ckcheng

I feel the analogy actually holds up well since brands often create retailer exclusive versions of their products specifically for Costco or Walmart, etc.

Like you can buy Henckels knives anywhere, but you can only buy Henckels item 1374080 at Costco [1] - that Costco item number is printed right on the box.

I sympathize with the rent-seeking walled garden argument against Apple. But would it be ok for Apple to require iOS IAP if their cut was not 30% but was instead 25% (like Walmart’s gross margin), or was instead 14% (like Costco’s)? c.f. Apple’s App Store Small Business Program rate is 15%.

Is the percent cut that’s the problem regardless of structure, or is it the structure that’s bad?

[1] https://www.costco.com/henckels-modernist-20-piece-self-shar...

I’m trying to understand the structure, is this about right?

Costco Warehouse = Apple App Store.

Dyson (vacuum vendor) = WidgetApp (digital goods vendor).

Dyson sold in the Costco Warehouse must use Costco as the merchant. WidgetApp sold in the Apple App Store must use iOS IAP merchant of record solution (??).

Costco Warehouse forces use of Visa (or Mastercard - depends on country). Apple iOS IAP does NOT require Apple Pay.

Of course Apple doesn’t buy goods from suppliers before selling them to customers, unlike Costco. And Costco Warehouse is members only, unlike Apple (unless you squint and view iPhone ownership as a membership).

So you’re saying Apple IAP as merchant of record forcing use of Apple Pay is ok (?) but Apple App Store forcing use of Apple IAP as the only merchant of record is not?

(You were clarifying the difference between these things elsewhere so hoping you could clarify this, thanks.)

(And I want to see more Costco v Apple comparisons!)

VLA = Vision-language-action model: https://en.wikipedia.org/wiki/Vision-language-action_model

Not https://public.nrao.edu/telescopes/VLA/ :(

For completeness, MMLLM = Multimodal Large language model.

> After eight years, the people in the structured exercise program not only became more active than those in the control group but also had 28% fewer cancers and 37% fewer deaths from any cause. There were more muscle strains and other similar problems in the exercise group.

https://apnews.com/article/colon-cancer-exercise-study-recov...

It just occurred to me that without the “before” version, one wouldn’t be able to do a compare and contrast of the revision proposed above. So here’s the original, similarly a little anonymized.

***

Sam Hayden Somewhere, USA

sam.example.dev | sam@example.com

linkedin.com/in/samexample | github.com/samexample

Professional Summary

• 20+ years full-stack engineering experience specializing in data architecture, JScript+ & ZetaScript, immersive tech

• Engineering team leadership experience

• Actively experimenting with ML-generating full codebases and intelligent tooling

• Strong focus on project methodology, automation, and comprehensive documentation

• Bachelor of Science in Computer & Information Science, State University of the Northwest

Goals

I am looking to be part of a small-to-midsized engineering team where I can have real impact by solving hard and previously unsolvable problems by leveraging machine intelligence tools. I am interested in growing my dev leadership experience. I am a motivated and self-directed learner, capable of deep diving into learning whatever new tools, languages, or frameworks I need to, on the job. I am seeking to leverage my problem solving skills and creative thinking rather than be a mere production-line engineer. I have a special interest (but not limited to) cutting-edge applied ML and XR, with experience in immersive environments since 2017.

Skills [n.b. was in 2 columns]

• 20+ years JScript+

• full stack ZetaScript

• Atlas Cloud and Sparkbase

• Glance & GlanceUI

• API development

• ForgeStack

• Data architecture

• Relational & DocDBs

• CI/CD pipelines

• Documentation & Technical Writing

• User experience & Developer experience

• SDLC methodologies

• Creative problem solving

• The vibecoding ecosystem

• Clear and kind communicator

Experience (selected) [n.b. “•” missing in original]

Lead Full Stack Engineer – WorldSpaceXR.dev (2021 – 2023)

Browser-based immersive meetings and more. Built full stack subscriptions system resulting in tens of thousands MRR. Built an in-world AI-assistant NPC with VectorX RAG vector search and ChainWeave. Built and maintained Atlas Cloud infrastructure. Led DevX and documentation improvements. Had creative input across the full stack. Designed and implemented user experiences in Glance/GlanceUI.

Sr. Full Stack Engineer – LinkLayerNetworks.io (2018 – 2020)

LinkLayerNetworks is a managed SD-VPN systems provider. Built management dashboard software integrating with over a dozen telecom APIs (HorizonLink, AlphaComm, TriWave, OmniNet, MatrixComm, etc.) for clients including Beanbrew Co., FitNow Gyms, ForgeGear, SureTrust and more to manage their networking and telecom infrastructure.

