Wikipedia:Reference desk/Archives/Computing/2021 August 26

= August 26 =

Google Play Services automatic downloading
is it possible on Android to prevent Google Play Store and Play Services from automatically consuming data, through downloading, by switching off background data? or will Google continue downloading, even with background data switched off? Gfigs (talk) 08:16, 26 August 2021 (UTC)
 * as example, if Google Play Store starts automatically downloading 15MB, and reaches 2MB..and I switch background data off..will it continue downloading remaining 13MB? or download other items later on, when background data is off? Gfigs (talk) 11:51, 26 August 2021 (UTC)
 * I have two "Download Manager" system apps on phone..is that normal? Gfigs (talk) 03:22, 27 August 2021 (UTC)
 * also, how long after data is activated, will the downloads begin? as example, with an hour of data, will downloads start after 30 secs? Gfigs (talk) 03:32, 27 August 2021 (UTC)
 * is it possible to debug code, by rooting..a dongle..virus that installs system apps..installing new Android OS..email, banking, or buying new phones..xx? Gfigs (talk) 07:28, 29 August 2021 (UTC)

Gfigs (talk) 07:38, 29 August 2021 (UTC)
 * in 2018 The New York Times published the exposé, "How Google Protected Andy Rubin ,the 'Father of Android'".Google announced that "48 employees have been fired over the last two years" for sexual misconduct. in 2018 over 20,000 Google employees engaged in a worldwide walkout, as result. see History of Google ,2018 Google walkouts Gfigs (talk) 08:15, 29 August 2021 (UTC)

I used the apps settings thing to revoke network privileges for google play. I don't remember exactly how. But, just about every day it gives me an alert of some kind of google play failure, which I ignore, and everything works fine. I figure the block is working since it also keeps nagging me to upgrade the photos app, but then doesn't upgrade when I click the thing to allow it. I'm jaded enough about computers to understand that upgrades of that sort usually make the program worse. 2601:648:8202:350:0:0:0:2B99 (talk) 07:27, 1 September 2021 (UTC)

Excel
I suppose, it should be quite easy, but for some reason I am on a dead end. I would like to put into cell B4 a formula which results in the cell outputting the percentage change A5 to A4 (which should be 12%, if I am not too much of an idiot). Ideally, then I would just copy it upward so that I get in B3 50%. I appreciate your assistance. Cheers, 202.87.166.165 (talk) 12:00, 26 August 2021 (UTC)
 * Use = to start a formula. You'd enter =(A4-A5)/A5 into B4. When you copy that formula to B3, it will detect that the row of B3 is one less than the row of B4 and subtract 1 from the rows in your formula, resulting in =(A3-A4)/A4 in cell B3. 97.82.165.112 (talk) 13:23, 26 August 2021 (UTC)
 * Most brilliant! Thank you very much, I am much obliged. Very best regards, 202.87.166.165 (talk) 13:51, 26 August 2021 (UTC)

Cost of computer components
Hello! So after recently discovering my PC is actually fairly good, I went to Amazon to look at a potential GPU choice, but they're all super freaking expensive! Why have GPUs and other computer components become so expensive during the pandemic? I know that bitcoin mining generally makes use of GPU's but that can't be the only reason why. Blaze The Wolf &#124; Proud Furry and Wikipedia Editor (talk) 16:41, 26 August 2021 (UTC)
 * There is a global chip shortage, not just GPUs but all kinds. Graphics cards in particular are in high demand, so thats why you see the prices so high.  RudolfRed (talk) 19:19, 26 August 2021 (UTC)
 * TSMC (large chip maker) is raising prices due to high demand, so prices will be going up more: RudolfRed (talk) 20:02, 26 August 2021 (UTC)


 * There have been recurring shortages of GPU's for years before the current general chip shortage, because if GPU's get cheap enough, it becomes profitable to mine bitcoins with them, so they get bought up for that purpose, and that drives up prices again. 2601:648:8202:350:0:0:0:2B99 (talk) 05:56, 27 August 2021 (UTC)


