Talk:Disk sector

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Zoned bit recording[edit]

I think we need good linking to Zoned Bit Recording, otherwise any understanding about sectors is really confusing. I was trying to work out how the data transfer rate "depends on the track location, so it will be higher for data on the outer tracks (where there are more data sectors) and lower on the inner tracks". This article, and the diagram in particular, indicate there are the same number from inside to outside. It now appears that this hasn't been the case for 20 years.

See also the following, particularly the diagram: http://www.pcguide.com/ref/hdd/geom/tracks_ZBR.htm. Amniarix (talk) 20:00, 9 April 2010 (UTC)[reply]

Sector size[edit]

It should be noted that there have been sector sizes that were not a power of two. Disk drives produced for the IBM System/38 and it's follow-on products used 520-byte sectors.drh (talk) 22:20, 17 November 2012 (UTC)[reply]

It should be noted that a sector size of 256 bytes was quite common. The HP 3000 operating system (MPE), and others, had that number well enshrined in their code and documentation. — Preceding unsigned comment added by 50.76.52.65 (talk) 18:47, 20 February 2013 (UTC)[reply]

I agree that the claim

The standard sector size of 512 bytes for magnetic disks was established with the inception of the hard disk drive in 1956.

is dubious. See Talk:Advanced Format#512-byte ... since the inception of the hard-disk drive in 1956. DHR (talk) 17:43, 9 April 2013 (UTC)[reply]

If you look at the citation to that claim, it is for a disk with 100 character sectors. On top of all that 128 byte sectors were reasonably common on single density disks and later CP/M formats used sectors up to 2048 bytes (that I know of) all with block sizes of 128 bytes. — Preceding unsigned comment added by 173.163.243.81 (talk) 19:54, 15 September 2015 (UTC)[reply]

Standard Sector Size[edit]

The standard sector size of 512 bytes seems to date from the introduction of the IBM 3370 drive in 1979.

Were there any FBA devices prior to the 3370?

Please refer to the articles :-
Fixed Block Architecture
History of IBM magnetic disk drives — Preceding unsigned comment added by Phil from fitzroy (talkcontribs) 3 April 2014 (UTC)

