karteum

> "GPUs, through a consequence of history, chose not to focus on real-time implicit geometry like curves, but instead on everything that goes inside them"

Just for information : 10 years ago, NVidia has released this paper and extension "GPU Accelerated Path Rendering" : https://developer.nvidia.com/gpu-accelerated-path-rendering , so it seems GPU can still help

Notice that Google Takeout is broken with regards to non-UTF8 email encodings (for some of you who archive old emails). I noticed it when coding a small tool https://github.com/karteum/gmvaultdb to (among other things) import the mbox from Takeout into an sqlite DB : Google performs some encoding conversions that permanently break all non-ascii characters in non-UTF8 emails (they are all replaced by 0xEFBFBD). In order to archive old emails, I had to use mbsync with IMAP (gmvault does not work anymore with Python 3, and I didn't try yet https://github.com/GAM-team/got-your-back ).

> "Also, it would be nice to fool the GPS through a jammer that emits false signals in order to make it report a fake location, after placing cameras at that place to record whoever shows up."

First : don't confuse jamming with spoofing. Jamming is extremely easy to perform by any idiot, while spoofing GPS is significantly harder https://archive.fosdem.org/2019/schedule/event/sdr_gps/

Second : keep in mind that both of them are actually illegal in most countries, and besides unless you exactly know what you are doing it's fairly easy to jam/spoof on much larger areas than you would intend to ! (and GNSS is not restricted to consumer products and maps but are used in a lot of industrial/serious applications so it can actually have consequences)

3G (UMTS) was reuse 1, not reuse 7 !

This is not the only mistake from this website you mention : e.g. in the page "difference-between-gsm-and-cdma" I read "GSM Data speed rate: 42Mbps in HSPA" (but HSPA is 3G and an evolution of UMTS and is using WCDMA, compared to GSM which is 2G and is using TDMA). I guess when they mention "CDMA" or "3G" (which is are generic terms that also apply to UMTS) they mean "CDMA from 3GPP2", which is not used in Europe. I didn't read everything but from the little I have seen I don't find this website very reliable...

"Infamously" ?

Well, there is for sure a lot to criticize about the CN government, but this precondition seems to me very natural (the OS is a very natural place for possible backdoors, otherwise...)

But wind turbines and photovoltaic panels are also neither green nor sustainable (especially when you take into account all the mining and the CO2 involved in their manufacturing and transportation, and especially when you take into account the backup system e.g. gaz or batteries to compensate for their intermittent nature).

Yes exactly. As a data interchange format I think it is much better than CSV (more compact, faster, more efficient, less error-prone, etc.)

I don't understand why people keep using CSV today while SQLite is MIT-licenced and can be used everywhere, has a very portable and lightweight implementation, has a stable file format with long-term commitment, and a good compromise on the type system that gives enough flexibility to be on par with CSV if some entries happen to have a different type...

I am not native english, but if I understand correctly : that woman (board member Debra Rose) made false accusations of aggression and child abuse against this dad because he was insisting for getting public data that she didn't want to give, and at the end there is no consequence on her ??

I understand that he had legal "victory" when the judge ordered the state to hand over the data and pay $35000, but I didn't see anything about that woman being punished for false accusations (I though that the author of false accusations would face consequences of the same magnitude as what would have been endured by the other party if the false accusations succeeded, in this case: he could have faced jail or have his children removed)

So you also think Linus Torvalds is a narcissist ? And every people who name their project or company after their names are potential sociopaths ? What about all scientists who gave their names to the effects or law of physics (or math formulas) they discovered... ?

Unless there is a good reason, I think in that situation I would probably try to avoid storing those tens of millions of small json files directly on the filesystem. For example : have you considered storing that data in jsonb/postgresql https://www.compose.com/articles/faster-operations-with-the-... ? (or even just in an sqlite db https://www.sqlite.org/fasterthanfs.html )

> I host a statically rendered wikipedia mirror. Tens of millions of gzipped html-files with a filesize in the range 1-5 Kb.

Have you ever tried Kiwix (https://www.kiwix.org ) and looked at the ZIM format (https://en.wikipedia.org/wiki/ZIM_(file_format) ), in particular those for Wikipedia (https://download.kiwix.org/zim ) ?

