HNHacker News
TopNewBestAskShowJobs

sbuttgereit

6,901 karma · joined June 14, 2015

Business systems implementation and development consultant Muse Systems (https://muse.systems) about me: https://buttgereit.com
submissionscomments
sbuttgereit··on Waiting for SQL:202y: Group by All
At least in PostgreSQL, both by alias and ordinal are possible:

  localhost(from SCB-MUSE-BOXX).postgres.scb.5432 [Sun Nov 16 12:02:15 PST 2025]
  > create table test (a_key integer primary key, a_group integer, a_val numeric);
  CREATE TABLE
  Time: 3.102 ms

  localhost(from SCB-MUSE-BOXX).postgres.scb.5432 [Sun Nov 16 12:02:25 PST 2025]
  > insert into test (a_key, a_group, a_val) values (1, 1, 5.5), (2, 1, 2.6), (3, 2, 1.1), (4, 2, 6.5);
  INSERT 0 4
  Time: 2.302 ms

  localhost(from SCB-MUSE-BOXX).postgres.scb.5432 [Sun Nov 16 12:02:58 PST 2025]
  > select a_group AS my_group, sum(a_val) from test group by my_group;
   my_group | sum
  ----------+-----
          2 | 7.6
          1 | 8.1
  (2 rows)
  
  Time: 4.124 ms
  localhost(from SCB-MUSE-BOXX).postgres.scb.5432 [Sun Nov 16 12:03:15 PST 2025]
  > select a_group AS my_group, sum(a_val) from test group by 1;
   my_group | sum
  ----------+-----
          2 | 7.6
          1 | 8.1
  (2 rows)
  
  Time: 0.360 ms
sbuttgereit··on Waymo robotaxis are now giving rides on freeways in LA, SF and Phoenix
What does happen in that situation sometimes is that if you're speeding along with everyone else, that sometimes gets allowed as a defense against a speeding ticket; basically the argument is that if driving the speed limit when all surrounding traffic is driving over the speed limit, that the relative speed difference can matter. At least that was true when I learned to drive (many, many years ago).
sbuttgereit··on Blue Origin lands New Glenn rocket booster on second try
I'm reading this thread and there are a few things that come to mind.

My sense is that SpaceX's goals with Starship are significantly more ambitious than what is being tried with New Glenn. I don't mean to underplay the difficulty of what Blue is facing with New Glenn, but if we take that "rapid reusability" goal seriously the problem set seems significantly larger and not so "been there, done that". This makes the development programs much more difficult to compare.... certainly on the surface of the public optics at the very least.

While it's one thing to talk about rockets, it's another altogether to look and the engineering and practices going into the manufacture process of those rockets. I'm not an engineer, but I do work in manufacturing and, at least looking from the outside, SpaceX seems to be dedicating some significant amount of effort into building a scalable manufacturing process. Many other efforts have always appeared to be more about "bespoke" production even if the designs of each unit produced are constant. I could be wrong and maybe it's just SpaceX is a lot more transparent (willingly or otherwise)... but looking in from the outside, they seem to be developing a very mass-production oriented rocket factory.

And if New Glenn is just finalizing things and Starship is just vaporware... well New Glenn still has to land a couple more boosters and re-fly one (or two?) to catch up to those vaporware numbers. :-) Sure, New Glenn has now flown a paying customer... but I think we'll see Starlink launches on Starship pretty soon... well before it gets to "final validation".

sbuttgereit··on Blue Origin lands New Glenn rocket booster on second try
Beautiful launch and landing.

I still can't stand the public relation heavy official stream... but even with all that static the rocket itself cut through.

sbuttgereit··on GPT-5.1: A smarter, more conversational ChatGPT
This. When I go to an LLM, I'm not looking for a friend, I'm looking for a tool.

