HNHacker News
TopNewBestAskShowJobs

paol

2,980 karma · joined August 10, 2010

submissionscomments
paol··on Show HN: IMDb SQL Best Movie Finder
Seems to be some normalization problem with the data, right in the 1st page of the default query there's a duplicate entry.
paol··on Apple's M4 Max chip is the fastest single-core performer in consumer computing
Well we're about to find out now that CDPR have announced Cyberpunk 2077 will get a native Metal port. I for one am extremely curious with the result. Apple have made very lofty claims about their GPU performance, but without any high-end games running natively, it's been hard to evaluate those claims.

That said, expectations should be kept at a realistic level. Even if the M4 has the fastest embedded GPU (it probably does), it's still an embedded GPU. They aren't going to be topping any absolute performance charts.

paol··on INNER JOIN ON vs WHERE clause
No, they are the same according to SQL semantics.

In other words it's not just that this particular case returns the same result, but that in all cases the result must be the same.

paol··on Why does ++[[]][+[]]+[+[]] return the string "10"?
This would have made a good addition to the (in)famous WAT talk:

https://www.destroyallsoftware.com/talks/wat

paol··on Why are my ZFS disks so noisy?
I know the article is just using the noise thing as an excuse to deep dive into zfs and proxmox, which is cool, but if what you really care about is reducing noise I thought I'd leave some practical advice here:

1. Most hard drive noise is caused by mechanical vibrations being transmitted to the chassis the drive is mounted on.

2. Consequently the most effective way to reduce noise is to reduce the mechanical coupling in the drive mounting mechanism. Having the drives in a noise-isolating case is helpful too, but only as a secondary improvement. Optimizing the drive mounting should really be the first priority.

3. If space isn't a concern the optimal thing is to have a large case (like an ATX or larger) with a large number of HDD bays. The mounting should use soft rubber or silicon grommets. Some mounting systems can work with just the grommets, but systems that use screws are ok too as long as the screw couples to the grommet not the chassis. In a good case like this any number of hard drives can be made essentially inaudible.

4. If space is a concern, a special purpose "NAS like" case (example: the Jonsbo N line of cases) can approach the size of consumer NAS boxes. The lack of space makes optimal accoustics difficult, but it will still be a 10x improvement over typical consumers NASes.

5. Lastly what you shouldn't ever do is get one of those consumers NAS boxes. They are made with no concern for noise at all, and manufacturing cheapness constraints tend to make them literally pessimal at it. I had a QNAP I got rid of that couldn't have been more effective at amplifying drive noise if it had been designed for that on purpose.

paol··on Damn Vulnerable UEFI
I think I've ranted about this in HN before, but UEFI is so pointlessly complex. Modern OSes don't rely on the BIOS for basically anything. When choosing the successor to the old PC BIOS we could have made something extremely minimal.

Instead UEFI goes hard in the opposite direction. It's basically a full blown operating system with all the attendant complexity and, unsurprisingly, vulnerability surface.

paol··on The effect of CRTs on pixel art
An article to the same effect, but with more example images: https://wackoid.com/game/10-pictures-that-show-why-crt-tvs-a...

There are several instances where you can clearly see the artists not merely "living with" the CRT rendering but actually using it to their advantage: the eyes resulting from a single red pixel in the Castlevania picture, the smooth gradients in the Streets of Rage picture arising out of just 3 or 4 colors...

paol··on The Typeset of Wall·E (2018)
I second the book recommendation. It's great.
paol··on OpenDNS Suspends Service in France Due to Canal+ Piracy Blocking Order
Of the 3 I wouldn't have expected Cisco to be the first to take a stand. Well done them.

DNS blocking bullshit is especially bad in Portugal because the law is set up in a way that has copyright interest groups just send their block lists directly to ISPs without any kind of judicial process or review whatsoever. It's common knowledge here to change your DNS provider to bypass the ISPs for this reason.

paol··on Are animals conscious? New research
If you enjoy fiction I urge you to read Blindsight by Peter Watts. One of the central concepts in the plot is "what if consciousness were unnecessary for intelligence"? Along the way it explores all sorts of angles on consciousness and cognition.

Still ranks as one of the best SF works I've read this century.

paol··on Intel Brings a Big Fork to a Server CPU Knife Fight
> Hyperthreading is turned off on m7a/c7a/r7a.

You made me go check and you're right! Thanks for the correction.

paol··on Intel Brings a Big Fork to a Server CPU Knife Fight
You're slightly off in your understanding of VCPUs. All AWS instances "get" VCPUs. A VCPU is a single thread in architectures with SMT and a single core in architectures without. So for example c7a instances which are AMD EPYCs have 1 VCPU = 1/2 core.

They have is a page that lists this for every single instance type which I can't be bothered to search for right now.

