Re: [dpdk-web] [RFC PATCH] process: new library approval in principle

[Thomas Monjalon <thomas@monjalon.net>]

17/04/2023 15:33, Jerin Jacob:
> On Wed, Mar 15, 2023 at 7:17 PM Jerin Jacob <jerinjacobk@gmail.com> wrote:
> > On Fri, Mar 3, 2023 at 11:55 PM Thomas Monjalon <thomas@monjalon.net> wrote:
> @Thomas Monjalon  Could you check the below comments and share your
> opinion to make forward progress.
> 
> > > 13/02/2023 10:26, jerinj@marvell.com:
> > > > --- /dev/null
> > > > +++ b/content/process/_index.md
> > >
> > > First question: is the website the best place for this process?
> > >
> > > Inside the code guides, we have a contributing section,
> > > but I'm not sure it is a good fit for the decision process.
> > >
> > > In the website, you are creating a new page "process".
> > > Is it what we want?
> > > What about making it a sub-page of "Technical Board"?
> >
> > Since it is a process, I thought of keeping "process" page.
> > No specific opinion on where to add it.
> > If not other objections, Then I can add at
> > doc/guides/contributing/new_library_policy.rst in DPDK repo.
> > Let me know if you think better name or better place to keep the file

Maybe that the contributing guide is the best place.
I'm OK with a new file doc/guides/contributing/new_library.rst
which could document more than the policy in future
(like things to remember and to check).

> > > > +Adding a new library to DPDK codebase with proper RFC and then full patch-sets is
> > > > +significant work and getting early approval-in-principle that a library help DPDK contributors
> > > > +avoid wasted effort if it is not suitable for various reasons.
> > >
> > > That's a long sentence we could split.
> >
> > OK Changing as:
> >
> > Adding a new library to DPDK codebase with proper RFC and full
> > patch-sets is significant work.
> >
> > Getting early approval-in-principle that a library can help DPDK
> > contributors avoid wasted effort
> > if it is not suitable for various reasons

It will be easier if starting with the goal:
In order to save effort, developers will get an early approval in principle,
or early feedback in case the library is not suitable for various reasons.

> >
> >
> > > > +   - Purpose of the library.
> > > > +   - Scope of the library.
> > >
> > > Not sure I understand the difference between Purpose and Scope.
> >
> > Purpose → The need for the library
> > Scope → I meant the work scope associated with it.
> >
> > I will change "Scope of the library" to,
> >
> > - Scope of work: Outline the various additional tasks planned for this
> > library, such as developing new test applications, adding new drivers,
> > and updating existing applications.

OK

> > > > +   - Public API specification header file as RFC
> > > > +       - Optional and good to have.
> > >
> > > You mean providing API is optional at this stage?
> >
> > Yes. I think, TB can request if more clarity is needed as mentioned below.
> > "TB may additionally request this collateral if needed to get more
> > clarity on scope and purpose"

OK

> > > > +3. Based on mailing list and TB meeting discussions, TB to vote for approval-in-principle and share
> > > > +the decision in the mailing list.
> > >
> > > I think we should say here that it is safe to start working
> > > on the implementation after this step,
> > > but the patches will need to match usual quality criterias
> > > to be effectively accepted.
> >
> > OK.
> >
> > I will add the following,
> >
> > 4.  Once TB approves the library in principle, it is safe to start
> > working on its implementation.
> > However, the patches will need to meet the usual quality criteria in
> > order to be effectively accepted.

OK