Wikipedia:Reference desk/Archives/Computing/2020 February 20

Welcome to the Wikipedia Computing Reference Desk Archives
The page you are currently viewing is a transcluded archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

February 20

Geek Squad Protection/Dell Warranty

Do any of these warranties cover missing keyboard keys? --Thegooduser Life Begins With a Smile :) 🍁 02:38, 20 February 2020 (UTC)

I don't know, but I have a funny story. When my daughter was a child, she pulled the arrow keys off my keyboard. Why? She said that she didn't need them! Bubba73 ^{You talkin' to me?} 02:43, 20 February 2020 (UTC)

Bubba73 For some reasons the backspace and ctrl key on my keyboard are "loose" Thegooduser Life Begins With a Smile :) 🍁 02:46, 20 February 2020 (UTC)

I don't see why not. Keys fall off / break off sometimes, or they get stuck or whatever. I've had to replace keyboards now and then, though they are usually old enough to be out of warranty by the time they are worn out or broken. 2601:648:8202:96B0:0:0:0:7AC0 (talk) 07:57, 20 February 2020 (UTC)

I don't know what kind of Dell keyboard you have, but I can usually find them for $1 apiece at the thrift store. Elizium23 (talk) 11:30, 20 February 2020 (UTC)

Digitized speech on the Commodore 64

The article Impossible Mission says:

The Commodore 64 version features early use of digitized speech: "Another visitor. Stay a while... stay forever!" and "Destroy him, my robots!" The digitized speech was provided by the company Electronic Speech Systems, who drastically raised their prices after Impossible Mission became a successful test case. Epyx did not deal with ESS again as a result. Caswell recounted:

I never met the performer but, when I supplied the script to the representative from ESS, I told him I had in mind a "50-ish English guy", thinking of the sort of arch-villain James Bond might encounter. I was told that they happened to have just such a person on their staff. When I was given the initial recordings, the ESS guy was apologetic about them being a touch hammy, but I thought the over-acting was amusing and appropriate, and they were left as is ...

Now, playing sounds on the Commodore 64 requires controlling the sound chip directly in machine code. There is no ready-made sample player. So how did Epyx incorporate the digitized speech into the program? Did they receive a recording which they scanned in as a sequence of sound values, and then wrote their own code which sort of looped through the values, playing each sound in sequence? JIP | Talk 11:26, 20 February 2020 (UTC)

I think they pretty much bit banged samples out to the SID with the CPU in a tight loop - that is, they set the SID to play a nominal tone, but then wobbled the volume level up and down to regenerate the sampled audio. When both Impossible Mission and Way of the Exploding Fist (which played a Kiai when it started) played their digitised audio, the rest of the action was stopped, suggesting the CPU was pegged out. -- Finlay McWalter··–·Talk 12:30, 20 February 2020 (UTC)

Yes, but where did they get the bits from? The secret sauce at ESS wasn't just the bitrate, it was the speech-specific compression algorithm, and an efficient algorithm for unpacking this. The Commodore 64 didn't have the space to go storing lots of sound, simply as sound-encoded and compressed bits, the speech was encoded as something more like phonemes, and ESS had an efficient decompressor which worked for sounds-like-speech much more efficiently than it could for sounds-in-general. Andy Dingley (talk) 14:29, 20 February 2020 (UTC)

Seeing as Impossible Mission can play the speech samples straight out of the box, does this mean that the code for Impossible Mission actually incorporates this ESS decompressor? Seeing as the Commodore 64 has only 65536 bytes of addressable memory, Epyx must have co-operated with ESS about exactly where to place the code. JIP | Talk 14:32, 20 February 2020 (UTC)

Yes, space is certainly an issue. Listening to the playback, there's a warbly wobble to it, which is reminiscent of a linear predictive coding scheme at a very low bitrate. Plus both games have very little sampled content - IM has maybe 3 pieces totalling perhaps 10 seconds, and WotEF had a single sample, which I'm pretty sure played only during loading (meaning they could then load over it), and some punch and kick noises that may have been sampled. -- Finlay McWalter··–·Talk 14:40, 20 February 2020 (UTC)

