Re: [MiNT] Network and USB problems with NetUSBee

[Paul Wratt <paul.wratt@gmail.com>]

On Fri, Aug 5, 2011 at 11:15 AM, Thomas Jürges <thomas@senmut.net> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> On 2011-08-04 14:03, Paul Wratt wrote:
>> if this is the case, a small workaround prg, that does enough to
>> slow down packet management (and user adjustable), would allow people
>> to use >=1.17.0 (maybe supplied with the patched inet.xdd)
>
> You mean something like this which uses up all of the available cpu time
> - - All of it?  :-)
>
> .. code ..
>
> Cheers,
>        T.
not all of it, and preferably user adjustable, but a percentage yes,
even a great percentage, something you could pair up with the cpu &
download (bug) speed of your particular machine. remember it is only
to make downloading stable on certain systems

But I was also thinking of some that physically takes longer, as the
curl dump described by mono/etc (although in that case I think it is
the quantity of data being dumped, not just the disk access). I think
floppy access would be the absolute slowest (without disk).

if you handle your downloads with curl/wget things are a lot simpler,
as you could script something else in the background only while
downloads are happening)

I think some people need to test my original theory, because atm thats
all it is, just a theory to provide a workable solution

but thanks for the start point :)

(knowledge of context switching may also help, ie vdi calls, maybe
that is why aftp often survives longer, because programatically it is
slowing down the speed at which data packets are managed - its a guess
anyway)

Paul