
|
[openbeosnetteam]
||
[Date Prev]
[01-2002 Date Index]
[Date Next]
||
[Thread Prev]
[01-2002 Thread Index]
[Thread Next]
[openbeosnetteam] Re: Hi all
- From: "David Reid" <dreid@xxxxxxxxxxxx>
- To: <openbeosnetteam@xxxxxxxxxxxxx>
- Date: Wed, 16 Jan 2002 22:59:11 -0000
> Quoth Zenja Solaja:
> >
> > Now for the rest of the NetTeam, I know I shouldn't be saying this,
> >since its pretty sneaky and has a "lurk in the shadows" kind of feel, but
> >have any of you checked out the recent changelogs to NewOS?
> >(http://www.newos.org/cgi-bin/perfbrowse.perl?@changes+//depot/...)
Travis
> >is implementing a kernel level TCP/IP stack, so that kind of diminishes
the
> >urgency of what we need to do right now, since the OpenBeOS kernel will
> >undoubtably have its own network stack, which is bound to be faster than
> >anything we can accomplish in user land. We'll also be given fd and
Posix
> >on a silver plate, so to speak. In order to implement the net_server,
all
> >we really need to do is provide a wrapper (the BNet*** API's) to assure
> >source code compatibility, and an entire set of apps/daemons (FTP
> >client/server, Telnet client/server etc). As I've already mentioned, it
> >does feel kind of dirty to add window dressing to work done by 3rd
parties
> >(there is nothing stopping you from assisting the Kernel team with the
> >TCP/IP stack).
>
> I've just talked to him briefly on IRC, and he doesn't seem to have a
specific
> timeline as far as his own network stack. He says newos is pretty close
to ready
> to handle a kernel or userland stack, having ports and one network card
driver
> there, if we want to start throwing it together. (I've also got
documentation
> on three cards that I own, so it shouldn't be too tough to start getting a
bunch
> of hardware support in there..)
OK, well does he have a roadmap? I'm very cautious about expending effort
only for him to blow it out in favour of his own ideas...
david
|

|