As for spot instances, they're a great if you can use them, but you can't hold their price as being representative. Now using savings plan prices as representative is defensible, but it won't bring the cost below the "hundreds per month" order of magnitude.

paol··on Intel Brings a Big Fork to a Server CPU Knife Fight
You're missing my point. Servers are very expensive and are sized to task. If your NICs (or whatever) bottleneck the server with 32 cores worth of load, then you buy a server with 32 cores, not one with 64 cores.
paol··on Intel Brings a Big Fork to a Server CPU Knife Fight
Not in AWS. 1 VCPU = 1 thread (so half a core in architectures with 2-way SMT, which is most of them).
paol··on Intel Brings a Big Fork to a Server CPU Knife Fight
You may be right in general about my take being too cynical, but I have to counter this in particular:

> and typically they stay idle while waiting for IO

That's just not how you size servers. If most of your cores are idle you will add more work to the server, or you will have bough one with fewer cores to begin with.

paol··on Intel Brings a Big Fork to a Server CPU Knife Fight
But the decision to make server chips entirely of E cores must come from some customers demanding core count density above all else. I'm suggesting where that demand comes from.
paol··on Intel Brings a Big Fork to a Server CPU Knife Fight
Funny enough c7a is just what we chose recently for our CI runners precisely because of the better CPU performance. It's only better relatively to the terribleness of the t* types though.

And it definitely costs hundreds of dollars per month. Just ran the calculator and 8 VPCUs (to keep the rough laptop analogy) comes out at $316.

paol··on Intel Brings a Big Fork to a Server CPU Knife Fight
I have a pet theory that the recent focus on CPUs with a high count of low-perf cores (this article is about intel but AMD has an equivalent offering) is being driven by hyperscalers like Amazon and the perverse incentive caused by their pricing structures.

Cloud providers have structured their compute offerings around dollars per core. This creates the incentive to deploy the densest possible core count processors, and zero incentive to make those cores faster.

It can't have escaped people's notice how slow cloud VCPUs are. My 6 year old crappy laptop CPU has the same single core perf as a cloud instance that costs hundreds of dollars per month. And the frustration is not just cost, when you actually need good performance for something that doesn't scale so well horizontally, it's hard to get it.

paol··on Firefox bug gets fixed after 25 years
Ha! I was subscribed to this bug for most of those 25 years. Note that Firefox didn't exist at the time, this was filed against Netscape Navigator.

Don't remember why I was subscribed to the bug, must have commented on it or something. Got the occasional email notification about it over the years, always got a chuckle out of it. Then a couple of days ago, lo and behold, it was fixed!

paol··on Radxa's SATA HAT makes compact Pi 5 NAS
> Power. You need quite a solid PSU to withstand the power-up of a couple multi-TB drives.

This is overstated. You need a lot more than "a couple" of drives to trouble even the most modest ATX power supply. Now if you were trying to run this off the RPi's power I would agree. But since this requires a separate PSU anyway it's a non-issue.

paol··on Intel Brags of $152B in Stock Buybacks. Why Does It Need an $8B Subsidy?
It doesn't need it, it's simply in a position where it can get it. From their perspective it's free money.

It's the same situation with car manufacturers, they do this all the time. They get massive subsidies (usually in the form of permanent tax breaks) by pitting the governments of prospective factory sites against each other.

(It's actually a form of the prisoner's dillema: governments would benefit in the net if no-one offered these subisdies, but the prize for defecting is just too great so someone always does.)

paol··on WebSockets vs. Server-Sent-Events vs. Long-Polling vs. WebRTC vs. WebTransport
Oooh boy you touched a pet peeve. I mean who needs authentication on the modern Web right? /s

The even more irritating thing is that there is nothing preventing this, and every server I've tried supports it. It's only the browser WebSocket API that was designed without this. Cookies are the only thing browsers will deign to send in the initial request.

paol··on The Getty makes nearly 88k art images free to use
I can't either. wget reports "Read error at byte 7843894" every time.
paol··on Bruno: Fast and Git-friendly open-source API client (Postman alternative)
Sounds like mockserver[0] might be what you want.

[0] https://www.mock-server.com/

paol··on Yuzu emulator developers settle Nintendo lawsuit, pay $2.4M in damages
I don't think it's that simple. For one thing emulation has been ruled legal in the US. By all reports Nintendo had an unusually strong case against Yuzu because the developers behind it didn't sufficiently distance themselves from piracy uses - the opposite if anything.

On the practical side, the fact that the developers were a for profit company also made Nintendo's job much easier.

Taking down a more legally careful, decentralized open source project should prove harder.

paol··on Yuzu emulator developers settle Nintendo lawsuit, pay $2.4M in damages
It's open source. If there are people interested in continuing development it won't go away. And even if there aren't the current version will still work fine, just maybe not for future games.
paol··on Incorrect Stack Exchange answers make me angry
I think the initial design principles of Stack Overflow - basically that the balance of incentives will determine the outcome - is still sound. And the initial years of the site showed the people in control were very smart and thoughtful about structuring and fine tuning the incentives.

The problem is that fine tuning (and occasional restructuring) needs to be an ongoing process, and it has stopped. It feels like the original creators left and the site has been coasting on auto-pilot for a few years now.

That's what worries me about the future of SO. It was an enormous improvement over what came before (in fact was created as a reaction to the horrible status quo - expertsexchange anyone?). But if it's left to continue to wither it will eventually become like those sites.

paol··on Berlin TYPE: The official type for the city of Berlin (2020)
Does it have an ampelmann glyph? If not I call fake.
paol··on I'm confused: what's with the project descriptions at HTTPS://pkgx.dev/pkgs/?
You don't have to wait it's already here.
paol··on How to make a better default Firefox UI
The Lepton theme looks good. If I'd known about it I may not have bothered to cook up my own userchrome hacks. I'm pretty happy with the current results[0] though so I'll be sticking with them for the time being.

[0] https://draic.org/scratch/ffoxuserchrome.png

← PreviousPage 2 of 13Next →