Talk:Hard disk drive/Archive 26

External links modified
Hello fellow Wikipedians,

I have just modified 3 external links on Hard disk drive. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20091122075109/http://sdd.toshiba.com/techdocs/MKxx33GSG_MK1235GSL_r1.pdf to http://sdd.toshiba.com/techdocs/MKxx33GSG_MK1235GSL_r1.pdf
 * Added archive https://web.archive.org/web/20110105162632/http://www.wdc.com/en/products/products.aspx?id=140 to http://www.wdc.com/en/products/products.aspx?id=140
 * Added archive https://web.archive.org/web/20150226053423/http://www.recover.co.il/SA-cover/SA-cover.pdf to http://www.recover.co.il/SA-cover/SA-cover.pdf

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 03:47, 30 October 2017 (UTC)

Time wasting article
Even Wikipedia editors can't be so colossally vain as to think the readers are too stupid to get the idea an external drive can be more than one color, without a picture. There is no point to including commented-out contents in the article, the readers don't see it and it makes editing more clumsy. I don't know why the multiple reverts, this article is grossly over long and needs to be trimmed down and less of a collection of press release clippings. --Wtshymanski (talk) 04:13, 30 October 2017 (UTC)


 * The other image I assumed was meant to show it as a variety but I do agree 2 images is pointless, The commented-out refs are fine and can be used - I've tried to insert these in to the article and each time it's adding more errors, I disagree the article isn't long at all and as I said on your talkpage the length is fine. Thanks, – Davey 2010 Talk 13:08, 30 October 2017 (UTC)


 * Acording to WP:TOOBIG at Prose size (including all HTML code)= 63 kB (as measured by Prosesize) this article "Probably should be divided." I agree with Wtshymanski that the article needs to be trimmed and suggest he be allowed to proceed.  Tom94022 (talk) 17:46, 1 November 2017 (UTC)
 * I'm being commanded to go to the talk page, and yet I don't see here any reason to retain all the commented-out text. --Wtshymanski (talk) 23:41, 1 November 2017 (UTC)
 * So we're not following WP:PRESERVE no ? ... Fuck it done. – Davey 2010 Talk 00:07, 2 November 2017 (UTC)

Older interface cables
User:Wtshymanski insists upon this statement which is incorrect in several ways:

Going into this much detail seems TMI so I propose the simple and true sentance now in the article:
 * 1) The data signals on the data cable of the ST-506 and other such interfaces are binary not analog and there are other signals on the cable that are binary and digital.
 * 2) The cables sent sent signals to and from the drive.
 * 3) Not all older interfaces had a data separator in the controller; notably Diablo, SMD and ESDI drives contained data separators, so citing ST-506 characteristics without clarification is misleading.

It avoids all of the above issues; any other simple constuction would be appropriate but an incorrect and misleading sentance should not be allowed Tom94022 (talk) 23:04, 21 November 2017 (UTC)
 * This is Wikipedia, the HOME of TMI. We go into everything but the screw hole thread specs so one sentence explaining the difference of the "old days" is hardly too much information. There is a huge difference between the ST506 style interface (in particular) and modern all-digital, block-oriented,  "smart" drive interfaces.  (The ST 506 approach was more like telling the printhead when to fire the pins and when to move the paper, as opposed to just sending an "A" down the wire.) The "sentence" is correct - the difference is the data separator was on the controller card, not part of the drive. The signals were not nice clean 1's and 0's and required considerable analog processing at the controller to turn into binary data. If anyone needs to know all the pinouts, we've got them in the linked ST-506 article.  The importance for this article is to stress how modern drives have off-loaded the details of disk control from the host. --Wtshymanski (talk) 00:28, 22 November 2017 (UTC)
 * We agree that one sentance is enough, but it needs to be accurate. The data separator is only one small piece of what was moved from the controller and so it is incorrect to ifocus on it and thereby imply that all older interfaces had a data separator in a controller. What about the ECC encoder/decoder, serializer/deserialiser, formatter/deformatter and the write encoder/precomensator and the read decoder.  SMD which preceeded ST-506 is just one example of an HDD with a data separator in the drive and it was just as dominant in its time period as the ST-506 was in its time period.  All of the signals on the ST-506 data cable were in fact nice clean 1's and 0's - you don't have to believe me - from the MFM article,

