Re: [MiNT] [CVS] Xaaes in freemint module

[Johan Klockars <johan@klockars.net>]

> > > No, it depend at the moment on some kernel source. It's also planned from
> > > me to integrate XaAES much better with FreeMiNT. So it's part of the
> >
> > IMO, that is a really bad idea. But that might just be me.
> 
> Can you also explain why?

I'm generally against integration, and integrating the kernel with
what is really an application (albeit a somewhat special one, with
kernel type entry calls (for now)) seems even less appropriate.

Perhaps more importantly, XaAES is not necessarily tied to MiNT.
Sure, it relies on MiNT right now, but in the future Fenix might
become viable, or someone might want to make XaAES run native on some
emulator, or do a variant that does not require multi-tasking, or...

> I only see lot of ugly hacks in the kernel for the AESSYS to get it
> working properly.

Are you talking about current hacks?
Surely those must remain in place to allow N.AES and MultiTOS to work?

If you're talking about upcoming hacks, it's been a long time since
I last tested XaAES, unfortunately, so I don't really know what
problems it currently has that would require new kernel hacks.
Can't say I've heard about any complaints at that level, though.

> > So, you're prepared to give a bunch of new people write access to
> > the whole FreeMiNT repository then?
> 
> Until now every developer had write acces to all modules in the
> repository. But only a few are really active and do something. Do you

Well, kernel level hacking is a long way from the AES.
There are a whole lot more people who know the AES and the VDI (and
who might be interested in doing something about XaAES) than there
are people who know device drivers and memory handling, for instance.

> think it's in our small world so important to restrict the module access

Perhaps not, but I'd never integrate unless there were extremely
compelling reasons to do so.

-- 
                        | Why are these |  e-mail:   johan@klockars.net
                        |  .signatures  |
                        | so hard to do |  WWW:      http://www.klockars.net
                        |     well?     |            (fVDI, MGIFv5, QLem)