 * Why do you want to upgrade your GPU anyway? If you were already a dedicated PC gamer, you would have bought a PC with a decent GPU for just that purpose. Most other uses of a PC simply don't need the computing power which an expensive GPU provides, except perhaps apps like graphics or video editing. All modern graphics cards provide completely adequate screen resolutions and frame rates for most purposes. A more powerful GPU won't make any difference at all to 95% of users, including just watching videos/movies, using a browser, creating documents with MS Office, etc. There are many brands of benchmark software, try some of these. MinorProphet (talk) 07:01, 29 August 2021 (UTC)
 * My PC technically doesn't actually have a GPU. It uses Intel's iGPU (Integrated GPU) which is the CPU basically being used as a GPU which works fine for most games on low graphics, however the CPU's iGPU isn't that powerful. Blaze The Wolf &#124; Proud Furry and Wikipedia Editor (talk) (Stupidity by me) 15:10, 30 August 2021 (UTC)
 * In other words, my graphics card slot is completely empty. Blaze The Wolf &#124; Proud Furry and Wikipedia Editor (talk) (Stupidity by me) 15:14, 30 August 2021 (UTC)
 * But what do you want your PC to do, if you feel it falls short in some department? If you are completely happy with its performance, leave it alone and spend the money on something else. If not, upgrade away, since you tend to get what you pay for these days: and those who trade on FOMO will enjoy their prophets profits. MinorProphet (talk) 05:30, 31 August 2021 (UTC)
 * I'm wanting my computer to play games with less lag. WHile it can currently run basically anything I throw at it (except Forza Horizon 4), it usually has to be on low graphics or I will get lag. I'm really just wanting a small performance upgrade because having an actual GPU will probably help me play games with less lag. I really just want my computer to be able to run games better than it can. Blaze The Wolf &#124; Proud Furry and Wikipedia Editor (talk) (Stupidity by me) 13:32, 31 August 2021 (UTC)

Please post the full tech specs of your PC, ie CPU, iGPU, RAM, motherboard (mobo) etc. If you are prepared to wave goodbye to between $200–$500, the lag may become less noticeable. Your results may vary. Beware: you may be setting out on a lonely path towards eternal dissatisfaction. MinorProphet (talk) 22:27, 31 August 2021 (UTC)
 * Ask your mum, or your dad, or both, if you are that lucky. Previous to your request, do the standard housework for a week without mentioning it, tidy your room, eagerly run errands as if it wasn't important, do your homework, put flowers in vases, say "I love you" occasionally. Result: 1 graphics card. MinorProphet (talk) 21:48, 2 September 2021 (UTC)

Generalist computer books
I'd like suggestions about books like Code:_The_Hidden_Language_of_Computer_Hardware_and_Software. Maybe something like this one but more modern?

That is, books that have a generalist focus, explain some seminal aspect of computing, are not overly technical, and are accessible to an interested beginner and might have a long shelf life. So, nothing like 'Java in 24 hours' or 'Java in one hour'. Bumptump (talk) 16:49, 26 August 2021 (UTC)
 * You want a long shelf life and a book about programming that isn't overly technical? How about 2 out of 3? The Art of Computer Programming currently has a shelf life of about 53 years. It is written for people who have an understanding of programming, but don't really know exactly how the processors and memory and everything works together. It primarily covers algorithms and how to program them because most of programming is implementing algorithms. The coolest thing about the book is that it isn't finished. Knuth is still writing it. 97.82.165.112 (talk) 19:11, 26 August 2021 (UTC)
 * I've read Volumes 1-3 of The Art of Computer Programming and cannot imagine how it would qualify as either "not overly technical" or "accessible to an interested beginner". Anyway, I'm sure it's widely available in libraries if you want to have a look. --184.144.99.72 (talk) 20:51, 26 August 2021 (UTC)
 * Algorithmics: The Spirit of Computing by David Harel and (in the third, not very recent edition) Yishai Feldman is less technical and more easily accessible. --Lambiam 22:09, 26 August 2021 (UTC)

Carl Malamud's "Exploring the Internet: A Technical Travelogue" (1992) is ridiculously old by now, but was very enjoyable and probably still somewhat valid. Is there a particular area of computing you want to read about? The algorithms books suggested above are probably not what you want (Knuth's in particular is fantastic, but extremely mathematical). This desk actually might not be the best place to ask for popularizations, since the regulars here are deeply enough into the subject that they mostly read more specialized works. Hmm, In the Beginning... Was the Command Line is also old, but covers the history of personal computing in an entertaining way. 2601:648:8202:350:0:0:0:2B99 (talk) 05:55, 27 August 2021 (UTC)


 * Tracy Kidder's The Soul of a New Machine is also quite entertaining. This peek behind the scenes is more sociological, though, than about computing per se; the team struggling to quickly bring a product to market could have been working on any highly complex device. --Lambiam 09:13, 27 August 2021 (UTC)

Lambian, yes, I figured OP wanted more "human interest" books than technical ones. Speaking of which, how could I have forgotten Hackers: Heroes of the Computer Revolution (1984), by Stephen Levy. 2601:648:8202:350:0:0:0:2B99 (talk) 03:56, 28 August 2021 (UTC)


 * I enjoyed The Cuckoo's Egg, the account of a specific instance of a computer security violation and what was done about it. It's somewhat dated in a number of respects, but I still think it's very readable.


 * A more obscure book that I also liked is Travels in Computerland by Ben Ross Schneider (not this guy, someone else of the same name), in which a literature professor has the novel idea that the project he wants to work on is best done using a computer. (Today it might well be routine to do that, but this happened in the 1970s and he starts with about zero knowledge of computers.)  Judging by the prices I see at Amazon and Ebay, the book is now out of print, but some libraries might have it. --184.144.99.72 (talk) 05:29, 29 August 2021 (UTC)


 * Brian Kernighan has written two books that might qualify: Understanding the Digital World and D is for Digital. (I haven't read either of them; there's a bit more information on his home page.) —scs (talk) 08:49, 29 August 2021 (UTC)

Instructure Set Architecture (ISA) monopoly
Dear Wikipedians:

I notice that for desktop/laptops we have PCs vs Macs. And that for tablets/phones we have Android vs iOS.

But in terms of ISAs, we have ONLY x86 for desktop/laptops, and ONLY ARM for tablets/phones.

I am wondering why such monopolies are allowed to exist for ISAs.

L33th4x0r (talk) 21:41, 26 August 2021 (UTC)


 * Holders of a monopoly are the sole suppliers of some commodity on the market. No player is the sole supplier of chips with an x86 or ARM architecture, so they are not monopolies. Also, The Coca-Cola Company is the sole supplier of Coca Cola. Does that mean they have a monopoly? --Lambiam 22:20, 26 August 2021 (UTC)


 * Thanks for your response. Upon reading it I realized that I didn't phrase my original question properly, allow me to rephrase it here: I am wondering why ISAs don't have the same "x vs y" that we see in terms of operating systems, in that there is only one ISA for non-mobile devices and only one other ISA for mobile devices. (Except for M1, but then it's going from x86 to ARM, so again x86 and ARM, no third ISA). L33th4x0r (talk) 03:20, 27 August 2021 (UTC)


 * The preference of ARM for mobile devices is related to their relatively low cost, minimal power consumption and lower heat generation – factors that are not equally important for non-mobile models. Manufacturers of mass-market products need to have a guarantee that chips will keep being manufactured in sufficient quality and quantity for the ISA they select, as well as continued quality support for software for that architecture, which may limit the choices. For example, Linux version 4.16-rc4 dropped support for several less popular architectures in 2018, including M32R. --Lambiam 08:54, 27 August 2021 (UTC)


