HNHacker News
TopNewBestAskShowJobs

endiangroup

111 karma · joined September 7, 2018

website: https://endian.io
submissionscomments
endiangroup··on Why can’t the world’s greatest minds solve the mystery of consciousness? (2015)
thats the interesting part! Can it be done as a team.
endiangroup··on Why can’t the world’s greatest minds solve the mystery of consciousness? (2015)
That's a subprocess and not the thing we started with (that's analogous to me imagining myself... its not the same self that is doing the imagining), the assumption here is consciousness is single threaded.
endiangroup··on Why can’t the world’s greatest minds solve the mystery of consciousness? (2015)
This. You have the set of all sets problem... As soon as some part of you steps aside to inspect the other, it's missing out itself.

In the words of Alan Watts: You can't touch the tip of your finger with the tip of the same finger, you can't bite your own teeth.

It's an irreducible problem. Lest there is some social way we can achieve it...

endiangroup··on Automated hiring software is mistakenly rejecting millions of job candidates
These guys have come up with a solution: https://news.ycombinator.com/item?id=28443610 -- Peer to peer recruiting.
endiangroup··on JavaScript Object Signing and Encryption is a bad standard (2017)
AD: I wrote a library that can deal with this for JWT https://github.com/endiangroup/compandauth

The skiny is that you place a copy of some monotonic counter inside every JWT you issue, you keep track of the counter server side and compare with each request's JWT copy + some delta (which is the equivalent of maximum number of concurrent sessions you wish the user to have).

endiangroup··on Mary Catherine Bateson: Systems Thinker
Thanks for the insight!
endiangroup··on Mary Catherine Bateson: Systems Thinker
I think your intro about covers it, it's a great intro book, not an exhaustive compendium. I certainly found it useful for my mental modelling and world views. YMMV!
endiangroup··on Mary Catherine Bateson: Systems Thinker
I linked to the same article! Isn't a leverage point an attempt to control a system?

Thanks for the additional info! Interesting to hear. I will check out Daniel Schmachtenberger.

endiangroup··on Mary Catherine Bateson: Systems Thinker
AD: Systems Thinking is getting out of date, but in the spirit of knowledge sharing, another fantastic systems thinker is Donella Meadows, heres her most linked piece on leverage points in systems: https://donellameadows.org/archives/leverage-points-places-t.... Her book 'Thinking In Systems' is a fantastic read to open your mind to continuous systems and modelling them, as well as the counter intuitiveness of larger systems.

To expand on 'out of date' - systems thinking is centred around command and control, IIRC it came out of Cybernetics ala Norbert Weiner, Ashby and others. The command and control angle is based on the assumption that with enough information any system can be mapped, predicted and controlled. Whilst not wrong, it excludes another type of system which is truly complex and precludes mapping and prediction by the fact there are either/all too many variables to model or the very act of measuring or acting changes the system or the agents of the system can change the rules of the system.

These systems are deemed by the sense-making community to be Complex Adaptive Systems or Anthro-Complex (specifically for human systems). To approach these systems you have to become comfortable with uncertainty and re-arrange your thoughts on cause-and-effect, which is a whole new world view for many of us in software.

https://cynefin.io/ is great resource, Dave Snowden is a big name in naturalising sense-making and these broader views of complexity.

endiangroup··on Show HN: Redact – Automated deletion for your content on social networks
AD: I can't see any source code? Did you redact it already?
endiangroup··on Write tasks not user stories – Linear Method
I think the insight here is that if you're asking your user you aren't going to get the answers you want. This is why survey writing is so hard, because people can't just explain themselves on the spot.
endiangroup··on Write tasks not user stories – Linear Method
So let's negotiate the true value? How else are you going to know if you're delivering technology thats useful?

You said there are other reasons, what are those reason? I've gone to restaurants to meet friends... and ultimately we all eat to satisfy our hunger.

My point is that tasks are derived from functional requirements, they aren't context free, and if you make them context free then we're just doing what ever we want? To analogise, why do you have relationships in databases and indexes? An index is a flat map to speed up lookups on known queries, you don't get at the meat of your query in the index, the meat of what you want is in the row and it's relationships. You have an object graph for a reason, and yeah it was the hard part, but it's what makes a model useful.