Sr. Full Stack Engineer – sam.example.dev (2011 – 2018)

Independent consultant. Built an IVR phone screener system that the world’s largest biotech companies use to pre-screen and match thousands of patients a day to participating clinics for clinical trials around the world.

Developed form-building and e-signing app used to close commercial property title deals. Used by clients like Liberty Mutual Bank. Custom API integration with CRM platform. Handles tens of millions of dollars in daily transactions.

Produced and self-released 5-star rated XR app experience Metro Canyon Express on multiple platforms. Highly optimized to achieve long-running 60fps performance on low-end mobile immersive hardware.

Web Dev Instructor – LearnCode.io (2011 – 2012)

Taught about a dozen students JScript+ and web development fundamentals, one-on-one.

Backend Engineer – SignalCore Agency (2010 – 2011)

Backend engineer on the team that built the award-nominated official tourism site for a U.S. state.

Brought a leading sporting optics company into the digital age, converting their technical IP into usable databases powering their web apps.

Backend engineer on the team that created an internal employee metrics and shoe ordering app used nationwide in retail locations selling a top athletic brand. Represented the backend voice of my agency in-person at brand HQ meetings.

Backend Engineer – ModernHue Agency (2009 – 2010)

Built an internal data management tool for Ikea Athletics to track compounds used in their products and the corresponding environmental regulations.

Built an internal employee recognition and rewards app for pets.example.com.

Built a backend server job that synced with a call center CRM for a major home security installer to coordinate and schedule customer events around their system installs.

Full Stack Engineer – EchoPeak Agency (2007 – 2009)

Built an employee training and rewards app for Samsung.

Maintained web accounting software for a leading global electronics brand.

Immersed in full agency environment working across the creative stack with screenwriters, designers, and video producers.

I find resume advice very difficult to implement.

Maybe you made some drastic resume changes at the beginning of this HN discussion that I didn’t see, but since then you’ve only had time and energy to fix the repeated “maintained” word in …_Resume_2025-7 (current version is 8 now).

Your latest two posts in substack are interesting reads tho.

So I decided to spend some time to rewrite your resume. IMHO this is better, but there is still so much room for improvement even here.

I anonymized it somewhat as I don’t have your permission for this, but perhaps it’ll help you or someone else to see this revision. I’m not a pro resume writer, so maybe I’ll get some feedback here to learn from too.

Best wishes!

***

Note: links to your other online presence was intentionally deleted and omitted from the resume. See other’s comments linked above for more great suggestions.

***

Sam Hayden

Somewhere, USA / Sam@example.com

Senior full-stack engineer with 20+ years experience in ZetaScript, data architecture, machine intelligence tools, and immersive web. Builds production-grade systems, leads high-impact teams, and values clear and kind communication.

Objective

Seeking a role on a small to mid-sized engineering team solving complex problems with machine intelligence tools. Focused on growing technical leadership and contributing to cutting-edge work in applied ML and immersive tech.

Experience

Lead Full Stack Engineer – WorldSpaceXR.dev

2021 – 2023

• Built full stack subscriptions system resulting in tens of thousands MRR.

• Built an in-world AI-assistant NPC using VectorX RAG search and ChainWeave.

• Built and maintained Atlas Cloud infrastructure.

• Led DevX and documentation improvements.

• Designed user experiences in Glance/GlanceUI.

• Implemented front-end interfaces with responsive UI components.

Sr. Full Stack Engineer – LinkLayerNetworks.io

2018 – 2020

• Built management dashboard software for enterprise networking and SD-VPN systems.

• Integrated with over a dozen telecom APIs including HorizonLink, AlphaComm, TriWave, OmniNet, and MatrixComm.

• Delivered custom tooling for clients like Beanbrew Co., FitNow Gyms, ForgeGear, and SureTrust.

• Enabled client management of large-scale networking and telecom infrastructure.

Sr. Full Stack Engineer – Sam.example.com

2011 – 2018

• Built IVR phone screener system used by top biotech companies.

• Enabled daily pre-screening and matching of thousands of clinical trial patients worldwide.

• Developed a form-building and e-signing app to streamline commercial property title deals.

• Delivered custom CRM API integration for clients like Liberty Mutual Bank.

• Supported tens of millions of dollars in daily property transactions.

