Re: [irsg] An IETF repository for working code in our protocols?

[Vijay Gurbani <[e-mail]>]

On Wed, Aug 19, 2020 at 6:42 PM Joel M. Halpern <[e-mail]> wrote:

> For IETF purposes, the existence of implementations is very important.
> For some purposes, open source code is useful.
> Depending upon the license, for many purposes looking at open source
> code can actually be harmful to the implementation work. [...]
>

Dear Joel: Thank you for your comments, much appreciated.

I will note two things that are important to keep in mind as we develop our
protocols.

One, you do not need to look very far to ascertain that there is a dearth
of reproducibility in the sciences today, be it in academia or industrial
research.  This has, in part, lead to the stance of IEEE and ACM that I
pointed to in my original email where they have made repositories available
to researchers for storing datasets and software artifcacts.  Making the
source code available mitigates the reproducibility problem, not in whole,
but in substantial part.

Two, for an organization that prides itself on open and transparent
development of protocols, it seems rather incongruous that a key component
of a protocol --- the source code that implements a protocol --- is kept
secret.  Companies and individuals that do not want to share code are free
not to do so.  However, that need should not impinge upon the need for
others who can benefit from sharing for the purpose of reproducibility.
Individuals who would rather not deal with compilation and other problems
associated with open source are just as welcome not to worry about it, and
let others who revel in it move the implementation and the protocol ahead.
In most of the cases, the source code is not going to be a production ready
version, so I really see no harm in making these available on a "best
effort" scale.

In closing, I don't think that expounding the virtues or vices of open
source versus closed source is necessarily the issue here.  All of us have
strong viewpoints on this, and for the record, I am more in alignment with
Stephen in the need for the code to be available.  It can be made so under
some license just as we handle IPR issues.  Bringing in open versus closed
source code detracts from the larger aim of the discussion here, with all
due respects.

Thank you.

- vijay