... and to counter the quote... What's the point of your users experience if it's not to get them to where they are deriving value?

endiangroup··on Write tasks not user stories – Linear Method
AD: So this has been downvoted without refute... any takers? Or is it just bots...
endiangroup··on Write tasks not user stories – Linear Method
AD: Agreed! I'd look into BDD as an exploration and specification process (it's not just a testing methodology!), it taps into the goals of your users and features that satisfy those goals. See my comment below for a crude example.
endiangroup··on Write tasks not user stories – Linear Method
+1, without examples it's theoretical none-sense.
endiangroup··on Write tasks not user stories – Linear Method
AD: Sorry you got bogged down in the details. Would you like me to put together a more thorough example? That one I wrote in 2 minutes.

To clarify I picked /some/ value for the restaurant guest, but we could negotiate what the /true/ value is. I go to restaurants when I'm hungry and I'm sure others do, but you're right there are other reasons why you might go, but that doesn't detract from my point being made.

endiangroup··on Write tasks not user stories – Linear Method
AD: I completely agree, value > ego's.
endiangroup··on Write tasks not user stories – Linear Method
AD: I call BS... There is a fundamental difference between functional requirements and the tasks required to implement them. Note there are NO examples in that article, and I bet you it's because it was hard to think of context free tasks... because context is everything and thats what your functional requirements provide your tasks.

User stories capture functional requirements:

```

As a restaurant guest

I want to consume food

Because I am hungry

Scenario: Guest is provided with menu when they are sat at their table

  Given there is a free table

  When I am sat at my table

  Then I should be given a menu
```

This is BDD style, I don't really care what you think of BDD it covers what I need to make the point... First of all context is king, there should be NO scenarios here that do not satisfy the guest being hungry, thats scope creep. Equally there is NOTHING here about HOW you implement this, it does not mention who gives the menu, what the menu contains, how it looks or anything about the rest of the restaurant. Why is that the case? Because we are at a level of abstraction that is talking about delivering value, not about the quality of what we deliver. You can only deliver value to users so we can only talk about users and their experiences. Now as for tasks, they are derived from these functional requirements, they connect what currently IS to what needs to be, e.g. if we currently give guests a menu when they enter the door, thats what IS, now we need to go about creating a task to change that to when the guest sits down, thats the task and it is a poignant task because it's derived from our functional requirements, there is no room for scope creep here.

Please do not write context free tasks, you'll run yourself into the ground in scope creep, because you're detached from what matters. That's why user stories are important... but most people don't write user stories correctly, they treat them like a task template and forget the value delivering aspect and all the context that goes with it.

endiangroup··on How Do You Know When Society Is About to Fall Apart?
AD: When we all say it is, followed by the actions that prove it.
endiangroup··on Launch HN: SuperTokens (YC S20) – Securely manage session tokens
Yes absolutely, as I stated in the comment before. The super fast verification is once you've done the DB lookup and do the actual comparing (as it's just int64 comparisons). CAA is intended as a set of primitives to be built on top of.

As I understand ST is in effect making the client act as the cache by giving them a refresh token (RT), you delay having to do a central DB lookup until the AT expires and RT is used... You could do the same with CAA, check the DB every X interval (30s) instead of every request, the security implications would be the same (some period of time that no matter who is using the token it will work) as I understand?

From what I've read Theft Protection is a matter of detecting reuse of RT (as I've understood), again with CAA and using the above, you could track each periodic "check" against the master with another counter and embed into the token, if you see a tracking number < or > what you've seen then a Theft has occurred. Would that be the same as ST?

endiangroup··on Launch HN: SuperTokens (YC S20) – Securely manage session tokens
Good luck with the audit, the solution sounds interesting on the surface!

Ah I think you've slightly misunderstood, there is a 'counter' per entity, each user has their own separate counter so when you are revoking it is only relevant to sessions related to that users account. This would fall under your database lookup version of JWT management, it's centralised but in the smallest way possible.

That might clear up your second question, you'd want to revoke all sessions like I'd want to 'Log out all other active sessions', logging in to my Spotify on my friends computer last week, and my sisters the week before, I want to log them ALL out.

endiangroup··on Launch HN: SuperTokens (YC S20) – Securely manage session tokens
AD: I authored a solution to the revocation of tokens[1] for JWT and even verified the solution through TLA+. It uses a monotonic counter as a claim per JWT, you can revoke X active sessions behind the last issued, you can also use time to do temporal permission inflating (sudo... etc).