• AIUI, ESS charged for two services: encoding the speech (which was done by witchcraft and hand-whittling the bytes to sound best), and then licensing some code to be embedded in the game. After the first games sold more volume than expected, ESS changed their pricing model. Again, AIUI from no basis, this meant that future games became very expensive to license this, unless they also sold in similar volumes. So the game developers get scared at having a modest success, then making nothing from it. Andy Dingley (talk) 17:59, 20 February 2020 (UTC)

I don't know anything about this ESS company but I doubt that they had anything really secret. There is/was a vast literature about low-complexity speech coding and synthesis and plenty of known techniques to choose from. "Complexity" in this context refers to how much CPU power the algorithm consumes. Linear predictive coding (LPC) at low bit rates characteristic warble that Finlay McWalter refers to, so if the ESS sample had that sound, LPC is probably what they were doing. E.g. LPC-10 is an old standard that already used intricate bit patterns to save bandwidth and maybe they used something like it, adjusting by ear to get particular clips to sound acceptable. 2601:648:8202:96B0:C8B1:B369:A439:9657 (talk) 20:56, 22 February 2020 (UTC)

According to Forrest S. Mozer and ESS Technology, ESS's work at the time often relied on part on Mozer's patents. (This may also be of interest. [1]) While I gather quite a few of these e.g. US4214125A [2] and US4458110 [3] are more about speech synthesis, it does seem not all of them are just about speech synthesis e.g. US4433434A [4] and US4435831A [5] and US4314105A [6]. And from what I understand given the techniques of the time, the distinction between synthesis and a compressed recording were not always that clear cut.

While I don't know that any of ESS's work for games relied on these patents, and software patents even Software patents under United States patent law at the time were less clear cut, it seems easily possible that while ESS's work did not really rely on anything secret, it did rely on stuff others could not legally do without licencing these patents. Which unless they did before the foundation of ESS, may not be possible if it was felt they would be competing with ESS, without paying a substantial amount.

This [7] claims all the speech including the software was 6 kilobytes. From [8], I gather that the total amount of speech was about 20 seconds with 6 different sounds (not the 3 suggested above), so about 300 bytes per second which doesn't seem any better than LPC of the time, although it did include the software. Our article suggests 2 kB of RAM for LPC-10 so I assume the software can't be larger than that, and I assume this is also for software which can encode. So bit rate was maybe nothing spectacular.

But our article also claims 20 MIPS, which seems very high compared to the 0.43 MIPS for the MOS Technology 6502 (used in the Commodore 64). Again I assume this is the requirement for encoding and decoding will be significantly less. It's also possible that the instruction speed for LPC-10 is a lot higher although I somewhat doubt it on non specialised hardware. So while I don't know how much decoding costs with LPC-10 but especially as I doubt that the whole CPU was dedicated the the speech during playback (even with the action stopping, I assume some CPU was still needed for other stuff), I wonder if LPC-10 is significantly too complex. Or to put it a different way, there are different levels of "low complexity", especially when we're talking about non-specialised hardware of that era. Mind you the 20 MIPS is uncited, and it does seem very high given systems at the time, even considering specialised hardware but then again, it was for military applications.

BTW, I came across [9], but I have no idea if the person actually knows what they're writing about.

Nil Einne (talk) 02:07, 23 February 2020 (UTC)

Should add that besides the processing cost of decoding LPC, you'd also need to consider the cost of generating the output. My understanding of the subject matter is way too limited to be able to estimate, but I'd guess the processing cost of bit banging the output using the SID is not so tiny to be irrelevant. Nil Einne (talk) 06:10, 23 February 2020 (UTC)

Decoding LPC10 shouldn't need anywhere near 20 mips, though encoding it might use that much. See also formant synthesis which is generally LPC based. You play a pulse train through a digital filter that approximates the frequency response of the vocal tract during the phoneme you are synthesizing, and change filter parameters 20x or so per second, IIRC. I don't know if the cpu requirements are in reach of a 6502 though. I'm sure there were lots of optimizations and tricks in use in that era. Various speech-synthesizing toys like the TI Speak and Spell were being made at the time under severe cost constraints. The Speak and Spell in particular was supposed to have been quite a technological feat for the time. It may have used one of those newfangled DSP's though. 2602:24A:DE47:B270:A096:24F4:F986:C62A (talk) 05:57, 24 February 2020 (UTC)

Laptop battery does not charge - fault of battery or AC adapter?

