Re: So, basically, no change there
Use FF instead of Chrome, it really is that simple.
All of my critical FF plugins work on Android the same as they work on desktop operating systems:
- Ad Nauseam
- HTTPS Everywhere
619 posts • joined 15 Oct 2008
It works for some people - namely those that like all the work prevention devices that enable them to spend an hour on some work, half a day on admin to actually work around the security measures, and the rest to slack off.
Of course it will drive everyone who actually enjoys being productive nuts in days and they'll probably choose to leave guilt free on 1 day's notice before their probation period is up.
You are overthinking it.
ssh -D 1080 email@example.com
Set Firefox to use localhost:1080 as socks5 proxy with remote DNS lookups.
No need to mess about with DNS or anything of the sort. You just have to find a port that isn't filtered by the firewall and run sshd on that port on your.server.
"Does it stop people using cars as weapons against people and infrastructure? No, but it massively reduces the possibility."
I don't see how licensing drivers in any way reduces the possibility of a driver deliberately using a vehicle as a weapon against people and infrastructure. At a push it might slightly reduce incidence of incompetence, but certainly not of a deliberate act.
Sure - rowhammer can be made to work - on certain hardware. I've yet to own a device on which it was reproducible using various test programs in 8 hours, and I doubt I'm that lucky. But if it can be made to flip 1 bit statistically in n seconds, then it follows that in n^2 seconds you should be able to flip 2 bits in the same row, and in n^3 seconds, flip 3 bits. But since we are into O(n^3) territory, I wouldn't lose sleep over it.
"But if three bits could be changed simultaneously, ECC would not catch the modification."
Wow! So they actually read up on how ECC memory that has been in used in servers for 30 years works! Are they hoping for some praise medals for participation?
It must be a really slow day if this is news.
"A fair point, but lets face it, at the moment somebody in China would copy your product anyway if there's a significant domestic market."
Similarly with Zotye SR9 (Porsche Macan clone). Interestingly, the Porsche dealers in China aren't at all concerned, and for a very good reason - anybody who can afford even a used Porsche won't be buying a Zotye knock-off.
The thing to take away from this is that competing on the cheap end of the market in a place where labour is much cheaper is a foolish thing to be attempting at the first place. If you can't compete on quality first, at least earn some good karma by not expending more of the finite planet on making yet more worthless tat.
"Up until a local copycat product (that's copied to such a point as including your copyright marks) goes on sale for less than your product"
Cost of doing business in China, it seems. The only winning move is not to play.
For what it's worth, some headway toward addressing that had been made at some point, as short lived it may have been:
You'd be surprised. The workload nature of databases is that you need a _LOT_ of fast memory. GPU has loads of compute on it, but VRAM is far more constrained compared to system RAM when you can buy a server with 1TB of RAM.
So provided your data chunks reasonably cleanly into VRAM sized portions, you can gain some advantage there, but then there's the overhead of loading that data from system RAM into VRAM, where you get a significant hit.
The reason databases aren't commonly GPU accelerated is because it would only be faster for a small fraction of atypical workloads, and massively slower for most typical workloads.
"It is also written in law that HMRC can go back up to 20 years if they believe you have been committing tax evasion, so the legality of a loan you are not expected to pay back is brought into question."
Except this is not tax _evasion_, and such schemes have not been ruled illegal. So the 20 year rule seems like it should not be applicable.
"The law that sets speed limits is the Road Traffic Regulation Act 1984 (you'll want to look at section 6)."
OK, I'll bite, and I just looked at the exact wording. The gist of it is that it states that it is not permitted to exceed the speed limit. Nowhere in that document does it explicitly state that it is permitted to drive under it. So the analogy does in fact seem to hold, and the general principle of the law is that everything that is not forbidden is permitted.
"There is no law saying that EBTs were acceptable practice."
And there is also no law saying that doing 70mph on the motorway is acceptable practice. Anything that is not prohibited is permitted, that is the nature of laws everywhere. This income was legally declared, open to HMRC scrutiny, and could have been challenged any time since. The fact they did not do so in even a single case within the time period in which they were obliged to do so implicitly makes it legitimate. So the comparison is in fact absolutely spot on. It really is equivalent of retroactively reducing the speed limit and then sending everyone who exceeded the new speed limit going all the way back 19 years speeding tickets.
Frankly, it'd be nice if sensibly sized (<= 4") phones of decent spec were to become available again. If I want a device that required both hands to operate, I'll use a 9" tablet. For a phone, I want to be able to reach the top corner with my thumb easily, and being human my thumbs happen to not be 6" long.
As for becoming taller and narrower, this is again the consumer optinion being shaped to maximized profits - the less square the screen, the fewer pixels there are for any given diagonal size. Fewer pixels means higher yealds, which means lower cost and greater profit margins. We have seen this on desktop displays where things went from 16:10 to 16:9 and more recently the trend toward "ultrawide" screens.
Not only is it a dick move, the letter ю is from russian and doesn't exist in some cyrillic alphabet languages. So it's even more of a dick move because it opens an additional can of worms, e.g. whether there should be a 3rd TLD for eu, this time ".ey" (y us cyrillic u).
It sets a terrible precedent for effectively going back to code pages and character sets like they were before UTF-8 was invented by effectively only allowing domains that use the same character set as the extension. What's next, are they going to, say, ban škoda.eu because š isn't ASCII?
