🚀 go-pugleaf

In <342A57CF.979@ibm.net>, John Littler <john.littler@ibm.net> writes:
>It's streets ahead of Active X but it ain't 100% see
>
>http://www.zdnet.com/pcweek/news/0922/22java.html
>
>yeah yeah it's Ziff Davis and we know how far we can trust them if it's
>not made by Micro$loth, but the article appears to be based on an
>independent 3rd party.

     So you can get some data structures past the verifier - so what?  This 
doesn't mean that Java will be able to suddenly have access to hardware to 
overwrite or corrupt files, for example.

     And as the story states, JavaSoft claims that all of this is based on 
bugs in the verifier in early versions of Java.  But even if it isn't fixed 
as claimed, I doubt if it's cause for alarm - the JVM itself is still fairly 
restricting as to what a Java program can do (a Java program cannot, for 
example, tinker with CPU registers, access memory outside of the JVM's memory 
pool, gain direct hardware access to any devices, or make BIOS calls).

     And even if someone came up with a way to use this to make a Java program 
dangerous, I don't see how we as OS/2 users need to worry.  Exploiting a 
problem like this tends to be highly system-specific.  And with OS/2's 
excellent application protection model, we're pretty safe.  A Java program 
can't do anything special that a native OS/2 program can't do, such as violate
protected memory spaces.

     If the data in this paper you point us is conclusive, then it is simply 
an oddity that needs to be fixed, and not something to be overly concerned 
with at this time.

Brad Barclay,
OS/2 Development Director,
Internet Direct Canada Inc.

---------------
From the OS/2 WARP v4 Desktop of Brad Barclay.
E-Mail:  os2@idirect.com  WWW: http://warp.idirect.com
Finger for Public PGP Key.  [ ] VoiceType Dictated.