Re: Pillage of the Open Source projects @Stabbybob
When it stopped using GCC and developed its own compiler, it open sourced that too.
2518 posts • joined 18 Jun 2009
It may look like looks a bunch of clueless amateur retro gaming enthusiasts secured capital but RCL delivered the original Vega and its most public face, David Levy, was present and active during the original microcomputer boom — he was part of the team behind the Enterprise, that leading 30 years later to probably his only positive coverage on El Reg.
Whatever the story is that has led to the Vega+, it's not the usual crowdfunded hubris.
Especially not with me. By default it changes ownership of /usr/local/bin to your login user. So anything you run from then onwards can install a shim to usurp any binary that ordinarily lives in /usr/bin. Such as sudo.
How often do you inspect which application named sudo is asking you for your administrative password?
I'm not sure it's accurate to say that other micros typically had to live-toggle a bit. Of the successful ones I'm pretty sure that's only the Apple II and the 16/48kb Spectrum.
None of them is a match for the feature set of the SID, but the 128kb Spectrum and CPC share the AY which is three channels of square wave and/or noise with volume envelopes; the 8-bit Atari has the POKEY which is four channels of more-or-less square wave; the BBC has an SN76489 which is three square waves plus a noise channel, etc.
The SID's killer feature is phase accumulation for pitch selection rather than simple division, giving much finer control — in a SID there's a 24-bit counter, the top few bits of which are used to form the output level, and an amount that is added to it at each cycle. Plus some analogue filters. On the other chips there is the input clock and then there is an integral divider. So you're controlling the reciprocal of pitch, reducing useful precision.
Nevertheless, the other chips don't require active CPU participation as the 48kb Spectrum does, and the musical opportunities are still fairly decent.
I think you're off by one; the shortest 6502 instructions take two cycles, and the most common ones — those which read from or write to the zero page — take three.
But the issue in a real machine is that a 6502 uses only half a clock cycle to perform an entire memory access whereas the Z80 uses at least two. So pick your clock speed as a function of those constraints and your memory speed.
The ZX80 fetches its display in software, but contains only static RAM.
Rather than bother with all that nonsense of counters and whatever for fetching video, the processor just executes the display buffer. Well, it tries to, but the parasitic video steals the opcodes it is actually fetching and forces a NOP onwards. That gives the character code, and hijacking of the Z80's refresh cycle gives it a chance to get the actual pixels for that row of that character.
So most of what the Z80 in a ZX80 is doing is executing NOPs.
A NOP takes four cycles because there's no memory bandwidth to fetch anything else until four cycles later; the Z80 spent two cycles fetching the NOP opcode, then decoded and performed it during the two cycles when it was issuing a DRAM refresh. As soon as the refresh ends it can seek out the next thing. That's why it's also four cycles for all the other single-byte instructions that don't imply any other accesses to memory — register-to-register arithmetic and moves, and a few others.
The Z80 in the Spectrum is not only nominally clocked at 3.58Mhz but also genuinely runs at that speed for as long as you avoid the physical chips that are shared with the ULA. E.g. on a 48kb Spectrum that means that as long as your code is in the top 32kb of RAM rather than the bottom 16kb.
The CPC is nominally clocked at 4Mhz but via use of the WAIT line permits a Z80 memory access on only one in every four cycles, regardless of what you're accessing. The standard fetch cycle is four cycles long, so single-byte instructions that don't cause a memory access run without a speed penalty (once you're in phase, anyway) but everything else is subject to delays. As a result code often ends up running more slowly than it would on a ZX Spectrum.
It depends how often the Spectrum code is seeking to update the display though, obviously. And the CPC's main problem isn't this clocking scheme or that one, it's the annoying large percentage of titles that are so lazy as just to be the Spectrum code plus some extra work at the end to translate the Spectrum graphics to anything that looks sort of right. It's almost a revelation every time you load a game that was converted properly, like Chase HQ, Robocop or Gryzor.
I don't see the issue; review sites that depend upon the revenue of persuading people to buy the apps they review immediately then and there are not on my list of fine upstanding gentlemen. The decent ones already make the majority of their income from running adverts, so that there's no ulterior motivation for positive reviews.
