[lwip-users] Re: [lwip] A Modest Proposal

[David Haas]

On this thread I have a question. I am starting to port lwip to my coldfire
platform. As I start to compile these files I am finding a number of
compiler warnings (I always keep the compiler's lint option turned on).
Should I submit these changes back to the repository? I did an anonymous
checkout of everything about 2pm EST today.

While I am not new to networking development, I am new to distributed
open-source development. What is the protocol for submitting changes?

Thanks,
David Haas


----- Original Message -----
From: "Leon Woestenberg" <address@hidden>
To: <address@hidden>
Sent: Monday, November 11, 2002 3:52 PM
Subject: Re: [lwip] A Modest Proposal


> Hello all,
>
> > Yeah, but would the developers stay involved in the implementors list?
> > A bunch of newbies posting questions with no one knowledgable to answer
> > will not help the proliferation of this fine effort.
> >
> Any sane implementor will have to understand the current status of the
> TCP/IP
> stack and in some way has to follow recent developments and at least
> understand the .consequences.
>
> However, I agree that the more open structure of the current Savannah CVS
> tree
> has spawned some IRC-like discussions on long-standing issues, and this
> causes
> too much traffic for the lwIP generic list.
>
> I can set up address@hidden as a mailing list.
>
> > extensive thread about memory leaks to decide if this is an indication
> > that the release version is unstable, or is this just a discussion of an
> >
> As far as I can tell, the stability of the stack has increased in time.
>
> The exception to this, is once queueing was introduced in etharp.c. In
cases
> where chained pbufs where used, especially of mixed type, memory leaks
> where introduced.
>
> > So, I'm in favor of your suggestion, but I hope the core developers stay
> > in touch with the implementors.
> >
> But also, the other way around. Each implementor submits his finding back
> into lwIP (such as device drivers, architecture adaptions/ports).
>
> Speaking for myself, I never made the distinction between (core) developer
> and implementor; I chose the stack because I though the architecture and
> implementation was quite nice, and thought I can expand on it in ways our
> company uses it, which is what I do now.
>
> What I liked most was that people using my code submitted back problem
> feedback, patches, ideas and generally, contributed to a more stable stack
> for me to use. Basically, the benefits of open-source and the BSD license.
>
> OK, anyone against address@hidden Anyone in favor?
>
> Regards,
>
> Leon.
>
> > Dick Wilder.
> > -----Original Message-----
> > From: address@hidden [mailto:address@hidden On Behalf Of Bill
> > Knight
> > Sent: Monday, November 11, 2002 12:00 PM
> > To: address@hidden
> > Subject: [lwip] A Modest Proposal
> >
> > There has been a lot of traffic on the list concerning how to use
> > the current version of lwIP as well as discussions concerning new
> > features, bug fixes, etc.  While I am interested in both, it is
> > sometimes a bit overwhelming.
> >
> > May I propose the creation of another list "lwip-develop", to be used
> > for discussions concerning new features, bug fixes, and such?  The
> > current list might then be more helpful for those just trying to get
> > lwIP operational on their platform.
> >
> > Respectfully
> > -Bill Knight
> > R O SoftWare
> >
> >
> >
> >
> > [This message was sent through the lwip discussion list.]
> >
>
> [This message was sent through the lwip discussion list.]

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