[lwip-users] Re: Re: [lwip] The future of lwIP

[David Figueroa]

                                I also want to put my vote for a very small
group with write access.  Let Adam select from people he trusts.  This is a
TREMENDOUS responsability, not to be taken lightly.

                                I also vote to keep the lwip development
focused on the "light weight" premise on which it was created.  Maybe the
gig eth should be a different cvs module that uses parts of the core?

                                I agree with Paul, that at this point lwip
is not robust enough for serious use.  I am know having problems with
multiple connections that I am trying to isolate.

                                Thanks all for your hard work.

                                dave


                        

                                ----- Original Message -----
                                Hello,

                                ----- Original Message -----
                                From: "Duncan Palmer"
<address@hidden>
                                > access. These guys should have worked on a
number of different parts of
                                the
                                > stack (not just writing drivers for
example). Everyone else will have to
                                > submit their changes to the list as per
usual for peer review, and one of
                                the
                                > 2 or 3 people with write access will check
in the changes if they, and the
                                > list think they're ok. It is very
important that these people with write
                                > access know enough about most of the stack
to be able to effectively
                                evaluate
                                > all changes!
                                >
                                This seems like a good idea. Use the mailing
list to issue problems and
                                features,
                                and then after much reviewing forward the
changes into the current tree.

                                > Based on what I've seen on the list, my
nominees for people with write
                                access
                                > would be Leon, Kieran and Paul.
                                >
                                Thanks for your trust. I am not familiar
with the TCP part of the stack
                                though;
                                the rest has become rather familiar with me.

                                I think Paul has a major point in the
complexness of the dynamic memory
                                management
                                issues, especially regarding the TCP part of
the code, and the resulting
                                flakiness.

                                However, I think this is the best available
open-source TCP/IP project I
                                have seen, so
                                it's definately worth every effort to make
it rock stable.

                                Leon.


[This message was sent through the lwip discussion list.]