 * Thanks for the detailed response and explanation. I just have one more follow-up question, the flip-side of your response, I guess: so why hasn't Windows and Android, which are much more dominant (by a huge margin) over MacOS and iOS, been able to "stomp out" MacOS and iOS for good in the same vein that ARM had been successful in "stomping out" rival RISCs such as MIPS and RISC-V? Thanks. L33th4x0r (talk) 16:04, 27 August 2021 (UTC)
 * Windows came very close to "stomping out" Mac in the late 90s. But, It's not clear Microsoft wanted to stomp them out. Microsoft was facing anti-monopoly investigations, and the existence of MacOS was part of their defense. Believe it or not, Microsoft actually invested $150 Million in Apple during their lowest period. (Apple_Inc.) ApLundell (talk) 17:21, 27 August 2021 (UTC)
 * Ha! You've hit on a very interesting point! If Microsoft faces anti-monopoly investigations for potential monopoly over operating systems for all computers back in the days, then why isn't Arm Ltd. facing anti-monopoly investigations for its monopoly over the ISA of all mobile devices (tablets, phones, watches, etc.) right now? L33th4x0r (talk) 03:38, 28 August 2021 (UTC)
 * This isn't really a computing issue but Microsoft never faced 'anti-monopoly investigations for potential monopoly over operating systems for all computers back in the days'. They faced an anti-trust investigation because of the belief they'd done illegal acts to obtain and preserve their monopoly. This relates to the fact that in a lot of the world, having a monopoly isn't in itself illegal, it's simply that you have to be careful of what you do when you have a monopoly or when you are becoming one. For that reason, the existence of a monopoly will often not by itself be sufficient to justify an investigation, but proving there is no monopoly or no dominant position may also help you. (There is a related issue that governments will often reject mergers if they will produce a monopoly or an excessively dominant position. One of the reasons why the proposed purchase of ARM by Nvidia is facing very heavy scrutiny, far more so than the earlier purchase by Softbank.) BTW, I don't think it's reasonable to say that ARM has been 'successful in "stomping out" rival RISCs such as MIPS and RISC-V' either. MIPS maybe, but RISC-V is still too new to know if it's going to succeed or fail. Nil Einne (talk) 04:29, 28 August 2021 (UTC)
 * BTW, while I particularly aware of ARM's legal history, Intel has faced some private and public investigation due to claims they were similarly abusing their dominant market position, notably the private case Advanced Micro Devices, Inc. v. Intel Corp.. While the cash was likely a help given AMD's position at the time, the settlement of the patent cross-licencing and fab dispute [//spectrum.ieee.org/intel-pays-amd-125-billion] was probably a bigger deal. It seems unlikely AMD would be in the position they are today if they still had to rely on their own fabs or Global Foundries, instead of TSMC. While this isn't legal advice and it's speculation without a source (although I suspect I could find one if I try hard enough) so technically a no-no, ARM likely has the advantage of what they do and their general open licencing terms. If the Nvidia buyout is approved, it's likely they'll find a lot of continued scrutiny given the fear they may start to prioritise their parent company over other possible licencees or other similar things. This isn't likely to be about the dominance of ARM over other architectures per se, but the existence of that dominance means there are likely to be greater concerns than would be the case if other architectures were much more common. Nil Einne (talk) 04:53, 28 August 2021 (UTC)
 * Thanks so much for this very clarifying explanation. So now I understand that it is not the monopoly per se, but rather what a company DOES with that monopoly, that may potentially become grounds for legal investigations.L33th4x0r (talk) 18:43, 28 August 2021 (UTC)


 * ARM Laptops are not unheard of. Quite a few Chromebooks use ARM, and there are a small handful of ARM Windows laptops.  However most Windows machines are going to all be x86 to maintain software compatibility with desktop systems. That's the real restriction.  However, as people become less dependent on local software in favor of web-based services, we may see more diverse architectures in laptops in the future. ApLundell (talk) 17:21, 27 August 2021 (UTC)
 * Apple is also set to use ARM for all their laptops and desktops sometime in the nearish future. Nil Einne (talk) 04:32, 28 August 2021 (UTC)
 * Thanks for the discussion. Excited to see ARM making further inroads into the PC market. I feel that ARM and RISC represents the future of computing and might eventually replace x86 which is CISC. L33th4x0r (talk) 18:43, 28 August 2021 (UTC)
 * Don't hold your breath. I can remember exactly the same claim that the future was RISC and not CISC when Alpha AXP chips were introduced by DEC to replace their VAX line 25 years ago. That's just reminded me: up until the change to Alphas a lot of code was still written at assembly level, RISC machines made assembly programming significantly more complex and the unofficial meaning of RISC was "Relegate Important Stuff to Compilers". Martin of Sheffield (talk) 19:06, 28 August 2021 (UTC)
 * You got a point there too. L33th4x0r (talk) 01:09, 29 August 2021 (UTC)