512-byte sectors existed in the minicomputer world long before 1979. The RK05, introduced in 1975, had 512-byte sectors when used with PDP-11s; I think most if not all disks for the PDP-11, except perhaps for floppy disks, had 512-byte sectors. VAXes also had disks with 512-byte sectors. Hard disks used on the Data General Nova also had sectors of 256 16-bit words, i.e. 512 8-bit bytes.[1]
And IBM's fixed-block architecture didn't last forever, although later supposed count key data disks were implemented in firmware or software atop fixed-block disks.
So I'm not sure to what extent you can credit a 512-byte sector standard to IBM. Guy Harris (talk) 10:46, 12 October 2023 (UTC)[reply]
AFAIK the IBM 3310 and 3370 were the only FBA drives that IBM sold, although some later DASD subsystems could be configured with virtual FBA drives. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 11:56, 12 October 2023 (UTC)[reply]
Fixed block size, that is, fixed sector size was standard in most disk subsystems from the beginning, except for many IBM mainframes which beginning in the early 1960's introduced variable length sectors called records. IBM had fixed block size in its earliest disk drives; the first disk drive, the IBM RAMAC had a fixed block of 100 characters and IBM had fixed sector size in its small systems. I think DEC had some disk subsystems where partitions on a drive could have different sector sizes, with the sizes corresponded to blocks of different word lengths of different systems. I do think the IBM PC success coupled with embedded servos in the drives resulted in a sector size of 512 bytes becoming standard. Tom94022 (talk) 20:06, 12 October 2023 (UTC)[reply]
I'm going to need a citation on "DEC had some disk subsystems where partitions on a drive could have different sector sizes". The RK05 cartridges were formatted differently for 16-bit PDP-11s and 12-bit PDP-8s, and possibly differently from either for 18-bit PDP-15s.
If you don't have a standard quantum of data, you probably can't have a standard sector size. I suspect that as 8-bit bytes became more and more common, with 12-bit, 18-bit, 24-bit, 36-bit, and 48-bit machines becoming less common, power-of-2 sector sizes became more common:
  • DEC used 512-byte sectors in disks for their 8-bit-byte machines (PDP-11 and VAX).
  • Data General appears to have done the same.
  • HP appears to have used smaller sectors. At least one disk for the HP 3000 series had 128-16-bit-word, so 256-byte, sectors.Cite error: A <ref> tag is missing the closing </ref> (see the help page). and the 12960A for the 2100 minicomputers[2] also had 256-byte sectors. The HP 9153/9154's hard drives, used in the HP-UX 68k-based 9000/300, also had 256-byte sectors,[3] so it appears that not all drives intended for UN*X systems had 512-byte sectors.
  • When I was at Sun from 1985 to 1988, the drives we used were SMD for bigger drives and SCSI for smaller drives; I don't know offhand what sector sizes they had.
  • I don't think IBM came out with any non-8-bit-byte computers after the arrival of S/360. The 1130 had 321-word - 642-byte - sectors, so they hadn't standardized on power-of-2 sectors yet. It looks as if System/3 had 256-byte sectors.[4] I don't know what sector size was used by System/38 disks, but, as the page size is 512 bytes, it's probably a power of 2 <= 512 bytes. Series/1 also had 256-byte sectors.[5]
However, the 10MB hard drive option for the IBM Personal Computer had 512-byte sectors.[6] Was that because 1) the floppy disk also had 512-byte sectors,[7] 2) the makers of mass-market hard drives had already gone with 512-byte sectors, or 3) something else? Seagate's 2nd-generation Winchester drives had 256-byte sectors as shipped from Seagate,[8], so 2) may not be the case. If it's 1), floppy disks weren't all 512-byte-sectored back in 1976,[9] so why did that drive have 512-byte sectors? Guy Harris (talk) 05:43, 13 October 2023 (UTC)[reply]
Western Digital credits IDE for this:[10]

Early disk drives did not have standardized sector sizes. Each drive was simply a physical device and the drive control logic existed outside the hard disk drive, in the host. The number of bytes per sector was dependent on the HDD manufacturer, the operating system or application accessing the drive on the host, and what sort of error detection/correction mechanism the host used to ensure data integrity. In the 1980s, Western Digital invented the Integrated Drive Electronics (IDE) interface, which moved the drive controller into the physical drive. This created and standardized a command set for host- drive interaction, part of which defined a logical sector as 512 bytes long, with each byte consisting of 8 bits of data.

although this happened in 1986, after the PC and PC XT, at least according to Parallel ATA, so it's not the reason why IBM chose a drive with 512-byte sectors. According to Parallel ATA § IDE and ATA-1, WD, CDC, and Compaq "developed the connector, the signaling protocols and so on, with the goal of remaining software compatible with the existing ST-506 hard drive interface", but the ST-506 appears to have had 256-byte sectors,Cite error: A <ref> tag is missing the closing </ref> (see the help page). as did the ST-412.[11] The ST-412 wa, apparently the 10MB drive for the PC and PC XT; perhaps IBM said "OK, you know who we are and what product we want this for, and we want a version with 512-byte sectors" and Seagate said "OK, 512-byte sectors it is, heck, for the PC we'll give you 377-byte sectors if you want". Again, maybe the 512-byte sectors was to match the floppy drive. Guy Harris (talk) 06:41, 13 October 2023 (UTC)[reply]
I think WD has the story partially correct. Note I did correct Parallel ATA § IDE and ATA-1.
Until the advent of embedded servos, the track of a disk drive was a blank slate that each controller could segment into as many fixed length sectors (most of the industry) or variable length sectors (IBM CKD) as desired and could be fit into the finite length of the unformatted track. Disk drive manufacturers published unformatted track length and then capacities with various fixed or variable sector length. IBM's 3330 class for example had an unformatted track length of 13,440 bytes and a capacity of 100 MB but only with a maximum track length of 13,030 if I remember correctly. Same concept for the ST412 and all such drives.
IDE was indeed invented by WD, but it was the huge success of Conner's IDE drive that led to the ANSI committee adopting the 512-byte block as part of standardizing the interface. Conner's drive had to be compatible with the IBM PC's 512-byte block and it had embedded servos, so it was difficult for them to support other sector sizes other than about 512 or integer multiples thereof. Note there are some IDE drives with 520- or 524-byte blocks. I've researched a whole bunch of ANSI committee publications and minutes and find no explicit statement of 512 bytes (256 words) nor justification thereof. It seems like it was driven by the Conner initial spec and just implicit until maybe ATA8 where there is explicit mention in the definitions! But ATA became 80% of the market and the OSes and their drivers expected 512 bytes so that today even SCSI drives support 512 as "standard" and/or 4096 with very limited other sector sizes due to their embedded servos.
All of the above borders on OR so it's not likely to make the article, but it seems like market success of Conner's drive and then the success of ATA drove the standard sector size. Tom94022 (talk) 22:23, 15 October 2023 (UTC)[reply]

