Re: [Tnfox-discussion] tests don't compile

[Niall D. <s_s...@ne...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 23 May 2004 at 19:26, Taras wrote:

> >I would have done if FXACLEntity::isLoginPassword() were essential.
> >The fact is it isn't and most code will never use it.
> >  
> I keep meaning to ask. Do you plan to implement extended attributes
> too? They are kind of in the same realm as acls for files.

Absolutely. I can only test it right now on FreeBSD though as I think 
my Linux doesn't support them.

It'll be compile-time configurable only however.

> >When Tn is done it'll be more capable than CORBA with far fewer
> >restrictions. And much, much faster.
> >  
> Is there an up to date description of Tn? You seem to have many 
> interesting ideas.

Not really as it defies all my attempts to describe it in a way 
anyone else can understand :(.

> In any case the IPC stuff looks interesting to the fact that it's
> basicly plug & play. This way I could develop something in a
> disconnected manner(using sockets) & deploy using shared mem. In
> general this kind of thing just screams "good idea".

You can only use synchronous devices (subclasses of FXIODeviceS) for 
IPC channels. I did experiment with operating a FXLocalPipe from 
within a shared memory patch and interestingly the kernel beat me 
every time. Kernel pipes are fast except on WinXP and later Windows.

Cheers,
Niall





-----BEGIN PGP SIGNATURE-----
Version: idw's PGP-Frontend 4.9.6.1 / 9-2003 + PGP 8.0.2

iQA/AwUBQLI3+cEcvDLFGKbPEQLbnQCgz21BEYF3TiArPxcAvS6+gUW5Q6cAoJPF
4zj3Oz70ULTNpoXNXbT1cx49
=mUMp
-----END PGP SIGNATURE-----