I been recently getting a new popup msg whenever I turn my computer on, at least from stand-by, that says "The AC power adapter type cannot be determined. Your system will operate slower and battery will not charge. This problem might be solved by: Connect a Dell 65W AC adapter or higher." Dell laptop Windows 7. Thanks. 67.175.224.138 (talk) 11:37, 20 February 2020 (UTC).

Have you... inspected the AC adapter and cables for damage? cleaned the contacts? tried to use another AC outlet on a different circuit? removed the battery if possible, inspected it for damage and dirt? tried operating without the battery? considered upgrading from Windows 7 (being end-of-life)? Elizium23 (talk) 12:45, 20 February 2020 (UTC)

• Dell Inspiron? There's one model of these that's infamous for this fault. Poor quality (brittle) plastic in the case made the cases crumble, particularly around the sockets. The power socket comes loose and no longer connects. It connects well enough to give a power connection, but not for the additional data connection needed to identify the computer and charger to each other, so that they start charging (rather than just powering the computer). It may be repairable by dismantling and hot glue, or just by very careful insertion of the plug. Andy Dingley (talk) 12:48, 20 February 2020 (UTC)

Yep Dell Inspiron. Would the battery not charging also cause the laptop to heat up quickly? And then I hear a loud motor sound, then put my laptop on stand-by to let it cool a little. 67.175.224.138 (talk) 18:59, 20 February 2020 (UTC).

• The loud motor is probably the fan, reacting to high temperature on the CPU or motherboard. I would suggest installing a simple hardwaare viewer like Speccy[10] which will show you the temperature.

High temperature is often caused by dust buildup around the air vents, or using it resting on a cushion. Laptops do need internal cleaning sometimes. Andy Dingley (talk) 19:43, 20 February 2020 (UTC)

I feel like there is a fan noise and a motor noise separate from the fan. Uh oh, I downloaded Speccy:

CPU 52-55 C

Motherboard 54 C

Storage 38 C. 67.175.224.138 (talk) 20:03, 20 February 2020 (UTC).

The other motor noise would be the hard drive, if your laptop has one. And if the drive has gotten noisier over time, it may be heading towards failure. Replace it with a solid state disk. You'll like the SSD a lot better than the HDD since besides being silent, it makes a night and day difference in the laptop's responsiveness. I'm quite happy using laptops that would mostly be considered obsolete, since SSD's still make them pleasant to use. 2601:648:8202:96B0:0:0:0:7AC0 (talk) 21:50, 20 February 2020 (UTC)

Also, those temperatures are warm, but if the cpu is busy they are not really excessive. Stuff does get heated up, and laptops are limited in airflow etc. Figuring that laptops aren't really built for heavy computation, I generally prefer to use remote servers when I have compute-intensive things to do. 2601:648:8202:96B0:0:0:0:7AC0 (talk) 06:38, 21 February 2020 (UTC)

Yeah these motor noises are bothering. I kept my laptop in a freezer last night, and turned them on this morning, the coldest the motherboard is 45 C. Then rebooted my computer, it started the mother noise when it is 54 C again but only after a few minutes... The only difference now is the motor noise is not as long, and is often softer in sound. Somehow I think the thermometer is off. 67.175.224.138 (talk) 13:53, 21 February 2020 (UTC).