"MFM is a modification to the original digital FM (digital frequency modulation also known as delay coding) scheme for encoding data ..." (emphasis added).
 * The signals on the data cable other than the read signals were noise free digital; the read signals were digital with noise in the transitions. It is just plain wrong to call them analog - they are analog at the head and all drives convert them to digital.
 * I don't have a problem with a longer sentance containing more informationbut if we are going to generalize the earlier interfaces it needs to be accurate, something along the lines of:
 * The earlier interfaces had a separate data cable to transfer bit serial data to/from the device and its controller with the controller containing both the write circuits necessary to convert parallel data from the host into bit serial data to the drive and the read circuits necessary to convert bit serial data from the drive into parallel data to the host; in intelligent drives all this circuitry is moved from the controller to the drive, sending both data and control over one cable and eliminating the need for a data cable.


 * This seems like TMI but it is accurate, so if you want to try to improve the article go ahead, but please be accuate. Tom94022 (talk) 08:23, 22 November 2017 (UTC)


 * Please look at the connectors for ESDI - the signals are more or less the same as ST-506 but an ESDI drive has the data separator in the drive. It is not an "intelligent" interface as u define it.  This is perhaps the best example I can give as to why the challenged edit is at best misleading.  Tom94022 (talk) 18:05, 22 November 2017 (UTC)
 * ST-506 is not ESDI. So if ST5-6 was sending digital data to the controller, why did they move the data separator in ESDI? --Wtshymanski (talk) 19:05, 22 November 2017 (UTC)
 * Sure ESDI is not exactly ST506 but it is very close; if you compare the interfaces ESDI is a superset of ST506; the interfaces are physically identical and the common signals are logically the same except that the data rate is up to 20 MHz NRZ rather than 5 MHz MFM (or 7.5 MHz for ST-412HP).  There wer a lot of bad data separators so the ANSI committee dominated by drive companies required noiseless read signals to avoid compatibility problems. Moving the data separator did not make ESDI intelligent (just less dumb) nor did it eliminate the data cable.  Digital signals can have phase noise but they are still digital.
 * BTW there are 5 signal pins in the ST-506 data cable, three are noiseless and two are noisy, +MFM Read and -MFM Read. So your phrase "analog data signals" is at best misleading since three of the five data signals on the data cable are noise free digital.
 * Perhaps you might try to explain why you are ignoring all the other functions left on the controller that operated thru the data cable and just focusing on the data separator. FWIW, having designed many controllers, the data separator is a small part of the development effort.  Tom94022 (talk) 21:34, 22 November 2017 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on Hard disk drive. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20121014191631/http://faculty.plattsburgh.edu/jan.plaza/teaching/papers/seektime.html to http://faculty.plattsburgh.edu/jan.plaza/teaching/papers/seektime.html

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 12:27, 22 December 2017 (UTC)

HDD market vs SSD markets
An IP is asserting that SSDs now dominate the "market" referencing an Register article NAND that's that... Flash chip industry worth twice disk drive biz. This Register article is about the industries and not relevant markets. Flash competes in many markets while HDDs are essentially in one market, that is secondary storage device for general-purpose computers - it is inapposite (apples vs fruit) to draw any conclusion from relative industry revenues. In the relevent market revenues are about equal and HDD capacity shipped is about 10 times that of SDD. Clearly SSD is eroding HDDs market share in the relevent market but at least based upon capacity shipped, HDD still dominates. Tom94022 (talk) 17:57, 23 May 2018 (UTC)


 * Also, The Register's status as a WP:RS is dubious at best. Jeh (talk) 18:23, 23 May 2018 (UTC)


