Hacker Newsnew | past | comments | ask | show | jobs | submit | tinkertamper's commentslogin

I'm curious about what you mean here. It sounds like you're saying that applications shouldn't concern themselves with consistency. Can you elaborate?


Love this!

Call me lame, but are there any open-source projects that accomplish this type of gui in typescript? I have an idea but it uses something written in JavaScript.


You can use any javascript library in typescript, and TS is all JS at runtime anyway. You can even add types without touching the JS library, just write a .d.ts file and stick it anywhere that TS looks for sources.


I created a sub 9 years ago that now has a few million subscribers (so now requires a bit more effort to mod). I _don’t_ get paid, though I certainty wouldn’t turn it down. I started the sub because I like the content that is posted and I want to continue seeing more of it. Why continue to mod? Because I like to be involved and keep this little square of the internet free of toxic and often hateful spam/content.

> It seems threatening to take it away is making them compromise on their values quite effectively.

What exactly are you referring to here? I see this narrative in a lot of places and I’m confused about what “values” mods supposedly collectively have? So much discussion online paints moderators as some sort of a connected cabal with alternative motives instead of unrelated nerds who enjoy niche content.


> What exactly are you referring to here? I see this narrative in a lot of places and I’m confused about what “values” mods supposedly collectively have?

Taking communities private was a stand for values (or shown as such). Now that's being reversed when these mods are threatened with removal. So there seems to be some significant value in modding they don't want to lose.

> So much discussion online paints moderators as some sort of a connected cabal with alternative motives instead of unrelated nerds who enjoy niche content.

I wouldn't describe it as a cabal (some secret, influential, political org). But these strikes have been coordinated half-publicly[0] and half-privately among mods of subs. And the goal is to influence Reddit in a way that the group prefers. In a political context, that would be close to a cabal. Regardless, I don't think cabals are immoral unless their goals are.

[0] https://www.reddit.com/r/ModCoord/


> So much discussion online paints moderators as some sort of a connected cabal

significant group indeed became well organized in aggressively pushing one sided narrative, it is possible you are not a member of that group.


> Because I like to be involved and keep this little square of the internet free of toxic and often hateful spam/content

So most of the personal value comes from being able to influence this part of the internet in a positive way?


I was hoping this was about my coworkers tuning on our cloudwatch alarms until pagerduty stops calling


That’s not really a fair comparison here — A more fitting metaphor would be whether or not you could intervene on assembly code not knowing assembly.


This is just what was just called “speaking tech” back in the early 2000. I see a lot of discussion online about “prompt engineers” or “ai whispers” as you put, but I think the reality is that these tools are just another UI paradigm that our current generation will (on the whole) struggle with and the next generation will find intuitive.


Well, to be fair there is a quantum leap: an ability to materialize abstract thought presented as a text. This will end a lot of jobs where anyone can judge the result or certain error rate is tolerable. But while we have a lot of jobs like that, way to many are not like that and they will not be replaced. But i like the idea of a new type of UI (UX?) being born in front of us.


An IDE that provides a linear view of your code base. Follow code paths as though they were all inline, e.g. expand views of function/method internals inside the calling function, etc. No more changing files/modules.


forever now I've been imagining a codebase as database, functions/expressions are entries and the IDE can query based on context, eg. filter by regex or other rules and filter differently when debug context available etc.

it should be transparent once you flatten the db to file(s) and run normally, but for IDE would need to be aware of this db structure. maybe a VS code extension might work!

that is the dream...


How is that not describing just a regular IDE like Intellij? The reason it has an indexing phase when you open a new project is because the IDE is building a database of your code base. You can then do many kinds of query against it.


I'm not aware of it. any examples/links?


https://www.jetbrains.com/idea/

They make IDEs for many languages but the indexability of languages varies a lot depending on (mostly) how strong their type system is and how easily parsed the code is. Check out their structured search feature for an example of how to do semantic queries over the indexed AST.


Before you wax poetic about this period, please read more on the complete disregard for the families and communities impacted by the large construction projects from that time. Replies comparing the US today to China would also benefit from the same. Public works has a lot of issues with cost and inefficiencies, but I think we should all be happy that eminent domain is not wielded with the same disregard for human life as it once was.


