sixdimensional

Give users an AI assistant they can ask to navigate them to the right screen or section of the application?

In a previous job, we built our AI assistant so that it could operate our UI in the front-end and it was very powerful.

It wasn't an ad. I used to work for the parent company of PWS years ago, and their weather model was really awesome.

I thought the TAMDAR tech reference might be if interest to those following and reading comments here.

I'm not associated with them anymore and I'm not sure how much of their model still exists after the assets were sold to flyht, but I hope it does because we need good weather models!

Fair enough being vigilant, but not an ad.

An interesting anecdote from a past job - Flyht [1] acquired the assets from Panasonic Weather Solutions (PWS) some time ago [2].

The weather model from PWS was incredible, using TAMDAR sensors mounted on commercial aircraft, which recorded data at multiple altitudes getting a great sample of conditions from many layers of the atmosphere as aircraft ascend and descend.

IIRC, during hurricane season some years ago, the PWS model outperformed hurricane path predictions from many other organizations in terms of accuracy, among other things.

I love this share too, and I suspect this weather data could be very useful!

[1] https://flyht.com/

[2] https://flyht.com/investors/news-and-media/view/flyht-acquir...

Is 2001 the stone age now (the year YAML was conceived)?

IMHO, modern society critically undervalues the importance of open-ended play in all aspects of work and life.

Parallel production lines with overlapping sequences also cause this, especially in aerospace- it is widely known that Airbus does it, for example.

Make = who made the thing

Model = what is the type of the thing

Serial Number = was supposed to be what order the things were made in (e.g. the number of the serial order), but this is often obfuscated or often repeats [1].

In cars, make would be like Ford. Model would be like Focus, serial number would be VIN (vehicle identification number - in cars, those are generally unique!).

Ford Focus + VIN, basically.

There is a theoretical concept of a unique identifier for everything... including people from ISO under ISO 8000.. combining a natural location identifier (eNLI)[2] and an ISO8601 timestamp - to represent "where and when a thing is considered to be born" - a point in time and space the thing is considered to come into existence.

I think the idea is called "natural person identifier" for humans.

This ID has to be assigned but I think you can see the idea at least.

I suppose this doesn't include make/manufacturer but realistically that isn't needed for uniqueness in this scheme, only as descriptive metadata for things that have one.

[1] This is related to the fact that if serial numbers were truly serial, one could estimate the rate and quantity of production which is considered sensitive information by most manufacturers. This relates to "the German tank problem" - during WWII the allies were able to accurately estimate the production of German tanks by analyzing the serial numbers off captured tanks.

[2] https://eccma.org/enli-eccma-natural-location-identifier/

Agreed!

Exactly!

Here's a fun true story..

Aircraft do not have a singular unique identifier that is time invariant.

While it is true that aircraft have serial numbers issued to their airframe, by itself, aircraft serial numbers are not unique.

The only unique identifier for an aircraft across its lifecycle from production to end of life is a combination of the manufacturer, make and serial number.

I know this because I am on (for better or worse) the patent that involves defining that as a unique identifier for aircraft.

The combination of ICAO aircraft type designator + serial number approximately is the most permanent identifier for an airframe - and even then - if an airframe is modified significantly enough that it no longer is the previous type - even then this identifier can change.

Personally, it boggled my mind that something as big as an aircraft did not have a simple time invariant unique identifier.

P.S. For those who might ask - aircraft registration numbers are like license plates, so they change - tail numbers can be ambiguous and misinterpreted depending on what is painted on the aircraft where, and ICAO 24-bit aircraft addresses are tied to ADS-B transponder boxes, which technically can be moved and reprogrammed between aircraft also.

I was not up to speed, there seemed to be a period of time when CentOS support was dwindling and it even looked like the project might no longer have upstream support.

Regardless of the complex history of CentOS, RHEL, etc it seems like the original thrust of CentOS lives on rebranded as RockyLinux for example.. and CentOS stream serves Redhat's needs/purposes.

I think forking still is a valid method to resolving challenges for NATS here.

I would wonder if paying for a legal fight to recapture the trademark is even worth it.

