Daily Reflection
Friday, October 16, 2026
12
Decisions
73%
Avg Confidence
high
Avg Importance
Summary
Two-week window, Friday evening 10/2 to Friday 10/16, dominated by the Core42 deployment in Abu Dhabi and by travel (Charleston, then London). Twelve decisions logged, all reviewed one at a time. Three set how CIQ runs Core42: it is an Engineering engagement with Ryan on point and senior engineering on the calls; all planning happens in CIQ-only rooms with zero friction toward Core42 and Adam as single coordinator; and Wolfgang on-site role ended once the VPN existed, with nobody sent in his place and Peter as backstop. From the 10/13 Nathan 1:1: enterprise GNOME and an opinionated AI desktop ship together rather than as separate SKUs, with AI non-negotiable; Nathan management layer moves now, with Peter coaching a promotable candidate weekly; and blockers get raised at department heads. Also: a research function exists only with Greg running it, and a Qualcomm prototype must not lock Engineering architecture. Coverage gaps: the gather calendar pull stopped at 10/13 03:00 so 10/13-10/15 Fathom meetings were recovered by hand, and two 10/15 Fathom meetings 429d in the gather and were recovered by hand.
Wins
VPN came up and Wolfgang flew home on 10/15 with the test-cluster path intact. Core42 relationship described by Peter as very good, able to get them to do anything needed. RLC Pro 10 CLK 6.18 shipped 10/15. Atlas is running as an internal service with a status page.
Challenges
Core42 coordination misfires on 10/7-10/8 (people missing calls, planning in a shared customer room) required resetting the rules mid-deployment. Ryan-Bjorn friction over overnight ownership. Nathan is overloaded with no management layer yet. The TPS overdue list keeps aging without new dates.
Learnings
Peter treats on-site presence as insurance against a named gap and ends it when the gap closes. With a partner, CIQ leverage is the relationship itself, so he forbids anything that looks like managing them. He will spend his own time to unblock a leader rather than let a structural wait drag on. Functions are justified by what they feed, not by existing.
What I Learned About Your Decision-Making
Peter classifies an engagement by who must be able to make calls in it, not by who runs it day-to-day - Ryan keeps point while it stays an engineering engagement. With a customer partner he optimizes for zero friction because the relationship is CIQ actual leverage, and he keeps CIQ internal churn in CIQ-only rooms. He bounds on-site presence to the specific gap it insures and offers himself as the backstop rather than spend someone else travel. When a structural plan stalls a leader, he converts the wait into his own weekly time investment. He refuses to let an upstream partner policy or a days-old prototype set CIQ product direction. He judges whether a function should exist by what it feeds, and a management seat by whether the person accepts a defined scope.
Team Status
View TPS ReportNew since the 10/12 report: RLC Pro 10 CLK 6.18 shipped on 10/15 with Secure Boot and production repos, one day past its 10/14 target; 10.2 FIPS has moved to "submitted to atsec testing" (still 10/30, 85%); and Steve's top priorities now list SOC 2 Type II (10/30, 100%) where ISO 42001 used to be, with Mirror Manager reliability up from 70% to 75%. Three items are due today: the Veeam Errata API (in acceptance testing, 77%), Pro H packages on 10.2 (84%), and retiring Jenkins from supported RLC image builds, which at 55% is the one most likely to slip. The overdue list has grown and none of it has moved: RLC Pro LTS 10.2 GA is now 7 days late, both kernel CVE batches 7 days, the Koji dev environment 2 days, and the four Fuzzball items in Chris's area (DGX Spark, v4.4 in-workflow image build, NVLink/NCCL, AMD evaluation) plus Warewulf Pro RLC Pro images are 9 to 28 days past target with no new dates or confidence changes. Worth asking Chris and Nathan for honest new dates rather than letting them keep aging; Core42 is still on track for 10/31 at 90%.
Tomorrow's Focus
Derived, not stated by Peter: propose a Wolfgang bonus amount to Bjorn and Greg; get Nathan to name a six-month promotable candidate and start weekly coaching; confirm the open desktop reqs are posted; ask Wolford and Nathan for honest new dates on the overdue TPS items.
Decisions Made
Core42 is a CIQ Engineering engagement, not Support - Ryan on point, Peter, Wolford and Nathan stay on the calls
strategy · high
Core42 partner posture - plan only in CIQ-only rooms, never program-manage Core42, zero friction, Adam coordinates
operational · high
Wolfgang on-site role ends at the pre-VPN window - no extra travel, send someone only if a problem appears, Peter is the backstop, bonuses earned
operational · medium
Desktop - enterprise GNOME and an opinionated AI desktop ship together, not as separate SKUs; AI is non-negotiable; open 1-2 desktop reqs now
strategy · high
Nathan management layer - stop waiting on promote-from-within; Peter coaches a six-month candidate weekly; Nathan moves out of frontline management
people · high
Greg research team - no management role for someone who will not take a defined role; a research function exists only with Greg running it
strategy · medium
Sensitive Decision
Sensitive Decision
Sensitive Decision
Qualcomm prototype - present it without letting a days-old prototype lock Engineering architecture
technical · medium
Blockers get raised at department heads so the whole company hears them
operational · medium
Sensitive Decision
Reflection ID: f0ea2001-13a5-4c87-bbd2-1963d8ce2c34