HN user
Perhaps the fact that 99+% of today's workloads could be running on the client if it were as easy as shipping Rayforce and the data directly to the client.
Besides that, pure C that you can embed into your app is much easier to deploy for some (and likely 100x more performant) than stuff that comes via Helm chart [cries in JVM 'big'-data solutions]
I've been experimenting in this space, where might I find a guide for what to build that would be useful to you? I suspect most existing approaches are an order of magnitude slower and harder to use than they need to be.
Did the decades of Oxy which caused this epidemic not count as safe and legal access? I suspect we've tried allowing opiates more than once or twice in the millenia we've been here.
What would you plug this FPGA into? Where will you trade? How much more will you make thanks to the FPGA's lower latency?
Magna Carta: 1215 Boolean logic: 1847
Email your audiologist and ask for your audiogram. Should be straightforward to apply an equalizer to match it, esp if can find an equalizer with db(a? m?) labels
Least equipped to provide? They've been working on machine learning algos for exactly this purpose for twenty years.
The best way to speed up is to practice. Alternative layouts are a waste of time unless you're trying to fix an ergonomic / pain issue, and even then the board itself and your technique and lifestyle are more important. I type faster than any Dvorak user I've met and I did nothing make that happen except be impatient. But typing is basically over with now that Whisper is a single button press away.
My m4 macbook had a weird flashing external monitor issue. One that eventually led to my monitor appearing to break. But have no fear, it's a known problem since m1 times and not a priority to fix.
If you read the gnokepitch in the first comment it makes a lot more sense.
You should mention that it offers phone/tablet viewing in the manufacturers pitch, that's awesome over a custom c++ thing. I don't know it they'll care about the bundle size, if you're worried about visual space / where to put it.
I like the idea, and am excited to see an experimental implementation. You will have to ignore many haters who don't realize that Excel is the most popular programming language in the world. "Stop writing dead programs."
Is this why MacBooks have had weird flickering issues with some monitors since the M1 with no end in sight?
As a Samsung user, I couldn't tell ya. I've been clicking cancel when it popped up weekly for two Samsung phones in a row.
Is the video not one long science experiment? With you confirming your one textile experiment resulted in a tear?
Also accepting gmaps work, if only it could preemptively cache the return trip for any trip longer than an hour, so that I'm not stuck with no service trying to remember how I got there.
If only folks wrote software with this in mind.
Have you worked in the industry? Would love to help ya with all of the "easy" infra parts that you're talking about, but indeed that secret sauce that decides what to actually trade is the part I can offer zero help in :-)
They might not be saying it's perfect in all cases, maybe just giving the line of thinking that landed us at Katy highway :-)
Tove2d editor demo?
Civil design software?
Have you considered hearing aids for the tinnitus?
Which hearing aid company?
If you believe adding it up linearly is valid, and you care about the results, you should be using more efficient storage software.
Besides memory use I'm not sure why his approach isn't easy for everything you suggested.
Want to group undo actions? It's pretty easy to define arbitrary grouping algorithms on the sequence of recent state snapshots, which yields...a smaller sequence of state snapshots :-) a nice composeable idea.
To show redo after an undo, store recent undo triggers in the state, and if there is undo in current state, show redo with a list of older mem states to jump to. This repr also makes it easy to wipe out redos when doing a redo-breaking operation like editing the doc.
Memory use can indeed become a problem. But if/when it does, there is probably some obvious trait of the sequence of recent states that makes it easy to compress. Like for example that it's 99% the same contents, after a small edit. Or that one edit type dominates. Or that folks don't often care if they can't Ctrl+Z through 8GB of typing.
I think most programmers would gain by not dismissing simple approaches out of hand as too simple.
If writing it off for memory, write out the math for the size of the state and the allotted memory. This is not the era of computers that our languages were invented in; the page size your CPU fetches from insanely fast SSDs can be 4kb. And it can do a lot of that in a blink. Trying to shift runtime work (that the computer does) back to compile/design time (that the dev does) can cripple a system's performance in the short run (for coding/delivery time) and long run (for performance after pointer chasing)...luckily most of the time it doesn't matter anyway :-)
Anything can fail, at any time. The best we can do is mitigate it and estimate bounds for how likely it is to mess up. Sometimes those bounds are acceptable.
This is where s2.dev could in theory come to the rescue. Able to keep up with the streaming bandwidth, but durable.
Sqlite can be run in process. Latency and bandwidth can be made 10x worst by process context switching alone. Plus being able to get away with n+1s could save a lot of dev time depending on the crew, before Claude (tho the dev still needs to learn that the speed problem is due to this and refactor the query, or write it fast the first time)
Why spend Anthropic developer time on charging you less when an MVP impl can charge you $1 for indexing a simple small Helm chart?