Couldn't Synadia just fork what they want to and rebrand for a commercial new product, and go forward from there?

If they wanted to backport or keep contributing to the open source project they could, or leave some parts open source going forward, while forking only the server for example, to do that.

Kind of like how RHEL used to still provide contributions to CentOS for a long time (although I think that ended at some point too).

This is a fair concern, I imagine. If it is highly user serviceable, maybe that isn't a concern.

That said, I think you raise a bigger issue - I'd like to see MORE things like Framework, Fairphone or Slate - user serviceable, customizable - maybe low initial cost.

To me, this feels futuristic, exciting, optimistic and positive.. we need more like this, so how can we make these kinds of businesses more likely to succeed, resilient, etc?

Hi zackmorris,

If I may so bold as to refer to you as "my friend" (having never met you)...

My friend, I think I understand what you mean. I am about the same age too.

I would like to propose an idea to you - and it is something I have been exploring very deeply myself lately.. maybe the thing we need to start spending our time on is exactly this meta problem now. The meta problem is something like (not perfectly stated): we as humans have to decide what we value such that we can continue to give our existence purpose in the future.

I don't think AI is going to be the be-all-end-all, but it is clearly a major shift that will keep transforming work and life.

I can't point yet at a specific job, or task - but I am spending real time on this meta problem and starting to come up with some ideas. Maybe we can be part of what gets the world, and humans, ready for the future - applying our problem solving skills to that next problem?

I mean all of the above in 100% seriousness and I am willing to chat sometime if interested to compare notes.

I don't like the idea of intellectual theft without attribution/citation. If it is shared publicly/into the public domain, I believe the least that should happen is attribution/citation where possible - especially if a for-profit entity is using the content.

On the other side, for many of my thoughts/ideas/creativity, I find it is often a question of, do I die alone with it in my head or put it out there, maybe for someone to find, use or be inspired by? If it's a question of one or the other, I think perhaps I'd rather let it go free so the value isn't lost (assuming, that what I have to write/say/share is of some value).

I have seen that mechanism be exploited.. sometimes you put something out there, someone uses it, and there is zero attribution/citation. For myself, I prefer to try (most of the time) to treat non-fictional writing in the traditional sense of an academic paper and include citations whenever they are useful and I can - even for non-academic purposes - for the purposes of giving credit to others. It's nice to share credit and show how things connect, IMHO.

FWIW, I like some of the inspiration behind Creative Commons, but even old school citations (MLA/ALA or heck, even just this kind of inline citation style [1] that is common on HN and other sites is better than zero).

There are citation generators online also [2] - for formal purposes/writing/papers you must get these formats right, but for informal/casual use, I think simple citation is better than no citation.

[1] simple citation style, or even just paste the URL - at least you're giving credit.

[2] example, https://zbib.org/

There is a docker image at least. Would be cool as a standalone tool though, I kind of agree.

I'm not related to the author or developers, but this appears almost like an Airtable-like (or Nocodb) UI on top of postgres, so it's really quite a different experience than a full database manager tool/IDE like DBeaver or DataGrip.

I love it personally and it looks pretty good so far. Also, amazing that the team started it as an open source project under a 501c3 nonprofit as well.

Having a more "spreadsheet" (or I would say, MS Access-like) UI is quite cool for certain use cases (e.g. low/no-code but backed by the full power of regular old Postgres). Can allow pretty agile development, although - it can get out of control if not managed with care... but nice to see the access controls/permissions there to help a bit.

All interesting choices, looks cool!

Not sure about the previous post, but also python’s OrderedDict collection guarantees order-sensitive equality checks. [1]

Plain dict maintains insertion order but equality checks only check that the key/value pairs are the same. [2] [3]

[1] https://docs.python.org/3/library/collections.html#:~:text=e...

[2] https://docs.python.org/3/library/stdtypes.html#:~:text=dict...

[3] https://docs.python.org/3/library/stdtypes.html#:~:text=dict...

I found the argument that it’s impossible to test SQL (and some other hard statements in this article) a bit too absolute.