Interestingly, the Golden Gate Bridge pioneered worker safety measures while the Bay Bridge (built at the same time) followed the status quo of almost no consideration to safety.

The difference was, basically, how much management cared about the issue.


I used to have a boss who’d ask me to rate my company satisfaction with the company 1-10 at every 1:1. This always struck me as one of the laziest/misguided management moves I’ve encountered, and I’m sure he felt it was both accurate and clever. If you manage people, and don’t understand that basic power dynamics will always trump encouragements for “openness”, you are naive at best and willfully blind at worst.


One of things I always looked out for as manager (which I am not currently) was a change in feed-back I got. If it turned from open, direct and honest to tounge-in-cheek, carefully worded and limited, punctuated with various versions of "nothing to report" I took it as a sign something was wrong.

Obviously, this effect is much easier when you are not in a management position. Because of that I was usually borderline paranoid about these changes in 1:1s with my directs.

In small teams I found the way to have informal, spontaneous 1:1s very effective. The basis, so, always was regularly scheduled ones unless you just forget to have them.

EDIT: For sure every 1:1 is different, and every 1:1 with different people need to be run differently. Some people like to discuss private stuff, others want in-depth tech discussions. Sometimes 1:1s are over in 5 minutes, sometimes they take an hour. Be flexible, and never use anything said in a 1:1 "against" the other person.


Exactly this. As an employee, in this situation, you are always one step away from saying something to the person who can make your life miserable if they take offence. At no point am I ever open in these things, always guarded and watchful of what I'm saying because of who I'm saying it to. It's a corporate pipe dream to think otherwise.


Heard one of the Warby Parker founders speak once, and he mentioned a similar process at their company. I believe once a week all employees anonymously rated their satisfaction 1-10. He was very bought into the thought that "happy employees = productive employees". He said the weekly ratings was a very powerful tool for the executive team to get a feel for how the team was feeling, and if they could push harder or needed to ease up. Compared with the alternative of waiting to see resignation numbers go up, seemed like a pretty brilliant idea to me.

This was many years ago so apologies if some details are a bit off, but the gist of the story has stuck with me over the years.


A key difference though is that was anonymous (safe) and aggregated (large sample size).

Giving a 1-10 rating directly to your supervisor is a completely different thing and the number is tainted by all sorts of other dynamics.


Not to defend the constant ratings, but imagine managers are not in fact all stupid for a second, and that they realize the power dynamic is there. You still need to do the same job and fulfill the expectations of your own manager, as a manager. It's easy to assume stupidity, and I'm sure there's plenty dumb moves made, but what you find cringy another report finds cool and easy. I bet there are people who would love a no fluff rating and be done vs a bullshit 10 mins of talk where in the end the manager still makes a mental note of your face next to a 7/10. Anyone who manages people sees completely different reactions from different people to exactly the same thing, so that isn't surprising. But sometimes engineers adopt a cynical view of it all where because you don't like something it must mean nobody does.


Very few people can be fully open with their Managers; it is inbuilt into the hierarchy and nothing to do with stupidity. It is the Manager's job to be aware of it and then devise techniques for getting around it and/or make allowances for it.


100%, this is exactly where my head is at too.


> imagine managers are not in fact all stupid for a second

I don’t really see where the commenter implied this.

To support their stance, I have also had lots of managers at various-sized companies (FAANG incl), who didn’t understand the inherent 1:1 power dynamic. They would expect full honesty, while covering up anything above my pay-grade.


"Stupid" is a pretty accurate description for someone who thinks something lazy and misguided is accurate and clever.


I don't know. It would be bad to have that replace other real conversations, but i could see it being easier for some people to just answer a 6 instead of a 7 instead of the IC explicitly bringing up out of nowhere that they are slightly less happy thus week than last week for unclear reasons that they themselves don't understand.


> basic power dynamics

power dynamics vary. some developers can always easily get another job, so they have a lot of power. in a case like that, asking the developer to rate their satisfaction seems normal & good, like a customer satisfaction survey.


There’s also a playground to play with modeling and typing here: https://electrodb.fun

Modeling, and actually implementing, single table design for DynamoDB has a lot of pitfalls and boilerplate. I wanted to make it easy and fast to define models and implement collections of models to get more out of a single query without scans. Having your entities dynamically infer typing from the model (without having to manually generate typedefs) is also huge win for usability and dx.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: