is the best place to offer what I
have to give.
I am a system engineer. I almost never see Perl articles, books, or modules
written for a system engineer. So, I bet your are wondering what a system
engineer does with Perl. I use it from testing system architectures,
modeling and simulation, performance testing, gathering metrics, packet
sniffing, and more. I began using Perl Oct 2002, here are some of the
things I have written:
A packet capture and analysis system integrated with a GUI test
automation tool, Robot. Robot starts and stops the captures as well as
taking timings of function being executed. This information is
correlated and displayed by a CGI script. This correlation helps, to give
developers a better idea of where there code can be optimized, find
duplicate database calls, ill formed sql statements, unexpected code
execution, and many other issues that may have gone unnoticed or been much
more difficult to trouble shoot.
A tool that creates user workflows of the application we are developing.
Our application has been ARMed and our customer is using Tivoli to collect
the ARM data. ARM is a basically code that has been inserted into or
application to take timings of specific functions. Tivoli is used to
collect these timings and store them in a database. My tool takes this data
and uses it to generate workflows with GraphViz. This has changed,
the way we test our application, the how we prioritize bugs, and prioritize
performance enhancements.
The tool I am about to finish at the moment is multi-threaded. Our
customer would like to change the architecture of our backend. As a result
some modeling and simulation of the proposed changes needs to be
performed. Packet captures of our client application talking to our backend
servers where taken. The algorithms I had already developed where used to
extract the sql statements. I then built a script to variablize the sql
statements. These statements are used in a multi-threaded script to
simulate different numbers of users against different hardware,
databases, and OSes. The script times each sql statement and records this
in a separate database. The data produced is used to determine which server
configuration is best suited for the job.
I am currently working on an architecture for enterprise Web Service for a
DoD project using Apache, mod_perl, mod_soap, COM, ATL, Oracle, and much
more. I am anticipating it will be difficult to sell because of Perl rep.
Some of how I plan to over come this will be with the home grown tools I
have described in the list above. Is this the correct place to contribute
the tools I have developed?
Thanks,
Jay Flowers
"Stephen Adkins" <[EMAIL PROTECTED]> wrote in message
[EMAIL PROTECTED]">news:[EMAIL PROTECTED]...
> Hi,
>
> As a follow-up to my conclusions from the last vote ...
>
> http://archive.develooper.com/p5ee%40perl.org/msg01105.html
>
> and particularly this point
>
> 2. The P5EE project should focus on explaining the various
> solutions to the Enterprise Development problem rather than
> trying to adopt or create a single one at this point.
>
> I would like to propose a new focus for the P5EE project.
>
> I think that the mission of the project is still intact.
>
> The mission of the P5EE project is to promote the development,
> deployment, and acceptance of Enterprise Systems written in Perl.
>
> However, I think the right strategy to achieve this is to allow
> people who do Enterprise Development in perl (or know how to and aspire
to)
> collect into groups with a more-or-less consistent way of handling
> enterprise development problems.
>
> Each group needs core members, who actively use and develop
> Perl components in an overall architecture which solves the problems
> of enterprise systems
(http://www.officevision.com/pub/p5ee/definitions.html)
> and who are interested in documenting that architecture and supporting
> others in the use of it through the P5EE mailing list.
>
> I think that every Perl *component* generally has a community (i.e.
> AxKit, SOAP, POE, Alzabo, Template Toolkit) through its mailing list,
> but what is lacking is a community surrounding a complete enterprise
> application architecture with all of its constituent components.
>
> I think that these groups will generally form around major/popular
> frameworks such as AxKit, bOP, Extropia Toolkit, OpenInteract, etc.
>
> However, what is needed to become a P5EE Group is core member(s) who
> are willing to pull together (or write) components in a coherent
> architecture which increasingly satisfy all of the requirements
> of enterprise systems
(http://www.officevision.com/pub/p5ee/definitions.html),
> *document this architecture*, and promote this architecture by
> supporting users of this architecture through the P5EE mailing list.
>
> Note that the following three elements are critical for a P5EE Group.
>
> 1. COMMUNITY
> 2. CODE
> 3. DOCUMENTATION
>
> For instance, I volunteer as a core member for a P5EE Group I would
> call the "App-Context Group". (I am recasting P5EEx::Blue as
App::Context.)
> http://cvs.perl.org/cvsweb/p5ee/App-Context/
> I am committed to documenting an architecture that incorporates many
> perl components (many of which will be shared by other P5EE Groups)
> and which increasingly satisfies the requirements of enterprise
development.
> I am committed (for the foreseeable future) to supporting other users
> who wish to adopt this architecture (subject of course to availability,
> like all open source efforts) and working with them to continually
> enhance the architecture, its components, and its documentation.
>
> I believe that if we could get several "champions" to make the same
> commitments, we could get a handful of P5EE Groups going and
> reinvigorate the P5EE project. (Note that a P5EE Group does not need
> to be able to satisfy all of the requirements of enterprise development
> on day 1. But it must be committed to *increasingly* satisfy these
> requirements.)
>
> This would allow some developer who is investigating the feasibility
> of enterprise development in perl to find the P5EE web site, identify
> the various *actively supported* architectures, and gain confidence
> to get on board with one and begin developing his application according
> to its guidelines.
>
> The reason that I think it is desirable for these to all operate
> on the P5EE list (or future sublists) is so that there is lots of
> interaction and interoperation, and perhaps some convergence.
>
> What do people think about this?
> Are there others who want to step forward to form other P5EE Groups?
>
> Stephen
>
>
