Posts by matjaggard 247 publicly visible posts • joined 22 Apr 2015
This is exactly what I thought. Some VPN company pointlessly getting advertising. Shame on the Reg for sharing this. Everyone uses simple passwords for things they don't care about don't they? I get pretty irritated about sites with no data requiring long passwords. I know my passwords for some sites have been pwnd and I'm just fine with it. Go ahead and login, view my zero balance on a gift card, I just don't care.
Too late. I already did before anyone had heard of Copilot. The problems with copilot are complex and it's pointless trying to say otherwise in either direction.
Everyone has a right to learn from code of all licences that give you permission to view it. Surely that should apply to AI learning too.
Copilot can and does spit out non-trivial copyrighted code with no indication of that it might be copyright. It also learns from code in a completely different way to a human so maybe only human learning should be allowed by these licenses.
Then again the licenses don't specify only human learning. Maybe someone should write one that includes this clause just to make a point?
I used CUPS to get my printer working with a Raspberry PI but nothing "just worked" at all - getting QEMU to work and run the x86 printer driver on ARM was pretty easy but getting my print to go from Windows or Android to CUPS was next to impossible. Cue messages about encryption being required or the printer being offline, or just no errors at all but no print output. Only MacOS could reliably print to my CUPS server and I had to give up.
Agreed. I use Java and Kotlin extensively (neither for Android) and once you get to recent versions there's not so much benefit of migration especially if you use Lombok to cut down on pointless boilerplate (even then Java records work better)
Immutable collections are now the "default" on Java too. See List.of and similar which are equivalent to the Kotlin collections.
Regarding condensing dryers not working. I think they actually work perfectly well but to meet efficiency targets/ratings the companies seem to choose "don't bother drying the clothes" as a valid cycle to get tested. All our cycles take significantly less time than advertised when I start them but I have to run 2 or 3 times if I want the clothes dry enough not to smell when they're put away.
That's true but it is more nuanced than the article implies. When I was working on a project once, we'd licenced proprietary source code doing a job and we wanted to dump that licensed code - first we needed to spec the work it did and get it implemented by someone who had never seen the licensed code because they would almost certainly have learned things from it and implement them based on what they learned. This is very similar except that previously we were asking humans to make a judgement call on when it was valid to use learnings from a differently licensed code. Now this tool is using that learning without giving the human any knowledge of the original license so the human cannot make the judgement call any more. That means that the AI is left to make a moral call on how and when it is valid to show some code to the human - maybe they have included some safety to codify this judgement, but I doubt it.
I think that's going a bit far but I do broadly agree. Most advertisers seem to believe the targeted ads are better lie* and it's the data companies (Google, Facebook) that make all the money.
* Actually targeting based on _historic_ data rather than current context is what's wrong - clearly ads on the Reg will be more technically focused and search term ads are important.
The real problem with lightning is that Apple make lots of money from licensing it to vendors of products that plug into it. Other than that it's on a par, if not slightly better than USB C. I've never had an iPhone but it's not the port that puts me off. I can definitely see both sides of the debate here - innovation will decrease if we stay with a single standard and law makers are not known for updating their legislation to the latest standard very often. On the other had, there are very few houses without USB C chargers currently so I don't see people throwing away lightning cables as very important compared to the benefits of a single standard going forward.
I'm fine with adverts. I want people to be paid for the quality content they create so that they keep creating it. It's a sad fact that the companies that are hosting this stuff are making so much money in the process but I'm honestly more worried by Amazon than a few YouTube adverts.
I'm sorry but I actually like Chrome a lot, especially the developer tools.
I've tried to get back to Firefox - not because of privacy, I know that's gone anyway (I use Gmail and Facebook products. Privacy is the cost and like most of the world I'm willing to pay it) - but because I think it's not good for there to only be one browser rendering engine and all the others use WebKit/Blink.
It is pretty frustrating for those of us who haven't written any of the code too though. Linux provides an excellent base for a laptop, desktop or mobile OS but only those UIs provided by Google ever get used by more than a fraction of a percent of the market.
@Voiceoftruth I don't understand why so many down-votes and it makes perfect sense to be frustrated with this even if you're not the one wasting time on YALD. I'm also annoyed because I'd love to use Linux for everything - I did for about 2 years and instead of spending my time waiting for filesystem operations to complete (which I did on Windows) I spent it fiddling with stuff which basically worked but just not quite smoothly enough to get out of my way.
Thankfully I'm not a designer type at all so the 3 or 4 different look and feels for different windows didn't bother me at all but for some it would grate.
The frustration comes from so much potential and yet so far away from a system that anyone could use.
That's still the norm in sensible countries (eg Europe) where software patents are almost completely banned (if I recall correctly you have to show a real life, non-computer application of an algorithm for it to be patentable?) and the patent offices properly investigate before granting.
Contractors are often tax dodgers - registering their work for a single company in a different way to pay less tax. Until IR35 that was tax avoidance, after it it's tax evasion.
Sadly the rules have been changed so badly that the whole system is suffering from not knowing what's in and what's out and the enforcement of those rules is somewhere between unlikely, difficult or impossible.
I don't understand the points you're trying to make. Encryption between my browser and mg bank is point to point encryption - HTTPS generally. It's not end to end - the message is sent to the bank's systems via SSL where it is then encrypted again with another SSL connection to the bank employee dealing with the message. This is the correct way to handle communication with the bank and is not End to end encryption.
End to end encryption might or might not be a bad idea but don't try to make that argument using some technology that it is not even used for!
End to end encryption is not required for the web to function - until recently nothing had it, you had encrypted connections from you to a server and encrypted again from the server to your friend. It seems Facebook decided most of the value was from who talks to who rather than what they say, or possibly they felt pressured into end-to-end encryption. Either way, the biggest problem I have with removing it is trusting the companies and organisations that have the data not to misuse it or be hacked. Whether that risk can be mitigated enough to be worth having access to criminal communications is debatable. The worst criminals would just move into another platform anyway and who knows which platforms use genuine end to end encryption. I think there's a good chance telegram's is broken
Stop empty bragging?! Seriously? This is a tech company - almost all of their income depends on advertisers believing that they have data and know how to use it. The truth is that they have less data than they claim and they don't know how to use it - if they did then you'd get relevant adverts all the time and I rarely do. Mostly I get adverts for the things I've already bought.
Although basically irrelevant to this article, investigations by sources I mostly trust and know which areas not to trust them into the immigration piece:
https://www.washingtonpost.com/immigration/kids-in-cages-debate-trump-obama/2020/10/23/8ff96f3c-1532-11eb-82af-864652063d61_story.html
https://www.bbc.co.uk/news/world-us-canada-44303556
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2022