> _but without Linux, Hurd would still eventually be ready_

"eventually", however with "if" we can imagine a lot of things. It could also been said that "if Linux didn't exist then maybe BSD would rule the world" ;)

I don't think they will, it would be suicide (their business model relies on the fact that they can collect data from you. It's free because you are the product !).

Yet, it's still useful to keep in mind that they can disable your account at any time (with _all_ associated services i.e. your email, contact list, youtube history, Drive documents, etc) without any appeal (e.g. because AI would have triggered supposed misconduct from you or because AI may think you are not the genuine account owner !). Because of that, it's always good practice to periodically backup your data (e.g. using takeout.google.com, noting that the export from GMail is half-broken with non-unicode characters - which is why I backup it with IMAP...).

Huawei/Honor phones now have a locked bootloader so unless there is an exploit it's not possible to install anything else than the stock firmware https://www.xda-developers.com/huawei-stop-providing-bootloa...

> After upgrading the display, CPU, disk, RAM, fan, battery it's a bit click-baity to say you've gone back to a 2012 ThinkPad

Well, I think the main point is : regardless of how better some modern components are (e.g. screen resolution, CPU energy efficiency with 10nm process, etc), the build quality, keyboard comfort, repairability and amounts of I/O are critical aspects that have dramatically regressed in the recent years. (Not even mentioning the Intel ME and Boot Guard that prevent flashing Coreboot).

Besides, it also shows that a 10 years old CPU/GPU is still perfectly usable (and since we only have one planet, it would be desirable to extend the lifetime of our devices... (upgrading the RAM/SSD/Wi-Fi has an environmental impact, but far less than changing a whole machine !)).

I understood this happened in a P3 lab. I would have expected a P3 lab would have better procedures to avoid being bitten by an infected mouse, especially since some P3 labs host experiments with viruses sometimes much more dangerous than sars-CoV-2, and that would be a disaster if such a virus escaped... https://www.science.org/content/article/exclusive-controvers...

> This is one thing I'm surprised the open source community hasn't developed a popular alternative for: Teamviewer.

That's because it's not only a matter of software but also a matter of service (STUN is not enough, TURN is required in many situations. And someone has to pay for the servers and bandwidth...). If IPv6 had succeded instead of IPv4+NAT, maybe the situation would be different.

In terms of software, the open-source x2go (based on nomachine NX libraries and protocol, see https://wiki.x2go.org/doku.php) has been very efficient in my cases when bandwidth was limited and when I had no big NAT issues.

Well, here in that case they were using an approach that would be illegal by EU regulations (not sure about US) considering the EIRP limits :)

And if you want real low-power long-range communications with the other side of the earth, you may be interested to have a look at WSPR https://en.wikipedia.org/wiki/WSPR_(amateur_radio_software) !

N.B. note that Google performs some encoding conversions within the GMail backup from Takeout that permanently break all non-ascii characters when an encoding other than UTF-8 is used in the original mail (those characters are replaced by the 0xEFBFBD sequence, and since information in lost it guess it cannot be properly recovered).

Obviously I am not the only one who have been experiencing the issue, e.g. https://webapps.stackexchange.com/questions/71153/takeout-br...

The issue does not happen when doing backup with another tool e.g. mbsync.

Great news. One remark: "The on-disk format for the table data is the same" => I guess this means there is still some kind of "type/length" header for every single record... (which could theoretically be dropped) ?

SerenityOS is licenced under BSD-2-Clause License while SkyOS was entirely proprietary (and therefore disappeared when its development stalled), so the comparison seems irrelevant (your point would be more relevant if you were mentioning Syllable instead, which was quited advanced and open-source, yet somewhat seems to have disappeared... https://en.wikipedia.org/wiki/Syllable_Desktop . But since source-code is available anyone could restart it :)

I also hope that SerenityOS has a long and successful lifespan, as it seems promising !)