So, for whatever reason, IBM went with 512-byte sectors on the floppy drives for the PC and, when they added a hard drive, again went with 512-byte sectors, perhaps for compatibility with the floppy drives. Then WD came out with IDE, and Conner built the first(?) IDE drive, and it had 512-byte sectors for compatibility with the existing PC hard drive(s), and, as a result of the success of the Conner drive, ANSI standardized on 512-byte sectors in an IDE standard at some point? Guy Harris (talk) 01:08, 16 October 2023 (UTC)[reply]

The first "IDE" drive that I have confirmed is the hard drive in the Compaq Portable II announced on February 2, 1986; it was a WD Controller strapped to a 3½-inch MiniScribe HDD (I disassembled one and have photos). Compaq was reported to have worked with WD and CDC on a 5¼-inch IDE HDD for the Compaq Deskpro 286 but so far I've found no reliable source to confirm any shipment. In any event, these products were not sold separately and none of the four companies, WD, Compaq, MiniScribe or CDC made the interface public. Conner did an IDE drive for Compaq and shortly thereafter announced an OEM version with a published specification. The products took off, Conner at one point was the fastest growing company in the history of commerce thanks in great part to the success of its IDE product line. Competition entered and the ANSI committee then took up the burden of standardizing the interface between the several variants that existed amongst the compeditors. The 512-byte standard block size was apparently just adopted without any discussion.
I would rephrase your summary as follows:

For an unknown reason, IBM went with 512-byte sectors on the floppy drives for the PC and, when they added a hard disk drive, again went with 512-byte sectors, most likely to minimize impact on device driver and BIOS development. WD invented the IDE interface but they did not promulgate detail publicly; it featured full compatibility with the IBM PC/AT HDD BIOS and device driver, including but not limited to 512-byte sectors. Conner's highly successful commercialization of IDE drives beginning 1987 led to a multitude of competitive offerings, which in turn led an ANSI committee to standardize the interface beginning about 1989. It appears that the 512-byte sector size was adopted without discussion. IDE went on to become the dominant interface in the HDD industry thereby establishing the 512-byte sector as a Defacto standard.

Probably OR Tom94022 (talk) 06:22, 16 October 2023 (UTC)[reply]
As to why IBM Boca chose 512-bytes, it likely had to so with price and availability of the NEC 765 FDC and compatibility with the earlier IBM systems using FDs, such as 3740 and System/34. The chip only supported a few sector sizes so their choice was limited. Pure speculation on my part :-) Tom94022 (talk) 16:54, 16 October 2023 (UTC)[reply]

