harmony-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Mikhail Fursov" <mike.fur...@gmail.com>
Subject Re: [classlib][nio]the conflict of RI and spec?
Date Thu, 20 Jul 2006 08:15:01 GMT
On 7/19/06, Paulex Yang <paulex.yang@gmail.com> wrote:
> I tried the test, and it failed as you said with RI on my WinXP. And I
> also tried to pass in a read/write FileChannel got from
> RandomAccessFile, the test passes. I consider this is RI's bug, because
> no way to have such different behavior in this clear test. So I think we
> should follow spec.

I agree that this is RI bug.
RandomAccessFile("rw") works OK because the channel is writable and can be
extended by map() method. FileInputStream or RandomAccessFile("r") returns
redable channel and we get 'Access Is Denied" when map() tries to change
its  size.
The description of such a process could be found here:
http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6416884  , see item 2 in

+ I examined Sun's bug database I have not found any bug like this reported
+ Note, that this test is failed even with writable source channel: the
'result' is 2*length

Mikhail Fursov

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message