[openbeosnetteam] Re: [Fwd: New Networking Project...]

  • From: "David Enderson" <DEnderson@xxxxxxxxxxxxxx>
  • To: <openbeosnetteam@xxxxxxxxxxxxx>
  • Date: Wed, 15 Dec 2004 16:42:29 -0600

Philippe,

Where does this leave my DHCP implementation?  I have also been pulled
away by work and personal, and have been slating a large block of
vacation around Christmas and New Year's time to finish the stand-alone
executable.

If the FreeBSD 5.x net stack was ported, would that include a DHCP
client?  Would this "marrow" project include a DHCP client piggy-backing
in with the FreeBSD net stack?

Should I continue working on the DHCP client?  I'd love to, I would just
prefer the time not be wasted.  :-)

On a personal note, good luck finding a job.  I have no idea what the
job market is like in your part of France for your skills, but I wish
you the best of luck.

--David


> -----Original Message-----
> From: openbeosnetteam-bounce@xxxxxxxxxxxxx 
> [mailto:openbeosnetteam-bounce@xxxxxxxxxxxxx] On Behalf Of 
> Philippe Houdoin
> Sent: Wednesday, December 15, 2004 16:20
> To: openbeosnetteam@xxxxxxxxxxxxx
> Subject: [SPAM%: 45%] [openbeosnetteam] Re: [Fwd: New 
> Networking Project...]
> 
> 
> Hi guys...
> 
> > Waldemar Kornewald <Waldemar.Kornewald@xxxxxx> wrote:
> > > the FreeBSD team started a new project to clean up their netstack
> > > and 
> > > make it ready for the future. Thought this is of interest for us.
> > 
> > Have you guys settled down the decision wether or not to port the
> > latest FreeBSD stack?
> 
> I'm the one to blame here for not having reported about this 
> project. Due to my personnal issues (finding a job *now*!), 
> I've abruptly lost 
> free time I used to spent on network team.
> Even more since last couple of months.
> 
> Anyway, we did talk Waldemar and I about this possible road.
> In particular, we agreed on a possible port of the FreeBSD 5.x net 
> stack, making it an huge stand-alone module.
> This big module, in that scenario, will replace the current core, 
> protocols and some of the interfaces stack modules.
> Yep, that's dropping modular design, mostly.
> 
> The pros were better, more stable, up-to-date, easiest to maintain 
> stack code. The 5.x net stack branch is known to be thread-aware on 
> most parts. Beside a really great and long features list, a 
> way better 
> BSD "exotic" network API's support is cool too.
> 
> The cons were the code will still be, well, not really easy to 
> understand (still many macros and low C *tricks* in the latest code) 
> and to evolve from, and that moving to a fully modular design again 
> from that point will most probably be very hard, dixit FreeBSD's net 
> guys.
> 
> But I fear that nor me nor him want or can do it (time issue) was the 
> biggest negative point.
> We may find more interrest in David's "Marrow" project 
> instead, as it's 
> sound more like "BSD stack meet BONE".
> Waldemar discussed with him, but since I dunno how it came.
> 
> > And if you're positive, any work done in that direction yet?
> 
> Nope. Not on my side.
> I've a huge personal todo-list for this team :-(
>  
> - Philippe Houdoin
> 
> --
> Fortune Cookie Says:
> 
> God must love the Common Man; He made so many of them.
> 
> 
> 
> 
> 

Other related posts: