Re: Q:did anybody think of porting scotty in java ?

Juergen Schoenwaelder (schoenw@gaertner.de)
Sat, 5 Jul 1997 20:28:29 +0200

"(U)(rsm) WETZEL Damien" <d.wetzel@cstelecom.cie-signaux.fr> said:

U> this is a totally innocent question, as i m beginning to look after
U> the java stuff?
U> first would it be usefull to port a very valuable software like scotty
U> tkined in java ? if anybody is thinking of doing it or has already
U> heard of something equivalent, i would be glad to hear about it ?
U> are tcl and java done to get along in an application like scotty ? or
U> is it stupid hype (sun are very good marketers anyway).

There is no way to "port" scotty and tkined to java. The only option
you have is to "re-implement" scotty and tkined in java. The questions
is whether this is a good idea or not. What are the reasons for
rewriting scotty and tkined in Java?

- Portability. Java programs are said to be highly portable. Experience
with different Java toolkits tells us that this is not 100 % true.

- Accessability. Loading embedded Java code from HTML pages looks like
an attractive way to distribute programs to the masses. Is the scotty
tkined really meant to be used by masses?

- Object-orientation. People say that OO leads to better and to reusable
code. However, you can get reusable code without OO as well. The
key is the design behind a library, not the programming model. Note
that the Tcl command layer of the scotty package has been reused
successful by many people around the globe without ever looking into
the C code. I am not sure what OO will buy us, especially when doing
network communication.

- Integration. People believe that Java will become a dominant language
and hence integration would be easier if everything else can be
integrated easily. I think this is a reasonable argument. However,
integration of existing code does not necessarily mean that you
re-implement the code. There should be other ways to do it.

There are probably other reasons why people want to use Java (the list
above is no attempt to look at the complete picture), but I do
currently not see something which is convincing enough to rewrite the
50.000 lines of C code and the 23000 lines of Tcl code in Java. I see
good reasons to provide interfaces to Java or other systems, but this
is a different story.
Juergen

-- 
Juergen Schoenwaelder     <schoenw@gaertner.de>     (Tel: +49-531-23873-0)
Gaertner Datensysteme, Hamburger Strasse 273a, 38114 Braunschweig, Germany
--
!! This message is brought to you via the `tkined & scotty' mailing list.
!! Please do not reply to this message to unsubscribe. To subscribe or
!! unsubscribe, send a mail message to <tkined-request@ibr.cs.tu-bs.de>.
!! See http://wwwsnmp.cs.utwente.nl/~schoenw/scotty/ for more information.