 * The user "Tom9402" has misquoted the reference, which is an article in The Register. Please, nota bene. The user "Jeh" has launched an ad hominem attack upon the reference, which is the publication The Register. The facts are real; deal with them. NAND is actually a super-set of SSDs. According to the headline of the article, NAND now dominates the market, "flash chip industry worth twice disk drive biz." The relevant market here is that for computing platforms. Since 2016, mobile (mainly smartphones and tablets) has been the dominant computing platform worldwide. 166.107.163.254 (talk) 00:43, 25 May 2018 (UTC)

PCs no longer the most popular computing platform
Usage share of operating systems shifted in 2016. Popular computing platforms use NAND secondary storage, not HDDs. — Preceding unsigned comment added by 166.107.163.254 (talk) 00:33, 25 May 2018 (UTC)
 * This is totally irrelevent to the question as to markets for SSDs or HDDs. You might also read Memory heirarchy. Tom94022 (talk) 00:43, 25 May 2018 (UTC)
 * The market revenue for SSDs is slightly larger than that for HDDs, as noted in the reference. You might also read Usage share of operating systems. 166.107.163.254 (talk) 00:46, 25 May 2018 (UTC)

Hard disk drive &equiv; platter drive?
The first sentence currently says that hard drives are in all cases platter drives. To disambiguate the article's disposition, I suggest renaming the article to "platter drive", since this covers portable platter drives as much as fixed platter drives. Other articles use "hard drive" (linking this article or not) as shorthand for essentially "secondary storage that cannot be easily removed from the device that usually accesses it." In that group is also SSD, etc. Yes, it's absolutely different technology, but the end user experience is largely the same even if performance varies. D. F. Schmidt (talk) 18:14, 15 June 2018 (UTC) D. F. Schmidt (talk) 18:14, 15 June 2018 (UTC)


 * No. That would violate WP:COMMONNAME as well as WP:NOR. The term "platter drive" never occurs in this article and afaict it is only used with any frequency in descriptive phrases such as "single-platter drive" or "three-platter drive". We don't make up names for things; we follow the nomenclature used in reliable sources. Jeh (talk) 18:29, 15 June 2018 (UTC)


 * I can understand that part of it, but you didn't address the matter that other articles reference this article as a catch-all for platter drives, SSDs, SSHDs, and whatever other permanent storage drives exist. D. F. Schmidt (talk) 16:36, 18 June 2018 (UTC)


 * No they reference this article because they all appear to the system as HDDs. If u think the references in the other articles are inappropriate then perhaps you should try to edit them out there.  Tom94022 (talk) 16:48, 18 June 2018 (UTC)


 * Hard drive redirects to this page. Do you really think everything that links to hard drive is pointing specifically to platter drives? For example, see Hierarchical File System and Paper shredder. D. F. Schmidt (talk) 00:15, 19 June 2018 (UTC)


 * They shouldn't link to a redirect in the first place. Again - that is the other articles' problem. Your suggestion is simply not supported by reliable sources, so it's not going anywhere. Please WP:DROPTHESTICK. Jeh (talk) 05:05, 19 June 2018 (UTC)


 * Please don't be a dick. I'm suggesting that something should be fixed about this situation. What should those articles be referencing instead of hard drive -> hard disk drive? Should hard drive be a disambiguation page that lists all the various devices that have permanent storage, or else redirect to an article that covers these at large? D. F. Schmidt (talk) 16:41, 19 June 2018 (UTC)


 * There is no need to address that matter in order to know that "platter drive" will not be accepted as a move target here. That other articles use this article's name when they mean other things is their problem. Jeh (talk) 19:47, 18 June 2018 (UTC)

Type of arm vertical movement versus comb
There was a significant technological change from the 1950s to the 1960s that should be addressed in the lede or in Hard disk drive. The IBM 350, 355, 1405 and 7300 had arms that moved both horizontally and vertically; a seek could take over half a second. With the 353 and 1301, IBM introduced the comb arrangement used by all subsequent moving-head disks, with one R/W head per recording surface (including synchronization surfaces). Shmuel (Seymour J.) Metz Username:Chatul (talk) 20:16, 26 December 2018 (UTC)
 * It is already addressed in the history section of this article (the 350 had 50 disks and 2 heads while the 1301 used one head per 50 disks). I'm not sure any more detail is necessary but if you want to revise or add to the history section please do so. IMO, this change does not rise to the level of the lede.  Tom94022 (talk)

