Hacker News: daneel_w

New comment by daneel_w in "A 25-year-old video patent just expired, ending a legal headache for Linux"

daneel_w — Wed, 19 Aug 2026 17:52:14 +0000

It was absolutely a good enough solution compared to the long wait for DVD releases and shelling out tons of money.

New comment by daneel_w in "A 25-year-old video patent just expired, ending a legal headache for Linux"

daneel_w — Wed, 19 Aug 2026 17:36:20 +0000

Agreed!

New comment by daneel_w in "A 25-year-old video patent just expired, ending a legal headache for Linux"

daneel_w — Wed, 19 Aug 2026 17:35:22 +0000

> There's no such thing as a codec that looks good or bad in an absolute sense.

There are plenty of examples of both awful codecs (software; encoders) and of digital audio/video standards (e.g. Vorbis and MPEG-4 ASP) suffering from material limitations.

> If you gave DVD levels of bits-per-pixel to MPEG-4 ASP you could get something that looks nearly perfect.

No. Not even if the source is encoded by XviD. Not all problems can be solved by throwing higher bitrate at it. The ASP only supports 4:2:0 chroma subsampling, which is the largest of several shortcomings contributing to the limited quality you can achieve with ASP video material.

People on at least an intermediate knowledge level of digital video know what a damning problem 4:2:0 subsampling is and how heavy a penalty it incurs on color and clarity. Your comments mostly hold a suggestion that you kinda have some rough idea about digital video. And that's OK.

MPEG-4 SSTP is a very different matter, but that's not what DivX and XviD encodes.

New comment by daneel_w in "A 25-year-old video patent just expired, ending a legal headache for Linux"

daneel_w — Wed, 19 Aug 2026 17:15:50 +0000

If anything it would be an issue with the encoder, but it's not. Both DivX and XviD are prime pieces of software but they live within unbridgeable limitations of the MPEG-4 ASP profile. This or that individual may not care about the outcome, and they may use various post-processing to smooth out the creases, but the limitations are all the same real and inherently visible.

New comment by daneel_w in "A 25-year-old video patent just expired, ending a legal headache for Linux"

daneel_w — Wed, 19 Aug 2026 11:58:01 +0000

DVD Video uses the much older MPEG-2 standard, not MPEG-4.

New comment by daneel_w in "A 25-year-old video patent just expired, ending a legal headache for Linux"

daneel_w — Wed, 19 Aug 2026 11:50:27 +0000

With emphasis on shit. Really. I can relate to the rose-tinted memories, but DivX and XviD both encode to MPEG-4 ASP which has a lot of shortcomings that manifestly result in the infamous sludgy green-purple banding artefacts of the time, no matter how good the chosen encoder is. XviD is the better of those two MPEG-4 ASP encoders, but it simply does not look good in an absolute sense, just in a relative sense compared to the alternatives of the era.

New comment by daneel_w in "Deus Ex creator Warren Spector is retiring from game development"

daneel_w — Tue, 18 Aug 2026 16:40:45 +0000

Thank you for an absolutely fantastic game (most notably the first part) and also thank you to Michiel for an equally fantastic soundtrack.

New comment by daneel_w in "France to ban unsolicited telemarketing calls"

daneel_w — Fri, 14 Aug 2026 17:08:46 +0000

The ones to really bear the consequences are legitimate businesses and the customers of these businesses. It sounds like your desire to see offenders punished, instead of the problem actually being solved with the help of functional regulations, makes you incapable of seeing the wider picture.

New comment by daneel_w in "France to ban unsolicited telemarketing calls"

daneel_w — Tue, 11 Aug 2026 12:06:17 +0000

Sweden has such a system*, and a couple of related laws, since the early 2000s. In my personal experience it works very well here, and I've never heard friends or acquaintances complain either about getting unsolicited calls despite having signed up on the list. An identical system would of course not perform equally in any other country, depending on local "telephone culture" and general moralities.

