That said, I think a signal to hold on this vote is that a vote
happened on [email protected], resulting in continuing the
podling.

This should work like a roll call to a TLP and indicate that at least
3 PPMC members (mentor included?) are actively watching the podling.

Best,
tison.

Jarek Potiuk <[email protected]> 于2026年11月11日周三 11:18写道:
>
> Everyone and Dave,
>
> +1 to retire.
>
> Since Dave admitted he did not use the skill from Incubator, I performed
> the analysis.
>
> This is the time and token cost of running the report:
>
> --------------------
>
> With caching, the whole Pegasus analysis probably cost about **$1.50–$3**
> at Claude Opus 5.5 API prices, and took an estimated **10–20 minutes** of
> working time. Both are estimates: I can't see the session's real token
> counts or timing, so `/cost` (or `/usage`) is the place for exact numbers.
>
> **Claude Opus 5.5 API prices, per million tokens**
>
> | Type                         | Price                     |
> | ---------------------------- | ------------------------- |
> | Input (uncached)             | $4.00                     |
> | Output                       | $20.00                    |
> | Cache read                   | $0.20                     |
> | Cache write (5-minute cache) | about $5.00 (1.25× input) |
> | Cache write (1-hour cache)   | about $8.00 (2× input)    |
>
> The cache-write prices are worked out from the input price and the usual
> 1.25×/2× multipliers, not quoted directly.
>
> **Cost estimate**
>
> | Part                                                       | Tokens
> (approx.) | Cost           |
> | ---------------------------------------------------------- |
> ---------------- | -------------- |
> | Cache reads (each step re-reading the conversation so far) | ~1.8M
>      | ~$0.36         |
> | Cache writes (new content added at each step)              | ~150k
>      | ~$0.75–$1.20   |
> | Output (including reasoning)                               | ~15–20k
>      | ~$0.30–$0.40   |
> | **Total**                                                  |
>      | **≈ $1.50–$3** |
>
> If nothing had been cached, all ~2M input tokens would be billed at $4/M,
> which comes to about **$8–$10** in total.
>
> **Time:** 10–20 minutes, not counting the private@ login or your replies.
> Over half of that went on the private@ pass, because I had to query each
> month separately; with a working date-range search, that phase would take a
> few minutes.
>
> ----------------
>
> Piunloc report (note that a more complete, private report contains much
> more detail)
> more arguments will be shared in the members@thread where we are
> discussing the viability of using tools for incubation process, but the
> SKILL
> warned me that I should not share it here because general@ is public list.
>
> ----------------
>
> Some background from the public record.
>
> Clutch suggests Pegasus looked ready around 2024: a release in Dec 2023,
> recent committer additions, and a mentor writing in the Oct 2023 report
> "We should start to prepare graduation now". The podling never acted on
> it. None of its 2023-2024 reports ticked "Nearing graduation", and each
> listed releases and performance work as unfinished issues. The 2.6.0
> release announced in Apr and Jul 2024 never shipped.
>
> Since then there has been no release in almost three years, and dev@
> has had little beyond automated notifications. The October report was
> missed, and the IPMC's questions went unanswered until this vote was
> called.
>
> Yingchun's dev@ thread [1] is welcome, and the community can carry on
> outside the ASF if this vote passes. If it does not pass, I'd want to
> see by the next report:
> - a release vote on dev@
> - at least one new committer or PPMC member
> - a report filed on time with mentor sign-off
> - named PPMC members who will answer IPMC mail
>
> A note on how this was put together: the analysis was prepared with an
> AI assistant using the Incubator graduation-self-assessment skill and
> the Incubator data tools (podling reports, list archives, release vote
> history). There is a longer "private" version that also draws on
> private@pegasus and contains considerably more detail. The assistant
> flagged that it should not be posted to general@ and proposed this
> short, public-only version instead. IPMC members who want the full
> version can ask me off-list.
>
> [1] https://lists.apache.org/thread/to1rmjk0wn49kw4wbjkgfngy21xq2lrx
>
> Jarek
>
> On Tue, Nov 10, 2026 at 5:04 PM Andrew Ahn <[email protected]> wrote:
>
> > +1
> >
> > On Sun, Nov 8, 2026 at 3:09 PM Justin Mclean <[email protected]>
> > wrote:
> >
> > > Hi,
> > >
> > > I’d like to call a vote to retire Pegasus from the Incubator.
> > >
> > > Pegasus has been incubating since June 2020, and its last release was in
> > > December 2023. I asked the PPMC and mentors on 2 August whether a
> > community
> > > is still working toward graduation, raised retirement here on 14 August,
> > > and asked again on the private list on 29 September, saying I would
> > start a
> > > retirement vote if there was no response by the October report.
> > >
> > > None of these has had a reply from the PPMC or the mentors. Pegasus did
> > > not submit its October report, and its dev list has had one automated
> > > GitHub notification a month since September.
> > >
> > > Retirement is a normal outcome and nothing to be embarrassed about. If
> > the
> > > community wants to continue, it can keep developing the code outside the
> > > ASF, for example on GitHub, under a different name.
> > >
> > > Please vote:
> > > [ ] +1 Retire Pegasus
> > > [ ] 0 No opinion
> > > [ ] -1 Do not retire Pegasus, because…
> > >
> > > The vote will be open for at least 72 hours.
> > >
> > > Kind regards,
> > > Justin
> > > ---------------------------------------------------------------------
> > > To unsubscribe, e-mail: [email protected]
> > > For additional commands, e-mail: [email protected]
> > >
> > >
> >
> > --
> > - AA
> >

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to