Thank you very much, this does clarify quite a few points. It seems
that AcePerl is the way to go for networking, at least over the Web, but
Java could be used on a controlled intranet (my situation).
Interesting to know that AcePerl was actually derived from Java. It
definately was confusing to see webace here and acebrowser there, but
you have cleared that up now, thanks again.
Gudmundur
In article <199907151932.VAA00492 at alpha.crbm.cnrs-mop.fr>,
mieg at alpha.crbm.cnrs-mop.fr (Danielle et jean Thierry-Mieg) wrote:
> following a question by Gudmundur zgudmunt at my-deja.com>> the jade code i wrote with lincoln was a real try at interfacing
> acedb to java
> it works very well, with local caching, threading etc and is
> a very evolved software in terms of database
>> what was not so complete in our 97 work was the graphic interface
although it
> did reproduce the sort of graphics we have in webace and in the recent
> aceperl.
>> In addition Jade has authentic java graphics following the
> model of exchanging data via tables which we discussed at length in
> (Stein. Cartinhour, ThMieg, ThMieg, Gene, 1998, 209,pp 39-43)
>> This implies that Jade also suppports data imported from other
> databases, as for example in our GDB demo which gets its data from a
> relational Oracle server.
>> However, over the last 2 years, the java graphic library, JDK, evolved
> in a very disappointing way. Most programmers i know who have tried to
> develop in java have been taken aback by the diffculty of writing
> applets: write once-debug everywhere !
>> Java really works well only in intranet where you have control of the
> browser beeing used or as stand alone applications
>> For those reason Lincoln ported the jade library in perl and i adapted
> the server, constructing in this way the central core of the AcePerl
> library.
>> Contrary to java, we have had no problem at all with perl.
>> AcePerl is a great success and the webace code of Bigwood et al,
> (thanks Doug for this excellent work that we did use for deveral
> years), has now been superseeded on several sites, including
> www.sanger, by aceperl, although, and this is very confusing, aceperl
> was renamed webace on the sanger site.
>> Webace is the code of 95 by Bigwood et al,
> Aceperl is the code of 98 by Stein et al,
>> Both import data from the acedb server but
> actually in a different format
> webace uses the -p[erl] format
> i.e full fleshed perl acedb objects
> whereas aceperl uses the same -j[ava] format
> i.e. the format we developped for jade
>> ---
>> As far as connection goes, AcePerl talks rpc to the acedb rpc server
> and Jade talks socket to a small daemon running on the web server
> called jade2ace which itself talks rpc to the acedb rpc server located
> presumably on a different machine
>> Note that even when Ed Griffith and I finish the acedb socket server
> (say in a couple of weeks), we will still need a jade2ace daemon
> because the applets are not allowed to talk to any machine but the
> webserver
>> however standalone java applications will be able to talk to the
server
> directly (i.e. intranet type java usage)
>> ---------------
>> The real benefit of the socket connetion concern aceperl and the
xaceclient.
>> xaceclient is a full fledged graphic acedb which automatically
> import data from the server
> The current xaceclient only works on Unix machines because it talks
rpc
> with the new socket protocol we will be able to port xaceclient
> to a mac and pc
>> Aceperl like xace, on a non rpc machine, can only talk to a local
database
>> ===
>> i hope this clarifies a little the discussion
> .
> ---
>
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