"No, your UK debit card won't work on the European network."
I call bullshit. I take a road trip across Europe every year and have to buy fuel, food and on occasion get cash out of an ATM, and can confirm that my UK debit card always works just fine in France, Belgium, Netherlands, Germany, Austria, Slovenia, Croatia, Serbia, Hungary and Bulgaria. Also always worked fine in US. Further, you pre-emptively contradicted your own previous statement on this very subject by saying:
"... card cloned and lost several hundred quid as a result. We only got it back (my mum was too embarrassed to argue) when I made a complaint, as the card was used in the Netherlands and then in Malaysia an hour apart"
So the cloned card did in fact work in Netherlands and Malaysia?
"Does anyone know whether HMRC have ever responded to this frequently-made point about benefits?"
Not officially, but self employed (note: important distinction vs. ltd. company) people have in the past successfully sued their clients for statutory holiday pay and pension contributions after they were engaged as self-employed contractors and then found by HMRC to be caught by IR35. So there most certainly is legal precedent for it.
The real kicker here is that the client-side determination of IR35 status here is also coupled by the client's insistence that the contractor be operating through a ltd. company rather than as self-employed. This key point allows the client to dodge the employer's taxes and palm off the responsibility for them onto the ltd. contractor.
In other words, the entire setup is actively rigged for the government departments to dodge both employer's statutory liabilities/obligations _and_ employer's taxes and shift them onto the employee.
The test for employment vs. self employment should be made the same for both employment rights and tax purposes. If your client isn't paying you statutory holiday pay and pension contributions, you are self-employed and should be taxes as such. If they are, you are employed, and your client is liable for deducting your taxes at source and providing at least the minimum statutory benefits.
Anything else effectively amounts to expecting a employees to pay employer's taxes.
The simple fact is that ECC _does_ make rowhammer ineffective in the few cases where it might plausibly be exploitable.
Flipping one bit is hard. Flipping two bits in the same memory row at the same time is exponentially harder. Flipping three bits is exponentially harder again.
ECC memory can:
1) Correct 1 bit errors transparently
2) Detect 2 bit errors (causes an NME and typically a kernel panic)
It takes flipping 3 flipped bits in the same row of memory to introduce an undetectable memory corruption with ECC memory. And flipping one is difficult enough. Not to mention you have to get extremely lucky to be operating on a row of memory adjecent to the one you are interested in corrupting.
TL;DR: Don't lose sleep over it, especially if you have ECC memory.
"Providers could end up targeting the same lucrative areas to invest in fibre, rather than creating a more geographical spread."
It is a GOOD thing that they are all targeting the same areas. It means there will be competition wherever it is available, forcing providers to ensure they are providing a good quality service at a competitive price. Once they cannot take the piss by charging sky-high prices (anyone seen FTTPOD prices recently?), they will have to spread further and further out from cherry picked areas in search of more revenue.
Overbuild is exactly what is supposed to happen, and it is a good thing.
"ditch the Hub 5 and replace it with a decent router/AP of your own"
The problem with the HH5 isn't the hardware. The hardware is actually pretty good. The problem is their customized firmware after a specific version. Never had any problems until that July 2016 firmware update. Before my contract was up I flashed it with OpenWRT (soldering required to get serial console up), and that worked just fine.
It's a major achievement for their customers that they even acknowledged that there is a problem.
BT Hub 5 got a firmware update back in July 2016 that made it unstable. It would reset under load. I was setting up backups at the time, and the uploads to the cloud would never complete because of this. After arguing the toss for a few months they sent me a replacement BT Hub 5, and that worked great for a week - until it, too got the same firmware update, and started resetting 100+ times per day.
After all this, I got one of the L2 support engineers to finally confirm there is a firmware big, and that they are working with people who work on the firmware to resolve it, but they I got a call back from his boss who flat out denied everything the one and only helpful person they have working for them told me. His advice? Get a standalone VDSL modem instead of using the one built into the BT Hub 5, but that was still totally not an admission of a problem (because there would be more people reporting it if I wasn't imagining it, and the 100+ hub resets per day they could also see from their side were obviously just a shared illusion).
The only bit of good news in this story is that my contract term was coming to an end so I took the opportunity to mentally note to NEVER sign up with BT again and switched to a different ISP.
The problem is that switching context is in the order of magnitude of 100x slower in a VM than on bare metal (addding microseconds to nanoseconds).
That is why some workloads virtualize with minimal performance hit (few threads, low concurrency, mostly userspace CPU burn), and some workloads virtualize extremely poorly with a huge performance hit even without meltdown patches (anything highly concurrent such as compile farms, databases). I have measured performance hit from virtualization on some such workloads to be upward of 30% - and that was before meltdown patches came into play.
I have to say I'm dismayed by the trend of moving toward phablets. IMHO, the ideal form factor for a phone is about 3.5" (ZTE Blade). I grudgingly upgraded to a Moto G (1st gen) when Android 2.x went out of support and Android 4.x was never ported to it.
I find Moto G's 4.5" size unnecessarily big and bulky, but have been resisting upgrading to a newer phone because finding anything that is:
1) not bigger
2) supported by LineageOS
is pretty much impossible at the moment.
Until the market provides a sensibly sized device, it won't be selling me a new phone.
Biting the hand that feeds IT © 1998–2019