References

  1. ^ Technical Manual for the Moving Head Disk Controller 4046 and Disk Cartridge Adapters 4047 & 4049 (PDF). Data General. 1972. p. 4-10.
  2. ^ "12960A Cartridge Disk Subsystem" (PDF). Hewlett-Packard. June 1973.
  3. ^ HP 9153/54 A/B Service Manual (PDF). Hewlett-Packard. May 1988. p. 1-2.
  4. ^ IBM System/3 Disk Concepts and Planning Guide (PDF) (First ed.). September 1971. p. 4. GC21-7571-0.
  5. ^ IBM Series/1 4962 Disk Storage Unit and 4964 Diskette Unit Description (PDF) (Third ed.). June 1985. p. 2-1. GA34-0024-02.
  6. ^ IBM Personal Computer Technical Reference (PDF) (Revised ed.). April 1983. p. 1-203.
  7. ^ IBM Personal Computer Technical Reference (PDF) (First ed.). April 1981. p. 2-89.
  8. ^ "The ST400 Family" (PDF). 1982.
  9. ^ "Shugart Associates SA400 minifloppy Disk Drive" (PDF).
  10. ^ "Recording Technologies" (PDF). Western Digital. 2022.
  11. ^ ST-406/412/419 MicroWinchester OEM Manual (PDF). Seagate Technology. January 31, 1984. p. 3.

Proposed merge with Storage block[edit]

The definition offered in this unsourced article makes it sound as though storage block and disk sector are synonymous Ringbang (talk) 21:12, 5 August 2014 (UTC)[reply]

Dsimic and Ringbang, did it really make sense to redirect storage block here rather than Block_(data_storage)? These two articles are intertwined and perhaps should be actually be merged into one so that the subtle differences are more cohesively presented. There are differences between the two, but they always coexist. Disk Sector and Block Allocation For File seems like a pretty good discussion, and one of the answers notes that on Linux you can get Sector Size with "fdisk -l | grep -E "Sector size" and block size with "blockdev --getbsz /dev/sda". There are around a dozen related questions with various takes on the question. For example, What does chunk, block, offset, buffer and sector means? says that "typically sectors contain blocks", which seems wrong (but this book has the same statement. Also, this answer parrots this article's uncited statement "Early in the computing industry, the term "block" was loosely used to refer to a small chunk of data. Later the term referring to the data area was replaced by sector", which is misleading because block is a modern storage term. II | (t - c) 18:22, 13 December 2015 (UTC)[reply]

Hello! IMHO, merging the Block (data storage) article into Disk sector might make sense, but you'd need to open a standard discussion about the merger proposal. That way we should have input from more editors. — Dsimic (talk | contribs) 01:27, 16 December 2015 (UTC)[reply]

External links modified[edit]

Hello fellow Wikipedians,

I have just modified one external link on Disk sector. 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:

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

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 18 January 2022).

  • If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
  • If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 09:55, 11 September 2017 (UTC)[reply]

Disk Sector History[edit]

Tom, I did not get a reply from you after March 15th so I am putting this response at the top. This work represent the first mention of disk sectors with blocks and tracks as a method of greatly speeding up data transfer. The disc layout described in graphics in 1963 was completely novel at that time. The patent you did not find is below:

https://www.google.com/patents/US3439340 for Sequential access memory systems. The patent provides a new and improved method and arrangement for increasing the information transfer rate of sequential access memory systems. He had several other patents in this area as well. --Gallahed (talk) 21:20, 29 March 2018 (UTC)gallahed[reply]

I was hoping to add some relevant information to the history of disk sectors and their development. I recently discovered files regarding the original designs for disk sectors from 1963 and all the related internal correspondence from Bell Labs (Homdel). These files can be reviewed at https://archive.org/details/DiskDrivePatent. I am aware of the rules regarding original works but given the internal documents from Bell Labs, I thought they might suffice. Please advise on how to proceed. Gallahed (talk) 15:09, 14 March 2018 (UTC)gallahedGallahed (talk) 15:09, 14 March 2018 (UTC)[reply]