• Produced and self-released 5-star-rated VR app Metro Canyon Express.

• Optimized VR experience to run at 60fps on low-end mobile XR hardware.

• Released the VR app on multiple major platforms.

Full Stack Engineer / Backend Developer – Multiple Studios

2007 – 2011

Studios include: EchoPeak, ModernHue, SignalCore.

• Built internal tooling for Ikea athletic, Samsung, and pets.example.com.

• Created data infrastructure for state-level tourism and sports optics organizations.

• Represented backend team in client meetings at major brand HQs.

Education

B.Sc. in Computer & Information Science, State University of the Northwest

Skills

ZetaScript, JScript+, Glance, GlanceUI, Atlas Cloud, Sparkbase, ForgeStack, Relational DBs, DocDBs, CI/CD, API architecture, SDLC, DevX, XR, ML (ChainWeave, VectorX), UX, DX.

> Did the University of California take out E&O on the Manhattan Project?

Interestingly, see [1] for

“a teletype from General Groves to Oppenheimer from February 1944, instructing the latter as to what to tell Underhill [UC’s secretary and finance officer] about the hazards to be insured against at an unspecified site”

[1]: https://blog.nuclearsecrecy.com/2012/03/28/weekly-document-i...

There really is a few comments here with great suggestions for how to improve your resume and online presence. It seems like some of it can still be tried, or if you feel it was already tried but got rolled back, then they can be tried again altogether.

Unfortunately, there's also a lot of noise here too... As a start, I'd say these few comments have solid, specific, actionable suggestions:

- https://news.ycombinator.com/item?id=43978225

- https://news.ycombinator.com/item?id=43978405

- https://news.ycombinator.com/item?id=43981072

- https://news.ycombinator.com/item?id=43983905

But I also want to say... I sympathize and agree with you that it's tough now out there. There's ageism, offshoring, AI, end of ZIRP, economic uncertainties, etc.

I hope you get to a position you're happy with soon.

Oja's rule dates back to 1982?

It’s Hebbian and solves all stability problems.

https://en.wikipedia.org/wiki/Oja's_rule

Above was talking about embedded systems, and I really don’t know if the James Webb Space Telescope counts, but…

The James Webb Space Telescope runs JavaScript, apparently [1].

[1]: https://www.theverge.com/2022/8/18/23206110/james-webb-space...

A comma would've sufficed?

cf: "SpaceX reached out immediately to the government[s] of Turks and Caicos[,] and the United Kingdom to..."

> Any solution Apple develops that enables "disable E2E for this account" makes it harder for them to claim that implementing that would be compelling work (or speech, if you prefer)

I think it’s really speech [0], which is why it’s important to user privacy and security that Apple widely advertises their entire product line and business as valuing privacy. That way, it’s a higher bar for a court to cross, on balance, when weighing whether to compel speech/code (& signing) to break E2EE.

After all, if the CEO says privacy is unimportant [1], maybe compelling a code update to break E2EE is no big deal? (“The court is just asking you, Google, to say/code what you already believe”).

Whereas if the company says they value privacy, then does the opposite without so much as a fight and then the stock price drops, maybe that’d be securities fraud? [2]. And so maybe that’d be harder to compel.

[0]: https://news.ycombinator.com/item?id=43134235

[1]: https://www.eff.org/deeplinks/2009/12/google-ceo-eric-schmid...

[2]: https://www.bloomberg.com/opinion/articles/2019-06-26/everyt...

>> Apple's defense against backdooring E2E is the (US) doctrine that [government can’t] be compelling work (or speech, if you prefer)

It’s really not "work” but speech. That’s why telecoms can be compelled to wiretap. But code is speech [2], signing that code is also speech, and speech is constitutionally protected (US).

The tension is between the All Writs Act (requiring “third parties’ assistance to execute a prior order of the court”) and the First Amendment. [1]

So Apple may be compelled to produce the iCloud drives the data is stored on. But they can’t be made to write and sign code to run locally in your iPhone to decrypt that E2EE data (even though obviously they technologically could).

[1]: https://www.eff.org/deeplinks/2015/10/judge-doj-not-all-writ...

[2]: https://www.eff.org/deeplinks/2015/04/remembering-case-estab...

> to even have a rheumatologist referral to begin with, some previous combination of [doctors] would have had to have suspected an autoimmune possibility and created some documentation to that effect

This is a good point. Even when a diagnosis is obvious, a doctor may not say it outright and instead make a referral to a specialist because it’s not their specialty to diagnose that condition. This is a good thing because they’re not that specialist so they could be wrong about it, and if they are right, then they should make that referral anyway - and end result is the same i.e. a referral.