Keeping faux relationships out of the interaction never let's me slip into the mistaken attitude that I'm dealing with a colleague rather than a machine.

sbuttgereit··on How I fell in love with Erlang
The one that I can share is at: https://github.com/MuseSystems/musebms

A few things to keep in mind. This is a "labor of love" personal project... sure a large one... but that's what it is. Another thing to consider is that I go off standard Elixir practices in a number of different ways and for reasons™; in this sense the project isn't representative of what a typical Elixir project looks like.

When I say that things take longer for me to initially code, I'm not including "self-inflicted" time dealing with novel issues which arise from taking a less common path... which does have its own time cost. In my previous comment I'm only really thinking of the typical aspects of Elixir development time.

Finally, I do need to emphasize: for me the cost is worth it. Once I get to maintaining that now-developed code or doing refactors, so long as I'm not throwing away everything and starting from scratch, the development time for me is rather less than I have with code of similar complexity developed in other languages. All personal experience... but there you have it.

sbuttgereit··on How I fell in love with Erlang
I do a lot of Elixir programming, which is different and not at the same time :-) . What I find is that it takes me longer to write the initial code than I might in, say, JavaScript, but that the refactoring and ongoing maintenance of code is significantly faster. This is because I think the functional & actor paradigms force you to think harder upfront, but once those decisions are made are easier to manipulate. For me personally, I'd take that trade since I'm either refactoring or maintaining much more frequently... but I understand there are other considerations which could drive a different preference.

Also, I'm sure my personal experience isn't necessarily true for everyone... so weight this anecdote accordingly.

sbuttgereit··on Writing your own BEAM
...and LFE and Luerl and probably others that I'm forgetting.
sbuttgereit··on BlackRock's Larry Fink: "Tokenization", Digital IDs, & Social Credit
It's a fair point. I actually posted the article mostly to get discussion... and before I realized exactly what the rest of the website was about. So I would urge taking things in that spirit, including feeling free to call out that the assertions in the article are wrong in fact or interpretation.

I do think it raises some fair questions about the drive to certain forms of digital commerce and identification. Perhaps, just because you're paranoid doesn't mean they aren't out to get you.... at least sometimes...

For the record, I've seen no convincing evidence for the existence of a divine presence of any sort and don't support most of the site's messaging. Sure, we could be on the edge of real apocalypse, but am very doubtful that, should that day come, it will be the Christian Apocalypse or the same prophesied from any other faith for that matter.

sbuttgereit··on Trump directs nuclear weapons testing to resume for first time in over 30 years
> In the 1980s there were huge protests about jucelar weapons that actually resulted in detente and a relaxation of tensions.

Really? Yes there were protests, in the west (that sort of thing was generally frowned upon in the Eastern Bloc countries), but they didn't do much.

What I recall happening was that one of the principle parties started to suffer from a worsening economy and could sustainably maintain the spending the cold war demanded. When that party became much more conciliatory, progress on arms control resulted. Didn't end up saving them in the end.

sbuttgereit··on Doomsday scoreboard
argumentum ad populum and argument from authority in one sentence... :-)
sbuttgereit··on I Test Drove a Flying Car. Get Ready, They're Here
This company must have some sort of marketing push underway... here's a video yesterday from Scott Manley that's basically the same thing:

https://www.youtube.com/watch?v=wncRFPd69rg

sbuttgereit··on SpaceX launches Starship megarocket on 11th test flight
> The booster part is actually just demonstrated

For the second time.

sbuttgereit··on Circular Financing: Does Nvidia's $110B Bet Echo the Telecom Bubble?
In leasing to OpenAI, they aren't deploying those GPUs to other productive uses.

The costs are not just the raw cost of the physical asset, but also the opportunity costs of doing "thing a" vs. "thing b".

The GPUs will also depreciate during the time they're in OpenAI's possession, so again, if that investment doesn't pay off they aren't just getting GPUs back, they're getting "used GPUs" or "more used GPUs" back, not the original or full value of the originally leased asset. Naturally, the lease has to have those costs built in to be a good deal, but the lease has to fulfill to terms for that to happen.