• " I kept my laptop in a freezer"

Great way to kill it (in a few ways, condensation being the first). 55 degrees isn't bad. Andy Dingley (talk) 14:16, 21 February 2020 (UTC)

Since no one mentioned this, clean the air vents with gas duster, which is good to do periodically. Dust builds up over time. Ensure that you have backed up any data you care about. As long as you have data backed up, you don't need to worry. As others have suggested, if you want to be proactive, you could replace the hard drive with a SSD, except if the charging port is messed up, you might just want to get a new laptop that comes with a SSD. --47.146.63.87 (talk) 21:07, 22 February 2020 (UTC)

I notice now if I tilt my laptop a certain way, it prevents the motor-sound from even starting. Sigh. 67.175.224.138 (talk) 14:16, 25 February 2020 (UTC).

I hate to bring it up every time, but it is important to remember that Windows 7 is at end-of-life and gradually becoming unsafe to run as an Internet-connected device. Now is a good time to consider upgrading to a supported operating system on modern hardware. Elizium23 (talk) 14:20, 25 February 2020 (UTC)

Indeed, but you can upgrade an existing Windows 7/8 computer to Windows 10 for free. --47.146.63.87 (talk) 19:20, 25 February 2020 (UTC)

There isn't any problem unless the fan is not spinning up and causing the laptop to overheat. You can monitor fan speed and system temperature with lots of free tools. The temperatures you posted above are not concerning, and modern computers shut off automatically if the temperature gets dangerously high. The noise might be from the fan being loose, or it could have dust buildup, or the bearing has lost lubrication. You might want to open up the laptop and give it a good inspection and cleaning. Dust does build up over time. The fan can be replaced if absolutely necessary. As annoying as it can be, anything with moving parts does experience wear and tear. --47.146.63.87 (talk) 19:20, 25 February 2020 (UTC)

Smash the NANP?

Well not really. And not seeking legal advice-- this is a theoretical question. If I had the resources to attempt it in practice I'd also obviously hire lawyers as part of it.

I'm wondering what it would take for a US phone carrier (or consortium of them) to introduce new phone number ranges outside the North American Numbering Plan, i.e. something like ipv6 for phones. There would obviously be technical, regulatory, political, and other hurdles but maybe it is time for someone to pursue this. NANP numbers are in short enough supply (especially in crowded area codes) that they are re-used all the time, leading to misdirected calls; and there are few enough of them for spammers to war dial them. Meanwhile, nobody really manually dials a given phone number more than once (after that, the phone remembers it) That makes long phone numbers less of a problem than they might once have seemed.

So I'd like something like a 20 digit phone number or range of them. Working numbers of such size would be sparse enough that it would be hard for spammers to find them by scanning. It would have to be callable from normal POTS phones without incurring extra costs, but it's ok if it had to be prefixed with 011 and a new country code, or something like that. Any idea what this would take? NANP (like ipv4) is heading for exhaustion anyway, and going "big" seems better than adding another digit. It would allow people wanting NANP numbers to still have them if they preferred and were willing to pay for them (they are around $1/month each from VOIP carriers, though I don't know how much of that is for the number itself). Thanks. 2601:648:8202:96B0:0:0:0:7AC0 (talk) 14:33, 20 February 2020 (UTC)

Is it really heading for exhaustion, or can more area codes be added where needed? Elizium23 (talk) 00:09, 22 February 2020 (UTC)

North American Numbering Plan expansion says "An April 2019 analysis estimated that the current numbering plan would not be exhausted until after the year 2049". -- Finlay McWalter··–·Talk 00:18, 22 February 2020 (UTC)

In terms of the technical capability of the telephone system, it's doable. It's all computers now; there are no more mechanical telephone switches. Some time and effort would be needed to prepare and ensure everything works as intended, of course. The main friction point will be stuff out there in the world that assumes all phone numbers are five digits plus three-digit area code. This has never been a good assumption, because calls outside of the NANP area will vary, but I'm sure there is equipment, software, etc. that does it anyway. As noted, doing this has actually been proposed for the NANP. With that said, the way the Internet has become "the network", we could just as well move to giving people their own IPv6 addresses. There's no fundamental reason why people can't have personal IPv6 addresses and associate them with personal domain names, so that I could just punch in your name and initiate a voice conversation with you. Technological, political, and social factors are what impede this. In cell phones, 4G and later is all IP-based already. And of course there's the question of what we want to do with the existing POTS. --47.146.63.87 (talk) 21:28, 24 February 2020 (UTC)

