On Wed, 21 Aug 2002, Dana Spiegel wrote:
> Bon, as a community, we certainly appreciate the work you've put forth
> with your database technology. You're developing a solution that is very
> robust and I'm sure quite useful for your purposes. I will say, however,
> that the features of your database system don't entirely match what we
> (the people who have been involved in this discussion) have been looking
> for in a replacement for our maps database. Foremost on the list is
> usability of the web interface, and unfortunately your database system
> doesn't yet match what is available in our current system.
I am not aware you or anyone in the NYCwireless owns any system when you
say "our current system". I DO know that current map db is operated (and
owned?) by groups out west. NYCWireless just uses someone's service. If
you do own your system, I will be interested in seeing what usability of
the web interface that you have.
Let me make you a counter offer. Since you have said below you have a lot
of experience on Oracle and other db systems, and you time is limited, how
about you take up the design and implementation on the web interface and
make it to the highest standard for the usability of the web interface you
have in mind. I will take care the rest, and I will offer to integrate
your design and implementation into our Go-Wireless system so that it can
also serve whoever want whatever usability of the web interface you
described. In this way, your limited time can be utilized
optimally to benefit the community.
If you accept the counter offer, I have no problem to make your design and
implementation opensource (or not) at your decision. In that way, you get
what you ask for as you position yourself as a developer and sysadmin, ---
and the whole community and others can benefit from your work.
How's that?
> As for opensource vs. proprietary systems, let me state that for us (the
> people who head up certain aspects of NYCwireless) this is not the
> issue. My primary concern for our maps database is that we (as
> developers and sysadmins) can easily make changes to the system whenever
> necessary. Unfortunately, were we to go with a system hosted by a third
> party (for example, queens college), we wouldn't have the direct access
> to make the changes that we need or want to make.
I do not know where you get this idea about hosting by a third party like
Queens college. John Klos has offered the hosting service, I have no
problem of it, or anyone who kindly offers resrouces for hosting. The
problem line is to serve the community. I am NOT interested in controlling
resources, nor am I a control freak. I have enough of other things to be
busy with.
Wouldn't it be great if there are voluntneers out there who are
willing to host for the community, if nothing else, as an alternative and
redundant option? Wouldn't this make sense when you mention you are just
like everyone else who has only limited time?
>
> Furthermore, we are all volunteers with limited resources and limited
> time. We need to go with a solution that will minimize the amout of time
> and effort needed to maintain and administer. I've had a lot of
> experience with Oracle, but I can't speak for the other people who do
> sysadmining for NYCwireless. I know from experience that Oracle requires
> alot of knowledge to operate properly and efficiently, and we don't have
> that knowledge or resources to make Oracle worthwhile.
Well, then you can take on the sysadm duty with your extensive
experience on Oracle by contributing your skill for NYCwireless even if
Oracle is used. No?
I do not know what you have in mind on the Oracle sysadmin. Our system
pretty much like being held on a system that's relatively self-maintained
like any good Unix system.
Just to give an example, we built one Intel system running Linux (not
reason why it couldn't be other Unix variant) to host 4 Oracle
application projects in one single Oracle instance. Yes, ONE single Oracle
instance. One of these application contains national wide and world wide
environmental data with one table containing 6,000,000 records for our
research. This, of course, is not just finanical application that requires
real time access to data that could easily be at least ten times more.
How much does it cost to get someone to do the admin of the system just
described? So far just few good students doing on voluntarily basis. I am
happy to see so far our response time and system reliability is no worse
than any system you embrace or systems that I witness in the NYCwireless
circle.
> MySQL, on the
> other hand, is easy to administer, is known by everyone, and operates
> quite well under any server load we could expect.
As John Klos has said it more than once, a system is only as good as it
serves whom it intends to serve. If NYCwireless really wants to stay on
MYSQL, why not give John Klos a chance by giving him the data, source
code, and schema, and see what he can do about it. I am sure anyone with
the reputation on the line will strive for the best ;)))
Bon Sy
--
NYCwireless - http://www.nycwireless.net/
Un/Subscribe: http://lists.nycwireless.net/mailman/listinfo/nycwireless/
Archives: http://lists.nycwireless.net/pipermail/nycwireless/