So Apple withdrawing an affiliate programme doesn't really feel like another effort to screw the little guy. Just one ugly profit-seeking corporarion declining furthet to help other ugly profit-seeking corporations.
I read this story entirely differently.
Bosses spot that divergence in platforms costs extra money. Decide to introduce mediocre middleware to minimise cost. Developers are upset that they'll have to work in a poorer environment, working on a worse product. Money is saved, so bosses use their platform to write the first draft of history: that everything they wanted is definitely a success, and everybody else was wrong.
Right now it seems to be Youtuber versus Twitterer* as far as data points go. Which person you've never heard of, publishing their results via the internet's various platforms for the attention hungry do you most prefer?
I think Ars at least has pushed back its review of the new Apple machine in order to test the claims and provide something like an evidenced conclusion; I hope other outlets are doing the same.
* to save everyone the reading: he seems to allege the problem is in the use of Adobe Premiere Pro for benchmarking, by posting figures that show it takes almost 2.5 times as long as Final Cut Pro X to perform the same task on the latest MacBook Pro. I think he's suggesting Premiere isn't well-adapted to modern processors. It's Twitter though, so mostly the word-based version of a Rorschach test.
So your argument is: Google's alleged placing of terms into its contract to effect penalties if the other side ships anything that competes with a Google products is not "clearly" anticompetitive? And that the main reason anybody here might think a legally-enforceable contractual term that prohibited competition was anticompetitive... is pro-EU bias?
Per the article, Google to phone manufacturers: "if you try to market even a single FireOS phone, we will withdraw your licence to ship fully-functioning Android phones".
Your definition of what is and isn't "clearly a violation of anti-trust rules" must vary from mine.
Competition law protects markets from distortion, for the benefit of consumers.
So a company with only a small slice of a market can do whatever it wants, because it does not strongly influence market.
When the company with 90% of a market prevents manufacturers from considering diversification, that's textbook anticompetitive behaviour, and it should be obvious why that's a substantial detriment to consumers.
From the article, Vestager said:
[Google] dominates licensable mobile operating systems ("over 95 per cent"), app stores ("over 90 per cent") and mobile search ("over 90 per cent in most European countries").
iOS is relevant to exactly one of those categories, and 90% is the correct number if you're factoring in iOS. Further:
The commission objected to three practices in particular: the requirement to preinstall Google Search and Chrome, payments to phone makers to make Google Search the default, and restrictions on creating "forks" of Android.
[Vestager] said manufacturers were interested in licensing Amazon's FireOS Android. But by making even one FireOS phone, the OEM would have lost the ability to include Google Play Store on its other devices.
The allegation is that Google's 90+% of phones mean that the anticompetitive terms it imposes on other companies — e.g. barring them from including Google Play if they offer any product featuring FireOS — is an illegal distortion of the market.
Do despite the appeal of bad-guys-on-both-sides whataboutism, I really think Apple's contribution has already been factored in here.
Structure and Interpretation of Computer Programs was still one of the favoured introductory textbooks when I got to university at the turn of the millennium. I took a Scala course a few years ago and thought I was doing really well until I realised that all of the problem questions were just direct adaptations of SICP originals so I'm not sure I was coming up with solutions so much as regurgitating them from some deep recess of my memory with different syntax.
... and it took me a while to find but Elite's source was transcribed to C and made available for then-modern platforms back at the turn of the millennium by Christian Pinder, but then taken down at the request of Elite's other coauthor, David Braben.
Nevertheless, that conversion remains available thanks to the persistence of anything ever put onto the internet, and Github. It's probably easier to digest than the original assembly though the games library it relies on to access the display, read the keyboard, etc, has introduced significant breaking changes in the interim. It shouldn't be a big job to replace though.
Ian Bell, Elite's coauthor, has been offering the source for download for years, along with what was produced as to ship designs for a putative sequel.
If memory serves, they used the assembler built into the BBC's BASIC. So it's a few different files that when run produce parts of the whole.
It's a bit more likely I'll bother to go somewhere physical, I guess, but the sales taxes here really aren't that much.