In the end, leasing them means they carry the early risk of the lease not being fulfilled... but they should gain more in the than if they just straight up sold them.

sbuttgereit··on Circular Financing: Does Nvidia's $110B Bet Echo the Telecom Bubble?
That Nvidia has to front the costs of the product at the beginning and arguably the risk that the allocation of assets end up not being paid off (bankruptcy, etc.) By carrying those costs early and the associated risk, Nvidia expects a return on that. If the risk is realized they'll lose but otherwise they'll gain. That has all the hallmarks of an investment.
sbuttgereit··on I built ChatGPT with Minecraft redstone [video]
If you like thinking about the possibilities of the worlds of Minecraft and LLMs colliding I recommend the YouTube channel @EmergentGarden (https://www.youtube.com/@EmergentGarden) it's not exclusively agents and Minecraft, but there's a fair amount of stuff tested over time.

Most recently: Can AI (actually) beat Minecraft? - https://www.youtube.com/watch?v=Wh4abvcUj8Q

sbuttgereit··on I built ChatGPT with Minecraft redstone [video]
Or just use Bedrock edition.
sbuttgereit··on Just let me select text
I don't think we disagree, too much. tab panes matches the "button" example.

However, I am sympathetic to those arguing translation. Sometimes I'll visit Japanese or Chinese websites. With some frequency, even if most of the site has an English edition, I'll find some UI element not translated, including buttons and the like... OK I think it was the commenter that I responded to, in a different reply said... just Google it if it's a single word. Great! But I don't even know where to begin to get the right characters from my old fashioned US keyboard. So now I have to Google for how to use my keyboard to get the characters I want, which also may need pre-requisite knowledge of the language I'm trying to translate (radicals and all that jazz)... that's a heavier lift than may be anticipated and where a simple copy/paste into an appropriate translator would make things much, much easier.

I would suggest this: make everything buttons, links, tabs, etc. selectable and copyable unless there is a real explicit and compelling reason to do otherwise. Now to be fair, I'm old enough to have been "online" in some fashion or another since before general public internet access availability was a thing... so my expectations for butter-like user experiences are low and my desire to do any damn thing I want high... but even today, there are probably still more websites which don't stop you from copying anything than there are searching for that polished experience where only the right things can be selected. The discontinuity and the deviation from the expectation that I can copy anything I also find as something which diminishes the user experience, even if occasionally I'm annoyed by over selecting things.

sbuttgereit··on Just let me select text
A case can be made for graphic like elements like buttons, but for text: treat it like text even if it's clickable.

In the Web version of Outlook, there are regularly times where the location of an appointment is a street address. That text is typically clickable. But the click action doesn't correspond to the choice of mapping service I might want to use in any one instance or to the fact that I might have other actions, like copying the address into another email/sms/etc. Outlook followed your philosophy. You can't select and copy that text, save for going through several auxiliary clicks just to get to a spot where you can. It's the most annoying behavior I can imagine.

That you think that you sitting in a meeting room talking it over with colleagues, or perhaps I'm a meeting in your own mind can assign legitimate uses and not, when something other than say security might be at stake, is just wrongheaded.

And by the way, that address being the link that it is is great 60%, 70% of the time. But when it's not it's clearly a design mistake.

sbuttgereit··on Kmart's use of facial recognition to tackle refund fraud unlawful
The store tends to always be there, there is often times need to have a lot of them, and they're available at well known (even published) times whereas the van isn't always at the same places at the same time and doesn't even carry predictable goods... just that you know there will be goods. Sure you can guess or make your luck by waiting for the van or search for it in good spots... but the cost is higher for the criminal to try and count on such a thing.

The truth is I expect stealing from a delivery van is ultimately simpler... or simply stealing the package off the porch easier still. The issue isn't the ease or difficulty.

