Vote on Treasury Withdrawal - IO - Maintenance Initiative

I am voting NO on the governance action with hash 73e171a4c0730b4b59ecae271ab89f12a9d56360b02920e1f95107dbdc1d6762#3.

The metadata for my vote is located here, and this blog serves as a more human-readable form of that justification.

Justification

Maintaining the Haskell node and related infrastructure are essential. However, I can’t, in good conscience, vote yes on this proposal.

• The funds asked for cover 9 months, representing a cost-to-employ of ~435k per year
• This allocates 36 employees at the above 435k per year, simply to keeping the lights on
• The proposal constitutes a laundry list of things that might need maintenance, or did need maintenance in the past, rather than a concrete action plan for what will be delivered. Many of them are either already complete (CC identity and guard rails scripst), not a good use of resources (db synthesizer, third part libsodium fork), or covered by other propsoals or teams (CDDL updates for Leios would be done by the 20+ people assigned to Leios; Cardano Blueprint has major contributions from other teams; etc.)
• I have concerns about the “Ensurable Systems” partnership; their website only lists 3 former IO employees who, while talented in their past roles, are a new company. No detail is given on how large they are; The proposal doesn’t specify how funds are split between IO and this new company with no proven independent track record; and the website is thrown together, unprofessional, and has very little details.

Ultimately, there are certain core functionalities that are non-negotiable:

• maintaining benchmarking clusters
• release signoffs
• community SPO calls
• paying down technical debt

But these represent 5-8 full time employees at most.

For reference, this proposal represents nearly 8 years of runway for Sundae Labs as a whole, a 20 person company that is consistently delivering on half a dozen critical projects.

I support IO’s role in the ecosystem. They’ve pushed through some extremely hard barriers, and built a foundation upon which we all rely. Maintaining that foundation is essential, and the skills to do so not easily fungible to some other company (certainly I’m not suggesting Sundae Labs could pick it up).

But this proposal, as is, is simply grossly mis-sized, or assumes many of the core maintenance staff will form the bulk of IOs other proposals, in which case it is double-counting the cost to employ.