I made a brief scan thru the Gallaher disclosure and off the top of my head inclusion would give undue weight to the disclosure. First of all this is an internal disclosure and not an actual patent - a quick search did not turn up any such inventions to Gallaher; had there been a patent that would indicate a degree of inventiveness and from the references we might get an idea of how pionneering and therefore should it be included in this article. Its very clear in the document that Gallaher was well aware of disk and drum art circa 1963 all of which used fixed length sectors so this disclosure is not the invention of disk sectors. It seems to relate to what we now call zoned recording on a fixed head Burroughs disk drive but zoned recording was well established on Bryand disk drives well before the date of this disclosure and again not particularly relevent to this article.
In that era (actually for much of last century) the communications industry (ATT) and the computer industry (IBM) were on separate storage tracks and this is an interesting example of the same thing (zoned bit recording) being developed more or less simulaneously, so it might be a footnote in the non-existant history section of the Zone bit recording article. But one would first have to write the history section starting with Bryant :-(
If the dislosure was actually implemented it might make a footnote or paragraph in the Number_One_Electronic_Switching_System#1AESS #1AESS article which apparently is where ATT introduced disk (from Burroughs?). Problem is this might be OR
FWIW to the best of my recollection the history of disk sectors is probably something like:
  • 1957 - IBM 350 disk storage: 1st disk drive, fixed sector, no header
  • 1961? - IBM 1301, fixed variable length sectors with headers
  • 1962? - Bryant 4000, zoned recording
  • 1965 - IBM S/360 - variable length blocks with key block
  • 1992? - EMC4400? - emulation of IBM variable length records on fixed block RAID
  • 19?? - Some IBM drive - headerless sectors
It's not uncommon to find simultaneous or near simultaneous invention so its clear Gallaher was inventive in 1963 but IMO it doesn't rise to the level to be included in this article. Tom94022 (talk) 19:12, 14 March 2018 (UTC)[reply]

This work represent the first mention of disk sectors with blocks and tracks as a method of greatly speeding up data transfer. The disc layout described in graphics in 1963 was completely novel at that time. The patent you did not find is below:

https://www.google.com/patents/US3439340 for Sequential access memory systems.

The patent provides a new and improved method and arrangement for increasing the information transfer rate of sequential access memory systems.

He had several other patents in this area as well.~~gallahed~~ — Preceding unsigned comment added by Gallahed (talkcontribs) 16:01, 15 March 2018 (UTC)[reply]

Sorry I didn't respond earlier - please use wiki protocol by indenting and placing at the bottom.
Since disks are more or less random access, I'm not sure that a "Sequential access memory systems" patent is at all relevent here, particularly a 1965 patent that admits to being an improvement over known art which we know goes back to 1957. It is certainly not the first mention of disk sectors and the patent so states in its discussion of prior art. Including it in this article IMO would be giving undue weight to what is a best a minor improvement over the then known art and as near as I can tell not used in practice. Tom94022 (talk) 21:54, 30 March 2018 (UTC)[reply]
To add it to the article you really should at least add the first three known prior arts (RAMAC, 1301, Bryant) and then find a notable resourse that states why it is significant enough for inclusion. Adding the first three probably should be done; I doubt if you will find any notable support for incusion of the patent or disclosure. Most patents do not rise to the level of inclusion. Tom94022 (talk) 22:19, 30 March 2018 (UTC)[reply]
"This work represent the first mention of disk sectors with blocks and tracks as a method of greatly speeding up data transfer." It's certainly not the first mention of disks, disk sectors, or disk tracks; see, for example, page 12 of 305 RAMAC Random Access Method of Accounting and Control, Manual of Operation, from 1957, which is the first mention of disks, disk sectors, and disk tracks, as it's a manual for the IBM 305 RAMAC, which was the first machine with a disk drive. The disk, as the manual says, has "100 consecutive tracks on each disk", and "the disk tracks are subdivided into sectors". How are, for example, Figures 2 and 3 of Gallaher's memorandum any different from what the 305 RAMAC had?
So, no, Lee Gallaher didn't invent disk sectors unless he was working at IBM on the 305 RAMAC. What he may have invented is the idea of queueing up disk accesses and sorting the queue so that you can read multiple sectors from a track in the course of one revolution of the disk. Guy Harris (talk) 01:36, 4 April 2018 (UTC)[reply]
The patent does not describe sectors, but rather sorting of access requests by angular position, similar to what other vendors implemented on their fixed-head disks. The IBM 350 had a sector organization in 1956. Shmuel (Seymour J.) Metz Username:Chatul (talk) 17:10, 16 December 2019 (UTC)[reply]

Requested move 16 December 2019[edit]

The following is a closed discussion of a requested move. Please do not modify it. Subsequent comments should be made in a new section on the talk page. Editors desiring to contest the closing decision should consider a move review after discussing it on the closer's talk page. No further edits should be made to this discussion.