Dual actuator drives
The article states that "Relatively new technologies like HDMR, HAMR and MAMR, Bit patterned media and dual independent actuator arms increase the speed and capacity of HDDs and are expected to make HDDs more competitive with SSDs.". It's a small point but dual independent actuator arms have been around since 1994 - Conner_Peripherals released their 'Chinook' dual-actuator drive in 1994. Loweredtone (talk) 14:23, 8 August 2019 (UTC)
 * Actually multiple actuator drives go back well before Chinook - all the way to the RAMAC which as I recall had provisions for 8 actuators and versions shipped with two. The Chinook was annnounced and a few shipped but it was quickly withdrawn so it is not a good example.  IBM had some dual actuator drives into the 90s but I think it is safe to note that there haven't been any dual actuator HDDs annnounced until recently (this century) which makes "Relatively new" relatively accurate and not worth changing.  I suppose a clarifying note might be acceptable but I think even that is TMI.  Tom94022 (talk) 17:20, 8 August 2019 (UTC)
 * Maybe you can replace: "Relatively new technologies like..." with either:
 * 1. "Recently developed and/or revived technologies like..."
 * 2. "Recent developments in technologies like...", Or
 * 3. "The revival of dual independent actuator arm functionality, paired with newer technologies like HDMR, HAMR, MAMR, and Bit patterned media increase the speed and capacity of HDDs..." Travisryno (talk) 09:32, 2 July 2020 (UTC)
 * The language was revised quite a while ago and is no longer misleading. Tom94022 (talk) 17:55, 2 July 2020 (UTC)

Form factor SI conversion
has added convert templates to form factor designations. My understanding is that these aren't really useful measuments but are more the name of the form factor. If so, I don't think the addition of an approximate the SI equivalent is helpful. I find that this has been discussed on this talk page at least once before but I don't see a clean resolution. ~Kvng (talk) 14:40, 27 July 2020 (UTC)
 * I think it was resolved in that there was no further response. I agree that all the form factors are names and not measurements.  Depending upon the generation some form factors were originally dimensioned in inches and others in metric but that has nothing to do with the names.  so I am going to revert Tom94022 (talk) 05:24, 28 July 2020 (UTC)
 * Thanks for verifying changes. To me it seems the linked earlier discussion didn't come to a conclusion. Even if the form factor doesn't define strict dimensions, it is still a name including a unit of length giving an indicative size of the object. To many readers, complementary metric units give a better understanding, even if the given unit is not an exact measure. I might add more conversion to this article, initially avoiding names and form factors, but adding conversion to measurements of length given in non-metric units only. -- Ws1920 (talk) 09:11, 28 July 2020 (UTC)
 * I'd have to research it to be sure, but AFAIK the 8-inch and 5¼-inch form factors were natively dimensioned in inches while the 3½-inch and 2½-inch form factors were natively dimensioned in metric units. The disk diameters I think also started out dimensioned that way but when they were subsequently dimensioned by standards bodies (e.g. ANSI) I seem to recall the went to native metric units.  So in adding conversions we need to be careful as to which system was native.  Tom94022 (talk) 16:07, 28 July 2020 (UTC)

A Commons file used on this page or its Wikidata item has been nominated for deletion
The following Wikimedia Commons file used on this page or its Wikidata item has been nominated for deletion: Participate in the deletion discussion at the. —Community Tech bot (talk) 00:01, 9 November 2020 (UTC)
 * Harddrive-engineerguy.ogv

