HNHacker News
TopNewBestAskShowJobs

7532yahoogmail

122 karma · joined June 7, 2019

submissionscomments
7532yahoogmail··on The automotive industry’s value-destroying addiction to capital (2015) [pdf]
You forgot to read the PDF
7532yahoogmail··on The automotive industry’s value-destroying addiction to capital (2015) [pdf]
Agree. In the US it makes sense to eliminate cars; consumers have voted for SUVs. I guess Detroit will need EV SUVs. I find pharma's roic and ebita suspiciously high; pivot to healthcare cost inflation. And I am surprised by retail numbers. Still the OP's point stands ... A damn interesting read btw. There are other articles talking about what Tesla has done to Germany. Their combustion based cars are falling in sales. The number 1 purchased EV is tesla not German. Yikes!
7532yahoogmail··on Lagrangian Neural Networks
Great post. This is why I love hacker news
7532yahoogmail··on Still at Guantánamo
Guantanamo un-American. I do not support it. Presidents including Obama have tried to close it. Congress - the absolute worst of the three Federal branches - has stopped it. It's another problem in a whole line of problems their inability to act has caused.
7532yahoogmail··on A new Go API for Protocol Buffers
Umm ... When will pb allow me to decode a field in buffer without having to decode the entire object in the buffer? I'm talking a field in a pb message NOT a byte stream I hand encoded without reference to a message eg more like flat buffers.
7532yahoogmail··on A new Go API for Protocol Buffers
But the protoc compiler already makes data structure definitions in the target, client language. So what's the point of a parallel macro system? Either it's a subset of capability of what protoc does or equal. Former questionable; latter redundant.
7532yahoogmail··on How I Did It: From Psych Degree to VP of Data Science at Top Startup
Look, there is an essential difference between what is an engineer and what is a technical person. The later want direction, projects broken into epics and then stories, and wants steps. Engineers can make that structure and work better in ill defined problem areas. That's not a flaw. What is a flaw is when tech never progresses to engineer. That's bad for the individual, team, company, and client. Leadership then (in the drucker or tqm sense) is important. Look at IBM pre/post gernster and Microsoft pre/post balmer.

In short rigidity on the part of individuals combined with countervailing organizational goals to combatting same is the issue.

7532yahoogmail··on A journey to searching Have I Been Pwned database in 49μs
Nicely done and explained.
7532yahoogmail··on Handling 350k Requests for $3 using Lambda
Title needs a rate ... 350k requests/month.

Ok: the rest of the article was great. Thanks

7532yahoogmail··on A Text Renaissance
If only the OP would have used one more flowerly, sales, marketing, adjective, or emotive word then I would have succumbed to his charm.

One more is all we're asking for.

But alas this piece is a bunch of cheap symbolism.

Op whatever is on your mind, spit it out and stop the bs.

7532yahoogmail··on I Want Off Mr. Golang's Wild Ride
Best back handed compliment I've seen in a while ... And right too.
7532yahoogmail··on I Want Off Mr. Golang's Wild Ride
I'm c/c++ over 20 years, go 1 year, Python 5+ years. I work at a company with a guy on the c++ standards committee with the internal sdlc and engineering training for large scale, commercial systems to boot.

This article is a rant. Not an engineering take down of go. There's just not much of substance here. Were I to care about windows (I don't) for serious cross platform os interaction, go isn't your hammer of choice.

I've turned to go recently for some I/O heavy apps of a micro-service type which it is fine for. I also turned to go because of God awful c++ build times and bad build systems in the sense that they assume all code is in a single branch. By switching to go I also prevent less experienced programmers from linking in legacy c++ libraries and the evil that comes with them.

Go has delivered. My needs are such that protobuf/flatbuffer are good enough for types and go's lack of generics is irrelevant. I'm pushing bytes across a network pipe in which each message admits simple transforms/operations.

Now I am keeping my eye on three things that I think go could burn me on:

- garbage collection

- channels ... cool but slow

- something unixy/multicore/close to the bare metal ... Like kv store

Those things I'd be reticent about doing in go.

Folks, we need 2-4 languages with their connections to libraries and tool chains in our toolbox.

While we remain dominated by c++ (a complex beast of a language) I am looking to add a functional language to my kit (ocaml/Haskell). Btw good engineers need a formal language too. I recommend tla+ and there's a guy in hacker news here that's got good books on it. Recommended! Highly concurrent code ought to modeled in tla+ first before leaving your app language gun and taking the canolli.

Cheers

7532yahoogmail··on Object-Oriented Programming – The Trillion Dollar Disaster
Like the thought. And list.
7532yahoogmail··on Object-Oriented Programming – The Trillion Dollar Disaster
(Repost & from 2019).
7532yahoogmail··on IOTA being shut off is the latest chapter in an absurdist history
I am laughing ... Laughing my b u double t off at this. In 2019 I spent too many hours debating crypto as currency replacement or substitute. Never bought it. But this is like m.pythons+galaxy hitch hiker+pink-panther level of crazy dumb.
7532yahoogmail··on Andreessen-Horowitz craps on “AI” startups from a great height
Thank you for the perspective. Now when we talk machine learning are we talking:

L. Pachter and B. Sturmfels. Algebraic Statistics for Computational Biology. Cambridge University Press 2005.

G. Pistone, E. Riccomango, H. P. Wynn. Algebraic Statistics. CRC Press, 2001. Drton, Mathias, Sturmfels, Bernd, Sullivant, Seth. Lectures on Algebraic Statistics, Springer 2009.

Or more like:

Watanabe, Sumio. Algebraic Geometry and Statistical Learning Theory, Cambridge University Press 2009.

My understanding (I do not do AI or machine learning) that AI is distinct from these more mathematical analytic perspectives.

Finally, might we argue that generally AI/ML is more easily suited to data that's already high quality eg. CERN data, trade data, drug trial data as opposed to unconstrained data eg. Find the buses in these 1MM jpegs?

7532yahoogmail··on Past Time to Tell the Public:It Will Probably Go Pandemic, We Should Prepare Now
This article does not read like cogent CDC advice. It seems written in a more page-6 OP-ED style. Not sure this is the perspective we need.
7532yahoogmail··on Confronting Indifference Toward Truth – Dealing with Workplace Bullshit
This topic is surely worth a lot more discussion. It's a systemic pronlem that needs a systemic solution. I would heartily recommend "The Human Element" by Schutz. in this regards. It's worth reading because of its simplicity, effectiveness, and because it deals head-on with this subject of BS.

Among the major arguments,

* The single biggest obstacle to good teamwork (well defined therein) is rigidity on the part of individuals

* Rigid individuals are unchanging because the person's self esteem is threatened. Teams are busted by individuals who don't change behavior.

* And ultimately the biggest problem in the long run for human organizations is lack of openness.

* The book explains why this is so.

The background and impetus for this work (Schutz was Harvard trained) came out of WWII Navy experience dealing with battle conditions. The Navy realized a major impediment to better performing battle execution was that the bridge officers could resolve conflict well. It's not that conflict exists; this is intrinsic to any open ended, unconstrained problems in an evolving environment.

To respond the Navy post war completely revamped training. They chose physically fitter officers, smarter, and trained them harder. The results however under test exercises were random: sometimes teams worked and sometimes they didn't. The Navy couldn't predict when or why.

Schutz was asked to consult on this problem and through this work and in corporations subsequently this work was born.

If you work in an environment where office BS reins, and where management of teams really amounts to people tinkering with a Markov like hidden variable random process until something works, this book is for you.

7532yahoogmail··on Environmental lawyer who won a judgment against Chevron lost everything
Right. Good explanation
7532yahoogmail··on Environmental lawyer who won a judgment against Chevron lost everything
You'd be giving money to the company if the shares actually came from them. If you're buying on an exchange ... you're exchanging cash for shares with the counterparty almost never ever the company itself. The company is not acting as a market maker on its own stock, for example. The only way the company would benefit is if the purchase increased the ask price so that at the end of the day the price per share was higher.
7532yahoogmail··on Why events are a bad idea for high-concurrency servers (2003) [pdf]
Threads and events are duals. So it is possible to do both well if, in the threads case, the implementation is done right. The author makes the case he's done implementations well. The salient question then is: can the average programmer do threads based concurrent programming well? As dual, I think we can all agree that more effort/intelligence will lead to better solutions for threads or events.
7532yahoogmail··on Ask HN: Simple cluster self-healing system?
Thank you for suggestions
7532yahoogmail··on Ask HN: Early in career – is it a bad idea to change jobs after six months?
Dear OP: here's some career advice: you're not thinking right or doing right if you can't figure this out for yourself. Just like I've said on reddit's lpt form reality is you gotta figure it out for yourself and stop shopping your life around on the internet
7532yahoogmail··on Ask HN: Simple cluster self-healing system?
No, that's a simple throw away response
7532yahoogmail··on Ask HN: Simple cluster self-healing system?
To further clarify all I need is a system to:

- ask for task heartbeat

- make sure nodes in cluster pingable

- run N copies of task on reachable hosts

That's it.

7532yahoogmail··on Why the foundations of physics have not progressed for 40 years
Agree!
7532yahoogmail··on Why the foundations of physics have not progressed for 40 years
I've read several of Sabine's blogs over several months. I think she has very good ends in mind, has courage to push back on corporate/academic inertia ... such inertia comes with any human organization ... On the negative side she's big on complaining but small on alternatives. She's also a bit too blunt/dismissive of people -- this from a person who also dislikes corporate happy talk. As such it's not clear if she'd confer distinction if she had a large budget, an institution, and group of experimentalists. An Oppenheimer? No.
7532yahoogmail··on Broot – A new way to see and navigate directory trees
For some at the office midnight commander still rocks their world....
7532yahoogmail··on Measuring mutexes, spinlocks and how bad the Linux scheduler is
No that was the conclusion: use mutexes
7532yahoogmail··on I asked my students to turn in their cell phones and write about it
Wow. Really, really bold leadership here. He actually asked students to go without their phones ... the student's family were happy the teacher didn't ask for them too. Such power!

Meanwhile, a university professor,

- isn't mom/dad by pseudo punishing students removing phones

- isn't moonlighting as pseudo pysch researcher who guessed phones were the problem

- isn't a cool guru by wowing kids into wisdom with this unexpected ploy

If your students did bad on their exam

- make improvements as a teacher if you feel some responsibility lay with you (there's no need to gas on about this to students)

- leave the grades unchanged

- stop confusing the issue with stupid exercises then self marketing it into media

← PreviousPage 3 of 4Next →