The result of the move request was: not moved to the proposed title at this time, per the discussion below. Dekimasuよ! 01:45, 24 December 2019 (UTC)[reply]


Disk sectorSector (computing)? – Disks are not the only storage devices to have sector organization. Shmuel (Seymour J.) Metz Username:Chatul (talk) 17:20, 16 December 2019 (UTC)[reply]

  • Oppose Seems like you are putting the cart before the horse here. Maybe make your "expanded" article on various other types of sectors and then talk about a possible merge into it.ZXCVBNM (TALK) 19:39, 16 December 2019 (UTC)[reply]
  • Oppose No need to replace a well understood universally accepted subject with an invented one. Tom94022 (talk) 23:51, 16 December 2019 (UTC)[reply]

The above discussion is preserved as an archive of a requested move. Please do not modify it. Subsequent comments should be made in a new section on this talk page or in a move review. No further edits should be made to this section.

Terminology[edit]

It may be desirable to identify a possible ambiguity in usage, in which the term is not always used as per the definition "The sector is the minimum storage unit of a hard drive." but instead "A sector is the minimum addressable storage unit of a hard drive."

Under the definition, modern 'advanced' hard drives have a sector size of 4096 byte (or more), but can address 512-byte sub-sector blocks. These smaller blocks very often inherit the term 'sector' from current (or possibly older) ATA standards. This may be made clearer if explained that ATA terminology is limited to host-device interface, and the above definition applied to HDD-platter (or equivalent) transfers.

It may be also desirable to explain why [1] is *the* authoritative source for the definition of the term, or if it isn't, why it has been used here. Athulin (talk) 09:27, 27 August 2020 (UTC)[reply]

I think the current definition is acceptable even though it is not supported by [1]. In a 512e HDD the minimum storage unit is 512 bytes, i.e. it is blocked into four blocks of 512 bytes with the assembly (during write) and disassembly (during read) handled by a translation layer with the minimum still 512. We probably should look for a better reference. Tom94022 (talk) 23:06, 27 August 2020 (UTC)[reply]

Inappropriate use of term sector[edit]

Data storage devices with variable block sizes fit neither the definition of sector in the lead nor the usage of the term in the literature. The term sector is totally inappropriate for the IBM 1301, IBM 1302 and the count key data[a] (CKD) DASD of the IBM System/360 and successors. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 09:03, 22 May 2023 (UTC)[reply]