I ordinarily pay sales tax of about 7%. If I drive twenty minutes I can get to a shopping centre in an incentivised tax zone and pay just a bit more than 3%. If I drive for two hours I can get to a shopping centre in one of the states where the sales tax is 0%. And here in the US the petrol is less than 60p a litre*, even after all the hand wringing over recent rises, so it's really a time and boredom calculation.
At least that leaves lots of spare money to try to deal with the awful healthcare system?
* okay, it's a shade below $2.90/gallon.
I'd read that more as "if the element was found before reaching the end" rather than inflecting it with negatives, but I agree with your point in the main part. C++ is a lot of really useful stuff and good ideas, hiding behind syntax kludges.
SFINAE is possibly an even better example. If anybody here isn't already familiar, look up std::enable_if, then calculate the ratio between the amount of time it took you to understand the purpose and utility and the amount of time it took you to be able to produce the syntax.
I disagree; templates are the best bit — the syntax could be cleaner but being able to write std::find(container.begin(), container.end(), value) and that result in the compiler being able to generate the proper code to search any iterable container is a neat alternative to more traditional approaches towards the same objective like dynamic dispatch. Specialisation (i.e. providing special cases explicitly) is the icing on the cake.
It's just a shame that (i) it's helpful for the template system to be Turing complete; but (ii) the look-at-me-I'm-clever crowd think that barely-comprehensible template hoops are a fantastic way to advocate the language.
Benchmarking is a fools' game, of course, but the ARM at introduction was sufficiently faster than the then high-end x86, the 386, that for a while Acorn sold it on an ISA card for use as a coprocessor.
The marketing puff is here; a PCW review is here, though it fails to come to a definitive conclusion on ARM v 386 it makes statements like "The 8MHz ARM processor is one of the fastest microprocessors available today" and "A fairer [price] comparison would perhaps be with other fast coprocessor boards for the IBM PC, such as the 80386, the 68020 and the Transputer boards" which certainly seems to bracket it with those others.
Sorry to ruin the joke but this attack doesn't fool any of the built-in OS-level security measures, just a bunch of third-party apps that check the signature on only the first architecture within a fat binary.
So the blame-claim would be: they're validating it wrong.
... though hopefully Apple will do something about whatever the APIs are to encourage correctness by default.
When you're standing up in court being confident that you can quickly find the piece of documentation you need, and see its contents, probably while also paying attention to what someone else is saying or while speaking yourself, is fairly important.
I can't speak as to the solicitors. They probably just like being able to work on the train?
From the realm of Brexit, why not apply the max fac technology to court digitisation? All documents will magically be recorded by cameras at the court's door, even if inside containers, because you know, technology. Get the same guys to work on it who are behind the governmental push for secure encryption that allows contents to be inspected by an unapproved third party, they seem to be able to be good at this impossible requirement stuff.
Of the trinity of 1977, the PET and TRS-80 are silent monochrome character-mapped devices which produce sparkles/snow if the programmer doesn't specifically avoid writing graphics during the active part of a frame, and the Apple II is a bitmapped device with a transparently-shared bus, some colour support and a toggle speaker.
That device was demonstrably better — the demonstration would be drawing some graphics or making a tone.
It was overtaken and undercut but I'd dispute that it wasn't better than its contemporaries.
Rule of thumb: if Woz was involved, the product was probably the best at something.
On the one hand, and despite their gross overrepresentations, I accept that Apple were one of the companies that helped to bring huge groups of people into the home computer market.
On the other, it seems counterintuitive that a company could produce the computer that started the revolution... if there was already a 'Byte-Shop' to carry the product. That sounds a lot more like a revolution that had already begun.
I guess that depends on how much further appreciation you foresee in the Apple I's future — in the noughties if you were lucky you could get one for 'only' $20,000. With hindsight that would have been a good investment.
I'm not expecting similar further growth but since the history of prices don't seem to follow any sort of rational pattern of growth, my expectations are clearly fallible.
So now I'm actually going to have to do the not-a-robot tests? Probably worth it.
If there is ever a robot uprising, all we need to do is drive at them in our cars. Apparently being able to tell which roads have cars on them proves you're not a robot.