Where I expect consumer delivery businesses to do better in the face of theft is on the cost of theft (assuming a certain scale in the delivery business). Given the economies of scale of a warehouse and the delivery model vs traditional retail locations, I bet means the loss for any item stolen from the van is less than that of the same item stolen from a traditional retail location.

sbuttgereit··on Kmart's use of facial recognition to tackle refund fraud unlawful
Or... just open up big warehouses, only do online sales, and then deliver to customer?

The truth is we have tried it and on a large scale: The Automat (https://en.wikipedia.org/wiki/Automat). Don't see to many of those around anymore, except maybe analogues in Japan.

With some perspective on the idea, would you invest in the retail real estate, the technology development, and later maintenance, and then still need to have staff to stop people from just breaking into the machine?

sbuttgereit··on M4.6 Earthquake – 2 km ESE of Berkeley, CA
And that was just the automated response... it was downgraded to 4.3 an hour after the event.

4.3 will certainly get your attention if you're relatively close-by... but yeah, worth a "did you feel that?!" on the local news and not much more.

sbuttgereit··on Zoox robotaxi launches in Las Vegas
I'm sitting in the back of a Lyft car right now... I had to prompt the driver via a phone call to actually try to pick me up at the designated airport pickup spot (you know, where the app has me go), he spent 10 minutes trying to get out of the airport parking lot because he didn't seem to have a ticket, and now his constant pumping of the gas peddle in the sluggish Los Angeles traffic is challenging even my ironclad resistance to motion sickness.

How I wish I was in a Waymo right now! I've never had remotely such a poor experience in a SF robotaxi.

sbuttgereit··on Starship's Tenth Flight Test
They've done so more than than this time, too. Granted, with a little less "Ship" than they left with on all occasions I know of.

No one (at least not me or anyone I take seriously) is arguing whether or not these suborbital profiles are designed to be safe even under adverse or full failure conditions; though the Caribbean air corridors might have been managed a bit more gracefully on some previous flights... still...

Nonetheless there is a valid criticism that in ten flights they still haven't mastered keeping the control surfaces of the space craft whole during the reentry phase of flight. 1500 miles isn't going to cut it as a safe return zone when they try bring this in for a catch. While I'm as impressed as anyone that they've hit the mark with compromised Ships as many times as they have, neither Port Isabel nor Titusville are 1500 miles from their nearest Ship catch towers and I wouldn't support any attempts for a catch until they can get the whole Ship back in good working order... reliably. While I'm a advocate for this program and SpaceX... I'm not such a fanboy that I can't see there are issues with this aspect of the program. This is ignoring the impact on rapid reusability and simply focusing on the basic safety of the program.

sbuttgereit··on Starship's Tenth Flight Test
They're never reaching orbital velocity on purpose. The reason is that they're still proving that they can fire rockets and deorbit under control; until they do that, any problem automatically puts the ship in the ocean no matter what while if they go to orbit and can't control the deorbit they end up possibly causing a disaster.

Now, they are getting it to pretty damn close to orbital velocity... which is why saying they still haven't reached orbit is a bit silly. They're clearly technically able to reach orbit if they really want to... that they haven't proved they can safely leave orbit is the problem.

sbuttgereit··on Starship's Tenth Flight Test
What's surprising is that people are still resorting to this silly complaint about not reaching orbit when there's a perfectly sensible complain they could be making instead: that SpaceX hasn't yet demonstrated that they can reach orbit and return safely. The safe return is important because I would expect a failure to return safely to be a big deal: it's not like this thing is going to completely burn up if they don't have control during a deorbit. The inadequate retry thermal protection is a large issue even if the Ship has managed to get to the landing areas on target and soft land in the right spot: the burn through on the control surfaces seems to mean that was as much luck as good engineering that the thing didn't crash somewhere less intended.

I appreciate none of that is as pithy as saying it simply didn't reach orbit, but it's a real concern versus something that is really irrelevant.

sbuttgereit··on Starship's Tenth Flight Test
Where are they claiming that?