What sort of verification have you done on your solutions? What are the down falls of your solution?

[1] - https://github.com/endiangroup/compandauth

endiangroup··on Plectics - The study of simplicity and complexity (1996)
AD: Its one of those areas I don't see come up AT ALL on HN, complexity theory [0] imho is a superset to all software engineering practises, because people are involved and not just because the things we are trying to solve are 'complex'.

A great place to start is Dave Snowden's Cynefin sense-making framework [1][2]. Study it and embed it into your core thinking, it can underpin how you approach almost every problem you come across in life. I was fortunate enough to attend a masterclass with Dave Snowden and Simon Wardley (Wardley maps, another thing worth understanding and practising related to strategy, he's published the whole thing for free on Medium [3]). What was fascinating and part of the result of the 2 day class is that Simon Wardley had been using a Cynefin like model in his head whilst making decisions and when he came up with his mapping method, additionally they had both been inspired by Max Boisot [4].

[0] https://en.wikipedia.org/wiki/Complexity_theory_and_organiza... [1] https://en.wikipedia.org/wiki/Cynefin_framework [2] http://cognitive-edge.com/videos/cynefin-framework-introduct... [3] https://medium.com/wardleymaps/on-being-lost-2ef5f05eb1ec [4] https://en.wikipedia.org/wiki/I-Space_(conceptual_framework)

endiangroup··on Show HN: CAA – Lightweight distributed session management [Go / TLA+]
fortunately there is a TLA+ spec so you can rewrite the system in any language you like!
endiangroup··on The Tragedy of Given-When-Then
Act 1 - G/W/T can be used to express iteratively aspects of an algorithm such that you can derive it without knowing or fully understanding it (algorithm triangulation: think GPS where each satellite is a constraint and you derive through iteration of each passing scenario the general algorithm).

Act 2 - is really about process rather than G/W/T (which is really just AAA, arrange, act & assert).

Act 3 - again process, G/W/T isn't the solution to domain understanding, conversations are. You should be continually communicating to between all stake holders to maintain a current domain understanding.

We wrote an article recently on the limits of BDD [1], G/W/T didn't really come up, there are other more glaring issues with BDD when it comes to systems that intersect mismatched understandings of the real world between experts and users. Unrealistic wants and goals are killer. Additionally we started writing a tool to attach metadata to scenarios (G/W/T) so you can capture technical details about things called SpecStack [2]

[1] https://endian.io/articles/limits-of-bdd/ [2] https://github.com/endiangroup/specstack

endiangroup··on The Tragedy of Given-When-Then
Thats where things like Domain Driven Design are going, and really it makes a lot of sense. Why express and model your domain in terms outside of it? Its like forever living through a translator instead of just learning the language in the first instance, you add more points of potential failure.
endiangroup··on Learn TLA+ (2018)
It uncovered an oversight in an algorithm we put together for managing distributed sessions. An invariant showed that our algorithm would accept and consider valid tokens that were ahead of the 'master' counter, which theoretically shouldn't of been a possible input in practice but now that we know we've added an additional guard. Original comment: https://news.ycombinator.com/item?id=19662123
endiangroup··on Learn TLA+ (2018)
We used this resource to put together a model for our distributed sessions management algorithm, heres a GitHub with Go implementation and PlusCal / TLA+ model [1]. We also wrote a more in depth article about it here [2] (with an interactive version in JS). It took a few attempts but eventually it clicked!

[1] https://github.com/endiangroup/compandauth [2] https://endian.io/articles/compandauth/

endiangroup··on Show HN: Five Integral C-Words of Maximally Useful Product Specification (2019)
We're the team that put this paper together, happy to answer any questions!
endiangroup··on How to Use JSON Web Tokens
A small Go library we wrote to centrally manage distributed session tokens such as JWT inspired by CAS: https://github.com/endiangroup/compandauth. Main highlight is central revocation, locking and unlocking of distributed sessions

The core concepts can be translated into any language that supports integers, happy to reference any alternative implementations.

← PreviousPage 2 of 2