I think your method of questioning might be at fault; Safari stores all passwords in the keychain. Open Keychain Assistant, switch to passwords, enter any website name in the search field and there are the entries that correlate to my passwords. Double click one, tick the 'show password' box, enter my system password and there it is.
It's stored associated with only a randomised ID as metadata, and it loses that association after six months. It then survives untethered for the remaining eighteen.
I'd therefore posit that it's being used for Siri training, not for the sake of "acquiring personal behavioural data".
That is to answer your question literally. It doesn't make me a lot happier to know that my requests would be stored for two years even if they didn't have details attached that more explicitly identify me. That's partly why I don't use Siri, though my main justification is the same as for the other voice assistants: I'm unclear of any situation in which they'd be useful for me.
As a fellow MacBook Air 2011 owner, it's also a little sad for me that it will exit the list of Macs that receive the latest version of the OS with the next release, 10.14. Certainly if your issue is that you find Apple devices only talk well to other Apple devices then you might as well write that out of the set of mutually talkative devices in the near future; they don't explicit switch these things off but proprietary protocol rot takes its toll. My iOS 5 and iOS 9 iPads can still play video content from iTunes but the iOS 9 device that's supposed to sync with my iTunes Match now inexplicably offers only maybe a twentieth of my content.
That being said, I'll be back for a newer Mac and an iPhone update if and when my 6s becomes a burden; I'm not particularly interested in home automation, habitually rent movies through Amazon rather than iTunes, and my iPhone works flawlessly with my Garmin watch.
I bought my Garmin for run tracking. I have unexpectedly also found occasionally useful its display of notifications from my phone, but often more so the fact that it buzzes when it displays one so that even if my phone is somewhere deep in a coat pocket I don't tend to miss texts or emails from the few people I'd already told my phone it was okay to generate notifications for. That, with its week-and-a-bit battery life is plenty smart enough for me.
So, I think I'm arguing: not only was it bought it for a purpose, which it performs very well, but it's exceeded that usage. So I'm all for Garmins.
Agreed; also it feels like the real offence would be continuing the behaviour now that the law is clear. Not being able to see this outcome coming doesn't seem all that troubling.
Put another way: even if the verdict stands and it definitely is true under US law that Trump's actions violated the First Amendment, do you think he was actually motivated by the desire to restrict people's legal freedom of speech? I feel like he's unwittingly stumbled into the non-obvious violation.
If someone is really crooked, get them for something that's really crooked.
The plaintiffs are seven people who were blocked and the Knight First Amendment Institute, which as the name suggests has the sole purpose of defending the First Amendment. The court's verdict was solely on the grounds of the Constitution and its case law.
Are you suggesting that the Constitution shouldn't apply to people who preferred the loser in an election? Or maybe that what the Constitution means or doesn't mean should be whatever the current government decides?
You're extremely naive if you've bought into the version of the world where everything that happens is entirely framed by partisan politics. The cause of the collision is: Trump is using social media in a completely unprecedented way, and the Constitution says a lot about freedom of speech. This has happened now because this is really the first time it could have happened.
I'm on the fence but I can understand the logic behind the conclusion.
The First Amendment speaks in absolute terms of making no incursions upon freedom of speech, and this case applies the usual sensible modification of that re: public forums. Any restriction is an incursion — that they could speak elsewhere (or elsewhere on Twitter) isn't a defence. The most controversial part of that to me seems to be in finding Trump's feed to be a relevant 'public forum'.
The court isn't compelling Trump to create a public forum, it merely found that he had already created one through use of his account for policy discussion. Having created one, his ability to exclude people is then subject to the law.
The difference between this and those kicked out of public meetings for being rowdy, obnoxious or rude is that the court was considering those excluded based on the content of their opinions, not their behaviour.
I'm not sure about that; the issue here seems to be that Trump's tweets and their replies count as a public forum because of the policy nature of their content, and that banning users from seeing his tweets while logged in limits their ability to take part in that public forum. So the breach isn't the not talking to them, it's the limiting of their opportunity to reply. It's the prohibition on their speech that's problematic, not their hearing.
I might be misunderstanding of course.
Biting the hand that feeds IT © 1998–2019