RE: [stellation-res] Windows Port

["Jonathan Gossage" <jgossage@xxxxxxxx>]

> -----Original Message-----
> From: stellation-res-admin@xxxxxxxxxxxxxxx
> [mailto:stellation-res-admin@xxxxxxxxxxxxxxx]On Behalf Of Mark C.
> Chu-Carroll
> Sent: Tuesday, March 18, 2003 8:09 PM
> To: stellation-res@xxxxxxxxxxxxxxx
> Subject: Re: [stellation-res] Windows Port
>
>
> Jonathan Gossage wrote:
>
> >I have now got Stellation working under Windows XP. The testing
> environment
> >included the Sun JVM version 1.4.1_02 and the database was
> Firebird 1.5 Beta
> >3 with the Jaybird JDBC driver 1.00 RC3. All tests were run on a dual
> >processor 1,6Ghz machine with Athlon P2000 processors and 2GB of memory.
> >
> >I have both local and server based operation working. It is
> interesting that
> >running a server environment rather than a local environment where both
> >client and server are on the same physical machine is nearly 3
> times as fast
> >with my test configuration.
> >
> >I found that running the ant script in a server based environment took 5
> >min. 45 sec. whereas running it in a local configuration took 14
> min. During
> >the remote tests CPU usage on both processors was approximately 50%. Go
> >figure :-)
> >
> >The next step is probably to address documentation. Our current
> >documentation is completely Linux centric and I think it would
> be difficult
> >to combine the two environments in a readable interleaved form.
> Accordingly
> >I recommend that I produce separate Windows oriented documents.
> If you all
> >agree with this approach, I will try to factor out common stuff
> that is not
> >platform specific into includable units that can appear in both
> documents.
> >
> >
>
> Congratulations on getting it working! That's wonderful news. Can
> you please
> post the patches for the change?
>
> I'm a little concerned about the overlapping documentation, but I think
> my concern
> is easily addressable.
>
> The root of my concern is this: any manual approach to sharing common
> documentation
> segments is almost inevitably doomed to failure. The solution, of
> course, is to automate it.
> I can write a simple processor to do includes, as long as we agree on a
> syntax. I'd
> propose the following, using the XML processing instruction format
> (which is exactly what PIs are intended for.)
>
> <?include file="filename"?>
>
> If there is general approval, I'll toss together a simple processor
> written in Java that expands this.
>
>     -Mark
>
>
>
> >Thoughts
> >
> >Jonathan
> >

Since I have absolutely no experience with Web programming of any kind I
will have to let you take the lead on this. All I want is to minimize the
documentation maintenance effort as much as possible.

I have created Bugzilla entry 35267 for the work I have done. Enjoy and tell
be what I will have to do to make use of the include stuff you are
proposing.

Regards

Jonathan