Windows 10 emptying recycle bin

When emptying the recycle bin, I typically right click and select the "empty recycle bin" option from the menu. It's muscle memory at this point. All good. However, I recently installed Jasc PaintShop Pro on my machine. It's more than a decade old at this point, but I needed a graphics program that I could use immediately with no learning curve and PSP is actually a fairly robust program and I'm familiar with it. Again, all good, apart from the help files not working correctly in W10. What I need help with is that the option to browse my recycle bin folder within PSP is now part of the right-click context menu and it's sitting in the position where the "Empty..." function used to sit so I keep on hitting it. Even weirder, I don't think the option to browse via PSP shows up anywhere else, though I haven't done an exhaustive search. Is there a way to remove this functionality? I'm assuming this is a registry hack, but I haven't attempted that since the days of XP so I'm not going to just muck around in there without direction. Matt Deres (talk) 16:06, 20 February 2020 (UTC)

Tried this? https://www.online-tech-tips.com/computer-tips/windows-right-click-context-menu/ 2003:F5:6F09:F300:4DD9:B04C:12A:4FB0 (talk) 17:30, 21 February 2020 (UTC) MPB

That would seem ideal, but I tried changing each of the entries (except the ones I could easily identify as not PSP) and none of them removed the line from the context menu. In fact, I'm not sure any of them made any difference at all, even though I followed the directions closely. Matt Deres (talk) 20:01, 22 February 2020 (UTC)

Whatsapp problem

Hi everyone, having a problem with Whatsapp and I am hoping someone can help, I've done a bit of googling and had a look at the FAQ section of whatsapp website but I am still stuck.

I would like to transfer my chats from my old mobile phone to a new one. Both phones are android and I am in the UK. The old phone is some old type of samsung which I have had for about 5 years. I have used WhatsApp on the old phone without any issue. The new phone is a moto G7. I have backed up conversations to google drive on my old phone. I have put my SIM from old phone into the new phone as I want to keep the same number. I have logged in with my old google account email address on the new phone. When I downloaded whatsapp to the new phone, none of my old conversations came up. Messages that people send me only go through to my old phone, which now doesn't have a SIM.

Not sure what the issue is as there is no change in number or google email address, merely a change in phone handsets.

Any help would be much appreciated! RichYPE (talk) 21:14, 20 February 2020 (UTC)

If you did not back up your Whatsap messages then your new phone won't receive the old ones. They are encrypted and this encryption is coupled to the hardware AFAIK. If the old phone still works you can use that phone to back up and then load these per https://faq.whatsapp.com/en/android/28030001/ and https://faq.whatsapp.com/en/android/20887921. Your new messages should arrive on your new phone, which is coupled to the phone number. If it does not, the system might not have updated yet, you could force this to happen by shutting down your old phone and sending a message using the new one. Rmvandijk (talk) 13:52, 21 February 2020 (UTC)

Thank you for your response. I have an update. The phone numbers on the 2 phones were different. However, the number of my account is no longer correct, it has a very old phone number from a previous SIM card. I uninstalled Whatsapp on the old phone and I am now no longer able to change my account's number as it sends a verification code to a phone number that I no longer have. Is there anything I can do to resolve this? Thank you!RichYPE (talk) 16:42, 23 February 2020 (UTC)

If you don't have access to your old phone number, you are out of luck. As the FAQ mentioned before says, you need to have access to both new and old phone for a number change and your backup can only be synced to the number they were made on (not sure how this combination works, but it probably says what to do on the FAQ's). Not sure how you managed to have a different number with the same sim-card though? Rmvandijk (talk) 07:58, 25 February 2020 (UTC)