Testing SQL is not impossible – SQL unit testing is a thing – identify a unit of SQL, such as a query, function, or stored procedure - mock the necessary inputs and any state that is modified and check the result of the unit just as you would in any unit testing framework (e.g. arrange, act, assert). Perhaps it’s a little difficult, or different to wrap our heads around (and many do not take the time to see how this can work) but it’s not impossible.

Also, as others have already pointed out, depending on the implementation of the SQL engine itself, different engines can have very different behaviors, some of which make composable SQL possible already today (e.g. table valued functions, CTEs that are inlined, stored procedures, etc).

And for what it’s worth, if you look at a lot of the data processing engines we use today (such as Apache Spark), they really solve the problem of being able to compose and modularize code and SQL together, through data frames or things like language integrated query, integrations of languages that compile down to code that can run and be orchestrated on common execution engines, and so forth.

A different org that is building small intermediate shelters that I think is cool is Pallet: https://palletshelter.com/

I'm not affiliated with them, I just think it's a good idea and they are making some real traction.

Also lark in Python too

May I humbly submit “parsing expression grammars”[1] for your consideration?

Fairly simple and somewhat fun.. Python has PEG parsing built in, but also the pyparsing or parsimonious modules too.

I have built EDI X12 parsers and toy languages with this.

[1] https://en.wikipedia.org/wiki/Parsing_expression_grammar

Thanks, fixed it!

I can go one step further with an anecdote- in the small town I live in, private equity bought our small hospital, and the land it sits on.

Consequently, it changed policy, raided the equipment to take expensive equipment to other hospitals, ran it into the ground, and then closed it permanently on the grounds that it was too expensive to operate.

The hospital sat derelict for probably more than 6 years as the city fought with the owner to try to get a hospital + emergency room back to no avail.

It was then promptly razed to the ground, and now the highly valuable real estate is currently sitting there growing weeds, as the owners contemplate what to do next. We believe they will build luxury senior housing there, but nobody knows.

The closest ER to my house used to be less than 5 miles away. Now it is more than 15 miles to the north, and it is shared by a much larger area. The nearest ER in any other direction is 30+ miles away.

It destroyed the hospital, and our local medical community in the process.

So yeah, at least in my case (and I suspect I'm not alone), private equity was completely destructive to our hospital.

Seeing this news is a nice tribute to Ed Stone, who was one of the core project scientists for Voyager and recently passed [1] (and all those who work/worked on the program).

I had the pleasure of meeting Dr. Stone at a public NASA event many years ago. I asked him, perhaps a silly question: "what does it feel like to know you built the furthest man-made known object in the universe?".

He paused for a moment, after which he responded, with a smile: "Pretty darn good".

RIP, Dr. Stone and go Voyager go!

[1] https://www.jpl.nasa.gov/news/ed-stone-former-director-of-jp...

That's quite interesting, would you be willing to share any sources?

I wouldn't be surprised if the term SQL/MED had been spoken in healthcare/medicine circles (I worked in that industry before). That said, it's the first time I've ever seen anyone connect the term/concept specifically to medicine as the driver for the term.

I'm actually someone who has gone deep down the SQL rabbit hole (I worked for a company that built a data federation platform previously), I'd be interested to read more if you can share.

My apologies, I thought the confusion of the acronym simple came from the "MED" part, which as I mentioned, stands for "management of external data".

In that respect we are in complete agreement :)

EDIT: Perhaps you might be referring to something like this paper about data federation in general? https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&d...

You're right, but this problem doesn't go away when you use Python - because you still have to interact with the type systems of the platforms you are integrating. It's a much more difficult and nuanced problem than I think most people realize.

Even a SQL query can suffer the same fate. Ever tried writing a SQL query against a distributed database that isn’t optimized for that read path?

I think that’s what’s really pointing out the root cause issues here, it’s not purely GraphQL’s problem, it’s the problems inherent to distributed systems.

The MED stands for management of external data, nothing to do with medical data.

It’s actually a chapter in the ISO/ANSI SQL specification.

In traditional security training, you learn the "AAAs" of security - authentication, authorization and accounting (and don't forget the last one!).

It's good to have ways to easily capture the meaning of these words, but permissions and login are implementations that fulfill the requirement of the As, not the As themselves.