[MUD-Dev] Custom Server Roll Call?

[Marc Hernandez]

Ola Fosheim [iso-8859-1] Gr=F8stad:
> Marc Hernandez wrote:
> >         You can get some pretty good emacs files for either c++ or java=
,
> > or write your own.  I currently use microsofts Dev Studio with Sun's ja=
vac
> > compiler.  The only thing it does not allow me to do is double click on
> > the error and go directly to it.  With a little hacking that could be d=
one
> > but it is not too difficult.  I am moving to emacs soon, but I do not h=
ave
> > it setup yet.

> Why moving to emacs??? :) Are there emacs setups which give you good
> class-browsers and such? What is wrong with microsofts visualstudio stuff=
?
> I've never used it, but will consider buying if it increases my
> productivity.

=09Well.  I do like Visual Studio, but currently I have been doing a
lot of work on a fairly diverse set of systems (Win95, Solaris, Linux,
FreeBSD).  I switched my main platform of development to Linux (and have
been very happy.  It is even pretty now).  I now need a system that is
fairly cross platform and emacs fits that.  As far as all the nice
browsing stuff in concerned I do not know.  There are some for Java and I
would imagine someone has one for emacs.
=09On the non personal side I am doing research on line coverage (as
well as variable definition/use coverage) and our tools can tell emacs to
highlight certain lines (typically the covered ones)[1].  EMacs can do
lots of stuff.  DevStudio might also but the typical users of this at
least have a passing acquantance with emacs and many might not have
DevStudio.=20

> >         Actually I plan on using Java applications for the client.  The
> > server is currently an application.  This has no security restrictions.

> Ugh. Then you miss out on Java's only advantage, but it is a good one. "C=
ome
> visit my MUD, it is only a click away!"  This is an advantage I would not
> like to miss out on! Even if you only can offer a limited client it would=
 be
> worth a lot. This is basically what The Palace are doing. They
> encourage/entice you to download the real client by telling you about the
> extras you will get. For many the Instant-Palace is enough though.

=09I agree that there should also be a possibly limited client.
However I would not say that that is Java's only advantage.  It is however
a unique advantage.  Java is still slow but is getting faster.  Cygnus has
a .class -> object || .java -> object || .java -> .class[2].  The .class
file format is stable and fairly easy to target (for other languages).  In
fact LPC would probably do well to target it :-).  Too many fun projects.

<speeding up client rendering snipped (and agree)>

> > We came to the conclusion that since:
> > The user has to download everything anyway.  Why not put it in one easy=
 to
> > install zip?

> Because clueless users will not know how, or even if they do, they might =
not
> want the hassle? They don't neccessarily have to download all the graphic=
s
> initially. If I read from an URL in an applet, will that file then end up=
 in
> the browser cache?  If you use progressive JPEG then you can start using =
the
> graphics before it is loaded as well. (Ok, the programming can be a littl=
e
> messy maybe. I haven't quite got the grip on clean programming using
> java-threads yet.)

=09Yeah.  Having a limited client probably is ideal.  I liked the
ideas put forth on this list earlier of sending sets of progressivly
detailed/specific graphics and sound to the client (put forth by JC I
believe).

> I think java archive format (jar) is zip, but will it be cached in tbe
> browser cache?

=09Probably.  However the jar could be larger than their cache
(default cache sizes are ~7 megs or less).

> > Sound is finally loadable off of disk.  Java sound support is not.  I d=
o
> > not even know why they say they have an 'api'.

> Isn't (nongenerated) sound loadable off disk using jar even in applets? B=
ut
> the sound support is really pathetic. They should at least let me set an
> amplitude/envelope for the sound, not to mention changing the pitch, or a=
t
> the very least let me generate soundfiles from raw data. :(  I don't real=
ly

=09Not sure.  I did not really mess with any sound stuff.  Sun saying
they have a java.sound API is certainly very misleading at the moment.

Marc Hernandez
5 billion neurons can't be wrong.
[1] We also then pull those tests out of the code then the user can rerun
their program.  After a few iterations of this you can easily tell what
your test cases are not covering.
[2] Yes .class -> object and .java -> .class implies .java -> object (but
only if there is a transative relation..)  Not that anyone would care but
that is why it is a footnote.



_______________________________________________
MUD-Dev maillist  -  MUD-Dev at kanga.nu
http://www.kanga.nu/lists/listinfo/mud-dev