W3C home > Mailing lists > Public > www-jigsaw@w3.org > September to October 1996

Request for try{}catch

From: Anselm Baird_Smith <abaird@www43.inria.fr>
Date: Fri, 18 Oct 1996 13:49:51 +0200 (MET DST)
Message-Id: <199610181149.NAA25864@www43.inria.fr>
To: Alexandre Rafalovitch <alex@access.com.au>
Cc: www-jigsaw@w3.org
Alexandre Rafalovitch writes:
 > I just ran into the problem that made me wish Jigsaw had a bit more
 > protection against the programmer.
 > I will give an easy example.
 > I created the class based on the other file. It compiled fine, but when I
 > tried to add it to Jigsaw, it bombed out.
 > Looking through the trace, I realized that the mistake was mine. I forgot
 > to change the name of the file in Class.getForName() so was getting
 > non-existing class. That caused Jigsaw to bomb out without any way to save
 > data or anything else.

Yes, that one was intentional (at least at the time I decided to go
that way). The real problem, in fact, is that by the time you detect
the error (ie in a static block), you don't have much infos on where
and how to report the error properly...

 > It would be very nice if the process of creating new resource had a big
 > try/catch close and used that to see if resource creation would bomb out
 > and if it will, refuse creation of resource the same way it does if it just
 > can't find it.

Hum, I guess you cannot thow an exception from a static block, so you
won't be able to catch anything outside that block...

Received on Friday, 18 October 1996 07:50:29 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:25:30 UTC