Also just 1 day ago: https://news.ycombinator.com/item?id=42587826

> seems like a signal

Also “mil-spec”, and in a totally different context i.e. big box retail, “contractor grade”.

https://en.m.wikipedia.org/wiki/Military_grade

This is referring to timescape cosmology and the recent paper (Supernovae Evidence for Foundational Change to Cosmological Models) that was discussed 11 days ago on HN:

https://news.ycombinator.com/item?id=42495703

I think they trained on one too many closet bifold doors [1].

If you look at the edge of the doors as they swing open, it seems their movement resembles bifold door movement (there's a wiggle to it common to bifold doors that normal doors never have). Plus they seem to magically reveal an inner fold that wasn't there before.

[1]: https://duckduckgo.com/?t=h_&q=interior+bifold+closet+doors&...

> Javascript's lack of tail recursion

Even in Safari? [1]

[1]: "As of July 22, 2023 Safari is the only browser that supports tail call optimization" https://stackoverflow.com/a/37224563

> then in Grade 12, there's Advanced (!) Functions."

Then in university, there’s Elementary Functional Analysis!

Tao comparing o1 to a mediocre, but not completely incompetent, graduate student went viral. He clarifies that he means it more like a mediocre or incompetent research assistant.

Tao means that it's not like a grad student who's able to learn and do more and more on their own with his specific guidance (it can learn, but Tao's guidance is watered down with all the other things the LLM learns from), but rather AI can serve as a kind of mediocre or incompetent assistant doing the manual, repetitive, grunt work typical in math research.

Tao seems to see that maybe a human can do much more "inspired" work with very little data, by using AI to do much more of the manual, repetitive, grunt work with much more data.

No, I don’t think so? I think in Ghostbusters 2016, comedic improv is used for funny dialogue, but the overall story narrative was pre-written?

Wong Kar-wai uses improvisation to explore characters and emotions to help write the story. Whole narratives can change based on the actor’s acting.

See: Tony Leung explaining how his character in WKW’s Happy Together changed drastically after a week of filming: https://youtu.be/KkNBh24cLvo&t=2m45s (it’s just 20 seconds)

So he was trying to be like Wong Kar-wai [1]:

> Wong has an unusual approach to film making, starting production without a script and generally relying on instinct and improvisation … cast are given a minimal plot outline and expected to develop their characters as they film

Edit: Yes:

> the director has shared a list of 19 movies that inspired Megalopolis … Wong Kar-wai's Chungking Express. [2]

[1]: https://en.m.wikipedia.org/wiki/Wong_Kar-wai#Method_and_coll...

[2]: https://movieweb.com/megalopolis-director-reviews-his-own-mo...

Wait, you saying in Canada a R&D software company can essentially sell a dollar (of SDE produced goods) for a dollar and get a tax refund from the government?

There was this alleged Alberta AHS privacy breach:

https://old.reddit.com/r/alberta/comments/1c7lk3z/ahs_privac...

Don’t know if that went anywhere… anyone know?

That’s a lot of different things to consider and try, thanks!

What worked for me was drinking massive amounts of water throughout the day.

Not so much to cause hyponatremia, of course, but it was a cheap easy way to establish a baseline for me (i.e. helps with dry climates, salty foods, dehydration, etc). And yeah, I tried some of those other things too.

> o1-preview is about on par with Anthropic's Claude 3.5 Sonnet in terms of accuracy but takes about 10X longer to achieve similar results to Sonnet.

> o1's performance increase did come with a time cost. It took 70 hours on the 400 public tasks compared to only 30 minutes for GPT-4o and Claude 3.5 Sonnet.

See details at https://arcprize.org/blog/openai-o1-results-arc-prize

Also discussed here: https://news.ycombinator.com/item?id=41535694

Jevons Paradox?

https://en.wikipedia.org/wiki/Jevons_paradox

> Golangs stupidly verbose error handling forces you to constantly think about the unhappy paths. Every function is littered with error handling and it makes you aware of possible problems and rigorously forces you to not forget them.

That reminds me of this [1]:

> However, if you are going to make a lot of state changes, having them all happen inline does have advantages; you should be made constantly aware of the full horror of what you are doing. When it gets to be too much to take, figure out how to factor blocks out into pure functions (and don't let them slide back into impurity!).

[1]: https://cbarrete.com/carmack.html