Title:

Using Threads to Isolate Bugs in Java Compilers that allow Malicious Code

Objective:

To analyze the lack of a Java byte-code verifier and its effect on possible 
bugs in compiling software that allow compilers to run malicious code.

Justification:

Malicious code has cost millions of dollars in damages, and nary a day goes by 
without Microsoft releasing software to patch its software for possible leaks. 
Cyber leaks can absolutely devastate systems, especially if run in the 
background, or some obscure loophole. Since TJHSST bases its Introduction to 
Computer Science course, as well as its AP Comp Sci course, in Java, hundreds 
of students use Java compilers everyday, with the possibility of malicious code 
slipping in. If TJHSST servers were to come down due to an attack such as this,
the loss would be unimaginable at this point.

Description:

Research will be done on the compilers themselves. Some background work has 
already been done, but the results are thus far inconclusive at best. By 
furthering this research,i.e., finding more bugs than the handful in the case 
study, I'll be able to test loopholes on an isolated machine to determine the 
possibility of malicious code worming its way into the TJHSST system.

Limitations:

Time, in that I don't know how much time I will be able to work on an isolated 
system. Also, research of the compilers could take additional time, and it 
could be difficult to gather data of the compilers currently used by TJHSST. 
Other than these two factors, I see no other limitations.