Talk:Floppy disk

Use unambiguous prefixes again

In 2017 I wrote:

The capacities stated in this article are difficult to understand since they are a mixture of IEC decimal and binary prefixes. I would like to edit the article so that capacities are always stated in unambiguous IEC binary (e.g., KiB) or decimal (e.g. kB) prefixes. ...
— Talk:Floppy_disk#Use_unambiguous_prefixes?

There wasn't much support then and the consensus seemed to be to use the prefixes of the source and live with ambiguity. A comment to that effect is embedded in the section of the article. @ShadyCrack: in a recent edit, now reverted, converted all of the values with IEC decimal prefix kB to values with conventional binary prefixes, KB. Going all conventional binary prefix is one way to resolve the ambiguity but if we are going to convert units such conversion probably should be flagged in some form. Personally I would prefer IEC Binary and Decimal prefixes but going all conventional, so annotated, works too. Comment? Tom94022 (talk) 18:41, 21 February 2022 (UTC)

I found the 23FD, 33FD, and 43FD presentation of size rather confusing in the context of this table, so I was attempting to convert to KB as is used in the rest of the table and in the List of floppy disk formats where I confirmed the values. I think these three should be presented as KB to be consistent with the rest of the table.

I did not appreciate the other two "kB" references potentially literally meant 1000 bytes, my bad. ShadyCrack (talk) 19:37, 21 February 2022 (UTC)

Wikipedia is not a reliable source. For example, an IBM 1981 JRD article, "The IBM Diskette and Diskette Drive" gives the 23FD, 33FD, and 43FD capacities as 81,644 bytes, 242,944 and 568,320 bytes respectively. In these cases kB is appropriate and there is no basis for conversion to KB. Tom94022 (talk) 01:08, 22 February 2022 (UTC)

I think the table should be updated to use the current binary standards of kiB and MiB, or at the very least include a footnote on the standard used. For the formatted capacity anyway, the marketed capacity should use whatever it was originally marketed as. StuartH (talk) 11:24, 4 March 2023 (UTC)

Generally I prefer KB over KiB per WP:COMPUNITS, but given some of the mixed units in the article, I could see the argument for conversion to KiB and MiB to improve clarity. ShadyCrack (talk) 15:12, 4 March 2023 (UTC)

Before we start on such a project there needs to be consensus here on doing it. Right now it is 2 in favor of unambiguous prefixes and one neutral. Given the number of conventional binary prefixes zealots on Wikipedia we probably need 5 to 10 in favor of such a rewrite before we dare attempt. I suppose we could ping the editors who have contributed the most to this article to see if there is such a consensus. Tom94022 (talk) 18:01, 4 March 2023 (UTC)

Upon closer review, the article appears consistent with KB = 1024 and kB = 1000. There are some confusing MB references that are worth clarifying. E.g. 1.68 MB and 1.2 MB are not MiB or decimal MB, they are same as "1.44 MB" which is explained or the unambiguous 1,440 KB is used. 1.68 MB could be 1,680 KB, and 1.2 MB could be 1,200 KB. I support leaving the article as KB for readability. But potentially note at the start of the article all KB are 1024 and kB are 1000 bytes. And we clarify anything else as needed. ShadyCrack (talk) 15:00, 5 March 2023 (UTC)

I take it that means you oppose IEC Binary Prefixes. A detailed note in the "Sizes, performance and capacity" section along with some sort of warning at the start of the article, perhaps in the lede, would work so long as we are careful that k==1000 and K==1024 is consistent throughout. I already put in explicit descriptions for the first usage of KB and MB. Tom94022 (talk) 17:29, 6 March 2023 (UTC)

@Tom94022 your recent changes look good. Going forward I feel we should just address any specific ambiguity that is brought up rather than any wholesale unit changes to the entire article. ShadyCrack (talk) 23:54, 7 March 2023 (UTC)

Spelling

Is there any significance to "Floppy Disk" spelled with a "k" vs. "Compact Disc" spelled with a "c"? 24.51.192.49 (talk) 04:24, 19 January 2023 (UTC)

"Compact Disc" is trademarked, "Floppy Disk" is not, the HDD folks adopted disk to avoid potential trademark issues with IBM and that carried over into FD. Tom94022 (talk) 07:30, 19 January 2023 (UTC)

If one reads HP manuals and marketing materials from the 1980s, the word was always spelled with a 'c'. Similar material from IBM spells with a 'k'. The 'k' won out at some point. Since the 'c' spelling was in use long before CDs were invented, I think it's doubtful that a concern over trademark issues was the cause. Christopher Rath (talk) 19:57, 1 January 2024 (UTC)

The decision was in the 1960s so HP materials from the 1980s are pretty much irrelevant. Tom94022 (talk) 21:03, 1 January 2024 (UTC)

Incorrect archive.org link for a lost citation

The SCS_2007 reference looks like this:

<ref name="SCS_2007">
{{cite web 
|author=(M)Tronics SCS 
|date=2007-05-20 
|title=Floppy-Disketten-Laufwerke 
|trans-title=Floppy disk drives 
|url=http://www.hardware-bastelkiste.de/floppy.html 
|url-status=dead 
|archive-url=https://web.archive.org/web/20170619194609/http://www.hardware-bastelkiste.de/index.html?floppy.html 
|archive-date=2017-06-19 
|access-date=2017-06-19 
|language=de
}}</ref>

The links themselves:

1. (dead) http://www.hardware-bastelkiste.de/floppy.html
2. (archive.org) https://web.archive.org/web/20170619194609/http://www.hardware-bastelkiste.de/index.html?floppy.html

That second link is incorrect, the page should be floppy.html not index.html with a parameter after it:

1. (archive.org fixed) https://web.archive.org/web/20170619194609/http://www.hardware-bastelkiste.de/floppy.html

Alas archive.org then rewrites the URL to a different date, so I'm not sure if I should edit it as a drop in replacement; or update all of the dates (including today as an access-date because I've just read and confirmed it).

I'm a new user. Whats the best practice for changing/fixing this? BrushDamp (talk) 09:25, 22 February 2023 (UTC)

5.25 Floppies

The older "360K" floppies (and I'm speaking about IBM-style formatted capacities) had a reinforcing center ring where the drive would clamp the disk. This ring was eliminated on the 1.2M floppies. What was the original idea behind this and why was it deemed unnecessary on the very similar-looking 1.2M disks? 57.135.233.22 (talk) 05:23, 8 July 2023 (UTC)

I'm pretty sure the ring was not standard on any 5¼-inch FD - AFAIC recall it was added by some vendors to prevent or reduce damage caused by the various FDD clamping mechanisms, which would then make the disk unreliable. Tom94022 (talk) 21:00, 8 July 2023 (UTC)

I'm looking at all the pictures of 5.25" disks in the article because I didn't believe you and I don't see any with that ring, so I guess you're right. "Mandela effect" or something. :) .... 57.135.233.22 (talk) 08:52, 10 July 2023 (UTC)

Fraction characters

@Tayste: Re: this revert, what's the reason for preferring vulgar fraction characters in these section titles? "5.25" and "3.25" are already being used in subsection headers of Floppy disk#Sizes. Per MOS:FRAC, articles that use vulgar fractions not on the approved list must use {{frac}}, but this cannot be used in link targets, so using decimals in section headers makes it less awkward to link here from those articles and for automated scans to show they have been fixed. -- Beland (talk) 01:10, 18 July 2023 (UTC)

Well, MOS:FRAC admits an acceptable exception "If ¼, ½, and ¾[k] are the only fractions needed" where [k] says these characters "are in ISO/IEC 8859-1 and work in screen readers". I think we want screen readers (and everyone else) to read these as "five and a quarter" and "three and a half" (rather than "five point two five" and "three point five") because that's what these floppy disks were generally called in their day. Tayste (edits) 03:28, 18 July 2023 (UTC)

Very well, I have changed all the instances of {{frac}} and decimal fractions to use the Unicode characters. The section headers actually have supplemental anchors with decimal fractions, so links from other articles should work fine. -- Beland (talk) 18:57, 25 July 2023 (UTC)

Thanks for making this change. The decision for consistency had been sitting out there for awhile. ShadyCrack (talk) 14:43, 10 August 2023 (UTC)

Single-sided 5¼" floppies and IBM-compatible PCs

A user has recently questioned and summarily reverted part of an edit which added the information that flippy disks (and single-sided floppy drives) were never common on IBM-compatible PCs. It's harder to google arcana that predate Google, but for clarity, this relates mainly to the 160KB (and 180KB) 5¼" floppy disk formats, i.e. those formats that show a 1 in the fourth column here. The actual common 5¼" formats on IBM-compatible PCs were double-sided double density (DSDD, i.e. 320 or 360KB) and later double-sided high density (DSHD, 1,200KB). (The capacity expansion from 160 to 180K and 320 to 360K was a software upgrade with DOS 2.0 in 1983.) Everything else was relatively uncommon, but single-sided drives were especially rare on the PC because only the earliest 1981 announcement referenced 160KB (i.e. single-sided double density, SSDD) drives. (Single density floppies were never a thing on IBM PCs.) When IBM PCs started shipping in volume, IBM had moved on to double-sided drives, which soon became pretty much the only option on IBM-compatibles. However, right around the same time, single-sided 5¼" floppy drives and disks were VERY common on HOME COMPUTERS, most notably the C64. Regarding the actual IBM PC, some of the best-respected sources don't even mention single-sided 160/180K floppy drives anymore, and for good reason: For practical purposes virtually every IBM PC with a floppy drive was double-sided capable.

The repeated reversions are especially uncalled-for because the relevant edit summary demonstrates the reverting user doesn't even understand what is asserted, which even when assuming good faith is still at least a stark lack of reading comprehension. The assertion they contest was specifically about IBM-compatible PCs, so anything beyond PCs cannot possibly be a counterargument to it. —ReadOnlyAccount (talk) 20:05, 29 April 2024 (UTC)