Re: idea for new TriJUG endeavor

   My one simple idea is a 15 minute analysis after the talk or before the
next talk. Alot of the presentations have some fundamental engineering
issues which could be best discussed by our group when the presenter is not
present. Most (all?) of the tools have precursors or competitors where
lessons have been learned. I guess if the meeting ended 15 minute earlier
then maybe it would facilitate people to talk afterwards. ???

  In general, I think we need to focus on Computer Science and try to return
the industry to qualified engineering. Many projects I've seen and companies
I've interviewed are using a seat-of-the pants design, follow keyword
architectures, and hire lowest-cost labor. The opposite of 15 years ago when
I entered the industry. If this trend continues then so will the downward
wage pressures. We will all be making too little to justify our love of
computer programming.

I'm not sure how to effect the change but maybe highlighting the projects at
local companies and what really good work is being done would help bring up
the overall level of professionalism in the area as well as give us insight
into what companies are desirable and have great projects and what ones are
not laudable/desirable. As one example:. I'm sure many on the list have
automated DAO generation so maybe we'd find that is standard practice and
really hand-coding DAO's is just not very professional. Etc.

A lot of the presentations on frameworks are very interesting but not
applicable to my day job. :-) Good to attend as maybe they could be useful
int eh future though. Maybe a focus on automation tools, GUI design tools,
load testing tools, and even a standard of best tools and practices would
help raise productivity. Heck, being a TriJUG member could one day carry
weight on a resume!

My $.02,
TimJowers
P.S> I tried to play with the Flex stuff but could never get their dev tool
to install. The install ran but the tool never appeared.
P.P.S> Some other thread was about clustering. Beowulf is the dominant (in
number of installs in Top 500) super computer clustering product and is open
source. Teradata is an MPP RDBMS with specialized HW. Oracle has tied itself
to its Star schema which is really better for single row queries and online
data processing rather than complex queries and clustered workloads. Oracle
did buy Times 10 which hacks MPP at the driver level from what I remember.
Web loads are inherently easily clustered so no special distributed
computing/clustering is needed other than a basic load balancer.


On 10/27/07, Phillip Rhodes <[email protected]> wrote:
>
> Robert J. Cameron wrote:
> > I have some ideas, and I'm thinking I may self-nominate for the board
> > to try and make them happen...
>
> Awesome.  You got my vote. :-)
>
> I'll do whatever I can to help as well, whether I'm on the board or not.
> I do like the idea of maybe forming some committees to handle various
> projects (if volunteers are available) so it makes it apparent that we
> can't expect "the board" to just magically make everything happen.
>
> A few ideas we spoke about on IRC the other day also include maybe
> trying to get a reading circle / study group together to study a
> particular book, or maybe a SIG/BOF type group that actually do some
> hands on coding (like the old Struts Study Group), and maybe
> try doing a "hack night" event like the Ruby guys do.
>
> I think any or all of these things can be done without needing
> any specific involvement from the board, other than maybe blessing
> the activity as an "official Tri-JUG activity" and maybe mentioning
> it on the website.
>
> Lee spoke a little about getting a Wiki going again on the website,
> which would be great for helping coordinate this kind of stuff.
>
>
> TTYL,
>
>
> Phil
>
>
> _______________________________________________
> Juglist mailing list
> [email protected]
> http://trijug.org/mailman/listinfo/juglist_trijug.org
>
>
>

_______________________________________________
Juglist mailing list
[email protected]
http://trijug.org/mailman/listinfo/juglist_trijug.org