And an admittedly uncharitable TLDR on the response is: "yeah... but most users just ask one thing and barely use the product so they never need the cache. Also trust me bro".

Which sure, fine. I'm willing to bet is technically true. I'd also bet those users never previously came close to hitting their session limits given their usage because their usage is so low. But now people who were previously considered low to moderate users are hitting limits within minutes.

They may as well have just said "we've looked at the data and we're happy with this change because it's a performance improvement for people we make the most margin on. Sucks to be you".

Anyway this wasn't meant to be a debate about rates. Just that "nobody that charges at home would want this" was an overly reductive claim.

* Soink: Not OSS

* Hitoku: Not OSS

* Fluid: Parakeet + Whisper again. Also Apple Speech which is even worse than all of the other options

I'm not going to dissect every example of why this absolutely is not an already solved problem.

Also our electricity rates fluctuate based on the underlying wholesale rate. It's going to be clear and sunny tomorrow at midday. Sure would be nice to be able to set my car to charge at midday when the price is single digits cents per kw, or maybe even negative. Instead I'll just have to drip it in with the higher rates at midnight-6am and know tomorrows cheap rates will average out to a much lower cost.

TLDR: definitely useful even for people who charge at home.

I'm a daily user of Wispr Flow, in part because all previous attempts to try and reimplement it myself have been terrible. Tortoise and all of the other suggestions people offer up have terrible performance in comparison.

Comments URL: https://news.ycombinator.com/item?id=47434820

Points: 1

# Comments: 4

I've unfortunately had a number of emergency visits over the past few years. I'm a bit torn on the public vs private situation. For certain classes of issue (e.g., broken bones) my friends who work in hospitals have repeatedly said they'd go public purely because the volume of patients those surgeons have to treat daily means the teams in the public systems are typically incredibly experienced. And yet, I smashed my hand to pieces mountain biking on a holiday weekend and when I arrived at the ER the place was absolutely rammed and it was going to be a many hours wait to even get triaged. We got straight back in the car and drive to the private ER 5 minutes down the road and were seen immediately.

In the moment I was incredibly appreciative of that option. It does make me feel uncomfortable that it was only an option because I could afford to pay to jump the queue though.

So much of the value was already delivered in that simple `git push heroku master` which gave you a container + load balancer + a database. The vast majority of people didn’t need more. And of those that were left that did far too few of them were willing to suddenly start paying $32/mo per dyno (you just gave me one for free! I only want one more!) or make the jump to multiple hundreds of dollars for a database.

Read any of the threads about Heroku over the years. The biggest complaint is always “it’s too expensive”. Even when a large percentage of what was on people’s bills were add-ons like databases, new relic, redis, logging, etc (i.e., not Heroku).

At the time it felt to me they were missing out on so many other useful resources. Maybe I was wrong. It's interesting to see things trending in a similar direction now.

I asked someone a similar question to you a year ago and they told me something like "just spend 15 minutes with it. Aggressively swipe past things you aren't enjoying, like the the things you like. Search for something you are interested in too and like anything you like there". My feed is currently entirely basketball coaching tips for kids, cooking & recipes, stand up comedy, basic DIY, fitness/running tips, local restaurant recommendations, and sports highlights.

The few annoyances we've had LR have resolved for us at zero cost, even when we were out of warranty.

Obviously that's not the optimal approach for every use case, but there's a lot where IMO it was better. In particular I was hoping to spend more time exploring it in an enterprise context where you've got complicated sharing and permission models to take into consideration. If you have agents simply passing through the permission of the user executing the search whatever you get back is automatically constrained to only the things they had access to in that moment. As opposed to other approaches where you're storing a representation of data in one place, and then trying to work out the intersection of permissions from one of more other systems, and sanitise the results on the way out. Always seemed messy and fraught with problems and the risk of leaking something you shouldn't.