For example, this is from their Flight 4 press release (https://www.spacex.com/launches/starship-flight-4):

"Flight 4 ended with Starship igniting its three center Raptor engines and executing the first flip maneuver and landing burn since our suborbital campaign, followed by a soft splashdown of the ship in the Indian Ocean one hour and six minutes after launch."

Note that they clearly say since the start of their suborbital campaign. And this from their Flight 6 press release (https://www.spacex.com/launches/starship-flight-6):

"Starship completed another successful ascent, placing it on the expected trajectory. The ship successfully reignited a single Raptor engine while in space, demonstrating the capabilities required to conduct a ship deorbit burn before starting fully orbital missions. With live views and telemetry being relayed by Starlink, the ship successfully made it through reentry and executed a flip, landing burn, and soft splashdown in the Indian Ocean."

And from today's pre-launch press release (https://www.spacex.com/launches/starship-flight-10):

"The Starship upper stage will again target multiple in-space objectives, including the deployment of eight Starlink simulators, similar in size to next-generation Starlink satellites. The Starlink simulators will be on the same suborbital trajectory as Starship and are expected to demise upon entry. A relight of a single Raptor engine while in space is also planned."

To be fair, there were two press releases where they didn't correctly use "sub-orbital" and used orbital instead. Releases 3 and 9. Neither said they achieved orbit, but more causally talked about the "orbital coast" and the worst: "Starship's six second stage Raptor engines all started successfully and powered the vehicle to its expected orbit" from flight 3. It's true these statements are incorrect, but they aren't asserting a direct claim to having reached orbit (though they imply it), when they make an assertion about the nature of the program they seem fairly consistent in talking about their "suborbital campaign" as well as talking about their orbital missions being in the future.

The way I'm reading it, it looks like they get sloppy with language sometimes, but it doesn't look like they are directly asserting anything other than being in a suborbital program.

sbuttgereit··on Starship's Tenth Flight Test
I think Scott Manley's position on the "still hasn't gotten to orbit" take is probably still the best and most accurate:

https://www.youtube.com/watch?v=8htMpR7mnaM&t=420s

sbuttgereit··on Against the UUID
> A UUID is a collection of random characters like f81d4fae-7dec-11d0-a765-00a0c91e6bf6

This statement is just wrong. More accurately, A UUID -can be- a 128 bit number made up of mostly random bits (UUIDv4) which can be represented as a string using a common representation. Sure, that may well be a common case for most, but there are very completely non-random UUID versions as well, such as UUIDv1 or, more recently, the somewhat random UUIDv7.

The author's proposed ID system is less defensible against some of the other UUID versions which aren't addressed. But that initial description suggests that the author doesn't have a complete enough understanding of UUIDs to really make a credible case about UUID problems (which do exist) vs. their own proposed system.

sbuttgereit··on Preliminary report into Air India crash released
And yet the preliminary report for the incident in question includes reference to that bulletin, indicates that the switches in the accident aircraft were of a very similar design and subject to advisory inspections:

"The FAA issued Special Airworthiness Information Bulletin (SAIB ) No. NM -18-33 on December 17, 2018, regarding the potential disengagement ofthe fuel control switch locking feature. This SAIB was issued based on reports from operators of Model 737 airplanes that the fuel control switches were installed with the locking feature disengaged. The airworthiness concern was not considered an unsafe condition that would warrant airworthiness directive (AD) by the FAA. The fuel control switch design , including the locking feature, is similar on various Boeing airplane models including part number 4TL837-3D which is fitted in B787-8 aircraft VT-ANB. As per the information from Air India, the suggested inspections were not carried out asthe SAIB was advisory and not mandatory. The scrutiny ofmaintenance records revealed that the throttle control module was replaced on VT-ANB in 2019 and 2023. However, the reason for the replacement was not linked to the fuel control switch. There has been no defect reported pertaining to the fuel control switch since 2023 on VT-ANB."

So while I agree that this being the cause sounds unlikely, referencing the switch issue is something relevant enough for the report itself.

← PreviousPage 4 of 34Next →