Actually the definition of sector in the lede and its usage in the literature precisely fits and is appropriate for the IBM 1301 and all IBM (CKD) DASD which used variable length sectors. According to the IEEE 2000 dictionary: "Block ... (8) A group of data that is contiguous in nature. Synonym: sector." [emphasis added]. IBM uses neither term to describe what it called "records" when introduced with the 1301 so given they are synonyms either term can be used. Furthermore, block is more ambiguous as noted in the article section Disk_sector#Sectors_versus_blocks. It seems to me that introducing the phrase "variable length blocks" as @Chatul: seems to suggest adds nothing to the article and his disptue tag should be removed. Tom94022 (talk) 19:18, 2 September 2023 (UTC)[reply]
The use of the term block for the 1301 is not just what I suggest; it is what the literature for the IBM 1301, 1302, [E]CKD DASD and FBA DASD actually uses. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 16:22, 3 September 2023 (UTC)[reply]
IBM also sees an equivalence between sector and block; one many contemporary IBM citations:
"In fixed block (FB) architecture, the data (the logical volumes) are mapped over fixed-size blocks or sectors." [emphasis added].
It seems to be that if a "fixed-size block" means the same thing as a "sector" then a record can equally be described as a variable length block or a variable length sector, but since this is an article about sectors, introducing the term block adds nothing but the possibility of confusion. Also I've looked at the early IBM "record" literature and mostly I find "block" used mainly in the sense of blocks of records, more like today's "cluster" so AFAICT there is no support for the proposed use of block in the early literature. Tom94022 (talk) 18:10, 3 September 2023 (UTC)[reply]
The usage in the literature varies between hardware and software, but in neither case is a variable length record referred to as a sector. The software documentation uses the terms block and physical record while the hardware documentation uses the term record. The usage of a block containing multiple records is in the software documentation. IBM, in particular, uses the term sector in the context or Rotational Position Sensing, and sector boundaries are not, in general, aligned with physical record boundaries. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 12:23, 5 September 2023 (UTC)[reply]
We are not quoting historical literature, just using English in ordinary meanings to explain what is a "disk sector" and how its meaning evolved from the introduction of HDDs. As such the lack of usage of the phase "sector" in IBM DASD many descriptions of "record" is pretty much irrelevant. The only question is it reasonably accurate to describe a "record" as a variable length sector in an article about disc sectors? You might take a look at how the issue of block, record and sector is dealt with in Fixed-block architecture and Count key data they deal with the issue. Tom94022 (talk) 22:42, 15 September 2023 (UTC)[reply]
BTW, @Chatul: u never really identified why the current description is "totally inappropriate" nor provided any appropriate language for replacement - both would be helpful. Tom94022 (talk) 22:45, 15 September 2023 (UTC)[reply]
It's inappropriate because the literature does not use the word sector for variable length blocks. The appropriate language for replacement is the original The 1961 IBM 1301 disk storage introduced variable length blocks, termed records or physical records by IBM,, The disk drives and other DASDs announced with the IBM System/360 in 1964 used self-formatting variable length blocks, termed records or physical records by IBM. and Ultimately all fields of disk blockss had ECCs., possibly using physical record in the last two.
At the same time there probably should be a sentence explaining that the rotational positional sensing introduced by the IBM 2305 used geometrical sectors that were not synchronized with block boundaries.
In general software documentation used the word block while hardware documentation used the word record. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 18:57, 19 September 2023 (UTC)[reply]
Actually I think the original language was The 1961 IBM 1301 disk storage introduced variable length sectors which you changed in 2018. IBM said in 1961 that:
  • ...self-formatting variable length blocks, termed records or physical records by IBM. [Emphasis added]
and today says:
  • the data (the logical volumes) are mapped over fixed-size blocks or sectors.[Emphasis added]
Why is it inappropriate using today's language to not describe a record as a variable length sector? If in today's language block == fixed sized sector and a record == variable sized block the why in plain contemporary English isn't variable sized sector == record? Just because you don't like this description is not a sufficient reason to interject the term block into this article. There is nothing inappropriate in such constructions. Tom94022 (talk) 23:36, 19 September 2023 (UTC)[reply]
That's begging the question; in today's language there are no variable length sectors.
In the text that you quoted, IBM wasNOT equating blocks to sectors; IBM was equating fixed-size blocks to sectors. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 15:25, 20 September 2023 (UTC)[reply]

To repeat myself:

We are not quoting historical literature, just using English in ordinary meanings to explain what is a "disk sector" and how its meaning evolved from the introduction of HDDs. As such the lack of usage of the phase "sector" in IBM DASD many descriptions of "record" is pretty much irrelevant. The only question is it reasonably accurate to describe a "record" as a variable length sector in an article about disc sectors?

Specifically, "variable" and "length" are being used in their ordinary English language meanings to modify the term of art "sector." It is used this way in other media (About 2,550 results on Google) and at least one place within Wikipedia. The fact IBM publications do not specifically use the phrase "variable length sector" to describe what it variously calls a record, physical record or block is irrelevant to this article and the note adds nothing of value to the article and should be removed. Since no one else has commented on this dialog I think it is time to go to the Village Pump. Tom94022 (talk) 20:15, 9 March 2024 (UTC)[reply]

Notes

  1. ^ The IBM 360/20 and 360/44 had sector-oriented DASD. The IBM 3310 and 3370 on the IBM System/370 had Fixed Block Architecture (FBA)