> It's simple, safe and efficient

We all know that having a single turn (like in the US) leads to a political landscape with 2 major political parties and few nuances (but the major problem in the US system is "the winner takes it all" policy in most states: if the number of members of the electoral college were proportional to votes in each states, the results would already be more fair)

But the French system (with 2 turns) is also very far from optimal and non-linear, and also leads (to a lesser extent, and some variants/surprises) to 2 major political parties + some smaller "satellite" political parties. And some issue are for example that

* sometimes, a candidate that would win against all others has a too low score at the first turn to pass to the 2nd turn (this is what happens in 2007 where all polls said that François Bayrou (who was center/moderate and more consensual) would have won both against Nicolas Sarkozy and against Ségolène Royal, but couldn't get enough votes to pass the 1st turn)

* sometimes, the opposite issue happens i.e. a candidate that has no chances to win the 2nd turn has enough votes to pass the 1st turn: this is what happened in 2002 when Le Pen (extreme right wing) supplanted Jospin and went to the 2nd turn (that Chirac was then 100% certain to win at that time, with a consequence that there were no real/constructive debate about the programs)

* some people who have new ideas cannot participate because you need 500 signatures of already-elected people in order to participate (so if you are not a member of an established political party, it's very hard to become a candidate to the presidential election)

Just to say that there is no binary split between "democracy" and "dictature" : the voting system and algorithm influences a lot the political landscape and therefore the debates. Maybe people here are familiar with the Condorcet method used (among other things) in https://www.debian.org/vote which is better in my opinion (yet a little bit difficult to implement in a wider scale such as a presidential election). Another approach / voting method that I quite like is this one (only in French, sorry) https://fr.wikipedia.org/wiki/Jugement_majoritaire : I like it because 1°/ it is simple to implement and to understand, 2°/ it would inherently avoid the issues above and forces the candidates to be more consensual (i.e. to become "president for everyone") rather than cleaving, while still enabling nuances to express during the debates (not only on a "binary" axis left/right).

> Its why should anyone build a house in the middle of nowhere and expect the govt to spend whatever it takes to get connected.

But the article says "he expected to pay a few thousand dollars, given that the nearest pole was across the street, about 45 metres away. He was shocked to find out it would cost $80,000".

So it's not the case of a house in the middle of nowhere...

> I think literally the only stuff on my phone that isn't backed up is local files

And besides downloads, there are many apps that just use local files and that you therefore need to transfert manually. e.g. apps from F-Droid, SMS, Wechat, etc. So even though "many" stuffs are automatically stored in Google servers, you still need a lot of case-by-case manual attention and care when transferring to a new phone.

Even if/when they don't fully "own" the phones, users ought to always be able to access their data (at least for backup) and call emergency services (which is a legal requirement in many countries).

More generally and regardless of time pressure, I can't believe someone can "play" with such test a script (used with a production tool) without even thinking it might fuck up and hit real phone numbers... (and if the developer was aware that his approach might fuck up, then he should make his boss aware that he had no time to fully test the feature rather than play russian roulette. Time pressure is not an excuse for everything...)

Considering the absolutely huge CO2/environmental impact of BTC and other proof-of-work cryptos, talking about a CO2-coin for the purpose of CO2 sequestration sounds like quite a joke to me...

> "some reason that I don't agree with[0]"

On my side I am quite happy that the hardware+software can support some kind of fast wake-up (similar to Macbooks and smartphones instead of needing >10 seconds to wake up when I open the lid). I just do a real shutdown when I don't need the laptop anymore.

And precisely : as it was said by someone else above, on modern computers with SSDs, I would rather disable "fast startup" so that the computer really shutdowns when I ask him to (which is also useful when you have dual-boot and your Linux systems wants to get r/w access to the NTFS partition)

But I admit that the computer could/should automatically enter in suspend-to-disk after being in "modern standby" for more than e.g. 5 minutes, and this should be configurable.

What we need is control on the bootloader, with the ability to unlock, load our own keys, and relock.

We also need to clean-up the mess with all those "partitions" (some of them with critical informations e.g. calibration, IMEI, etc) so that only one partition would have all those static information (reasonably protected against overwrite, e.g. colocated with bootloader and device-tree). We should be able to re-partition the storage (like we do on PC) without bricking the device...

I wouldn't say Pascal completely "failed". e.g. Delphi was quite popular at some point.