Images
There is currently an image of a contemporary SATA magnetic disk drive. Does anybody have an unencumbered image of an IBM 350 or IBM 355 that they could link to for contrast? Shmuel (Seymour J.) Metz Username:Chatul (talk) 15:17, 9 December 2020 (UTC)
 * This IBM Image of the 350 can be used per IBM's usage policy, "Users may download these files for their own use, subject to all applicable federal and state statutes. Fair educational use of the contents of this Web site is permitted. Specifically, an internet user may download images and text files and share them with others for limited noncommercial and educational purposes only." I would upload the image to WikiMedia giving IBM full credit per their policy and then copy the image into any article.  Alternatively if you want to make a collage with a current drive, do so, upload to WikiMedia as your own work giving IBM credit for the 350 part of it.  Tom94022 (talk) 20:33, 9 December 2020 (UTC)


 * Thanks. Doesn't that run afoul of Note that images that are licensed for use only on Wikipedia, or only for non-commercial or educational use, or under a license that doesn't allow for the creation of modified/derived works, are unsuitable. at Image use policy? Shmuel (Seymour J.) Metz Username:Chatul (talk) 23:47, 9 December 2020 (UTC)
 * I'm not a lawyer, but since the IBM license doesn't explicitly exclude such a usage, a collage is consistent with their stated intent would be permitted. OF course you could always ask IBM for permission to use it by sending an email to Archive1@us.ibm.com. Finally, your original request was for a link - you have it, don't u?  Tom94022 (talk) 16:09, 10 December 2020 (UTC)


 * Thanks - I sent them an e-mail.
 * To put it in context, the article currently has a picture of a contemporary SATA disk with text describing the invention and patenting of magnetic disk drives. I'd like to strip off that text and put it on a picture of an IBM 350. Shmuel (Seymour J.) Metz Username:Chatul (talk) 16:44, 10 December 2020 (UTC)
 * It would be great to have a person holding the SATA drive in her hand next to the 350 - when u get permission and if you want such I can try to take a hack at it Tom94022 (talk) 07:06, 11 December 2020 (UTC)
 * I found an image from another article, and it even has a hand for you. Shmuel (Seymour J.) Metz Username:Chatul (talk) 16:22, 11 December 2020 (UTC)

Control
It would be useful to know where control data like SMART data is stored - is it on the disk or within the electronics? If it is in the electronics, the platters and the electronics are matched, if not one can be replaced without the other. Can't find information about this on the Web. Chris.Bristol (talk) 18:00, 7 January 2021 (UTC)
 * This isn't the place to ask questions about the subject, but I'll answer anyway. In modern drives, the disk and electronics are fine-tuned to each other during manufacturing; if only one is replaced, the drive likely won't work anymore, regardless of how this is stored.  --A&#8239;D&#8239;Monroe&#8239;III(talk)  01:27, 17 February 2021 (UTC)

SAS vs. SATA connector
SAS uses a mechanically identical data and power connector to standard 3.5-inch SATA1/SATA2 HDDs...

That is not true. While the (combined data and power) plug on an SAS host adapter interface cable can be attacheded to a SATA drive, the reverse is not possible.


 * 216.152.18.132 (talk) 18:08, 11 February 2021 (UTC)


 * Are you sure? Just based on what I've read elsewhere, it is true; the reverse attachment is possible, physically, but the result is that the SAS drive will not be usable, because the signalling is different and incompatible.--50.201.195.170 (talk) 07:18, 14 February 2021 (UTC)


 * I think the issue being raised is that you can't plug a combined power and data connector onto a drive that has them separated. You can plug separate cables into a drive with combined connector. As a quick fix, I have changed "identical" to "compatible". ~Kvng (talk) 15:37, 14 February 2021 (UTC)
 * I'm very sure. My company builds servers that use SAS disks and tape mechanisms, and SATA DVD-RW drives.  It is not possible to connect SATA power and data cables to any SAS device.  It is possible to connect an SAS cable, which has power and data combined in a single connector, to an SATA device. 216.152.18.132 (talk) 05:59, 3 March 2021 (UTC)

A Commons file used on this page or its Wikidata item has been nominated for deletion
The following Wikimedia Commons file used on this page or its Wikidata item has been nominated for deletion: Participate in the deletion discussion at the. —Community Tech bot (talk) 01:23, 7 August 2021 (UTC)
 * IBM 350 RAMAC.jpg