* https://nixtelefon.org/

New comment by daneel_w in "France to ban unsolicited telemarketing calls"

daneel_w — Tue, 11 Aug 2026 10:46:28 +0000

No, that's a bad idea. There's no hurdle or legal issue, and money isn't behind it either. Those "prefixes" (area codes) are all the same segments of this or that national numbering plan that are also allocated by legitimate businesses both for placing legitimate outbound calls to their customers and for numbers that connect e.g. switchboards/IVRs to serve customers who are calling in.

New comment by daneel_w in "Oracle cut its Always Free ARM limits to 2 OCPU / 12GB, enforced Aug 18"

daneel_w — Wed, 05 Aug 2026 18:28:39 +0000

What you mean to describe is a free tenancy vs a paid tenancy, the latter having provided a validated payment method which gives them a higher priority on allocations, even if they are of the "Always Free" kind (a single x86-64 vCPU + 1 GB RAM) or "Free-Eligible" kind (e.g. a certain amount of the Arm stuff), even if the user never paid anything.

New comment by daneel_w in "Oracle cut its Always Free ARM limits to 2 OCPU / 12GB, enforced Aug 18"

daneel_w — Wed, 05 Aug 2026 18:18:26 +0000

It has for me. I've had several permanent Arm-VPSes running in OCI for almost 4 years, without paying a single cent. But I'm on a paid tenancy, by having provided a verified payment card, which I think is the distinction that the OP is trying to describe.

A "free tenancy" has limitations: e.g. your provisioned Free(tm) resources can suddenly be yoinked in order to satisfy an allocation from a user who's in a paid tenancy, even if that allocation also fits inside their Free(tm) envelope and ends up costing $0.00/mo. Simply put, the "paying" customers have dibs, even on free stuff.

New comment by daneel_w in "Harvesting SSH Credentials: Insights from My Honeypot Network"

daneel_w — Sun, 02 Aug 2026 21:12:33 +0000

Anonymizing network proxy.

New comment by daneel_w in "Harvesting SSH Credentials: Insights from My Honeypot Network"

daneel_w — Sun, 02 Aug 2026 18:58:27 +0000

No "credentials" are being "harvested" here. It's all worthless data, save for the statistics.

New comment by daneel_w in "DMARC has been public since 2012 but most company domains still don't enforce it"

daneel_w — Tue, 28 Jul 2026 19:27:42 +0000

The "logical OR" of DMARC is absolutely a glaring caveat. I run my own MX and would personally under no circumstance omit SPF, because I consider neither SPF nor DKIM complicated enough to warrant consideration.

New comment by daneel_w in "DMARC has been public since 2012 but most company domains still don't enforce it"

daneel_w — Tue, 28 Jul 2026 16:46:44 +0000

Yes and no. DKIM signs part of the envelope to help recipients detect alteration (by verifying authenticity), SPF locks down the permissible origins for the sender. SPF is in itself imperfect and can in some situations be exploited on open-access shared systems. If the two are used in concert they offer decent protection.

New comment by daneel_w in "GrapheneOS protections against data extraction from locked devices"

daneel_w — Mon, 27 Jul 2026 11:58:54 +0000

I personally run iOS but have given GrapheneOS an extended run on a development phone I use for work, and I have to say it really is a much tidier, less distracting, smoother and more responsive experience than any other Android setup I've used before.

New comment by daneel_w in "GrapheneOS protections against data extraction from locked devices"

daneel_w — Sun, 26 Jul 2026 14:53:06 +0000

No one is stopped from backing up important data. It is, in fact, kind of boneheaded to keep all "valuables" on a single device. I don't understand the scenario of not trusting a device to safely access the Internet or the telephony grid while also insisting that they need a PHONE to keep all their stuff on where they're going, and at the same time somehow trust that both themselves and their possessions are perfectly safe from seizure and extortion in the very same location.

New comment by daneel_w in "GrapheneOS protections against data extraction from locked devices"

daneel_w — Sun, 26 Jul 2026 14:51:52 +0000

You're misinterpreting. They mean that there are additional options next to only keeping these things on your phone.

New comment by daneel_w in "Future euro banknote design proposals"

daneel_w — Sat, 25 Jul 2026 13:49:52 +0000

I personally consider these people European more than their specific nationality, but more importantly they to me represent European humanities, science, and so on.