Talk:Floppy disk/Archive 5

Archive 1

←

Archive 3

Archive 4

Archive 5

Archive 6

How does it work?

It occurs to me that our proverbial bright 12-year old, looking at this article, might want to know *how a floppy disk works" - how do the 1's and 0's get saved on a sheet of plastic? Why does it spin? Why does it need a hole? Why do they tell you to keep it away from magnets? We need some fundamentals, which should have a diagram of a cross-section through a drive head and at least some pointers to FM, MFM, and GCR. Maybe a diagram of how a disk track is formatted for the original 8 inch size. Explain what is physically happening in a track and sector. Why are there inter sector gaps? What does spindle speed have to do with anything? Why are both sides coated if only one side is used? Etc. --Wtshymanski (talk) 14:45, 9 August 2010 (UTC)

A lot of what u suggest is already in HDD article's Technology Section which is not surprising since they are basically the same other than the flexible medium and lack of flying heads (contact recording). A Tape drive's recording technology is also quite similar, again with differences primarily relating to the medium and head. Like this FD article, the tape drive article also lacks a how to section. All of which suggests a separate article linked from the three articles. The basis of the article would be any of the several books and articles on magnetic recording, e.g. Magnetic Storage Handbook, Mee and Daniel, 1996. In the mean time perhaps just adding a link in the Floppy_disk#Structure section (or renaming it Technology) of this article pointing to the HDD article's Technology section will suffice. Tom94022 (talk) 16:19, 9 August 2010 (UTC)

Yeah, or we could just tell the user to look it up on Google. Much more work is needed here and I'm temporarily without Net access in my leisure hours....--Wtshymanski (talk) 13:17, 11 August 2010 (UTC)

We need to explain how 1s and 0s get turned into magnetized spots on the surface of the disk, then how the spots get turned back into 1s and 0s (chanting FM, MFM and GCR here). Diagram, cross-section of a read/write head with coils being dragged over some magnetic medium. What limits how close the spots can be spaced? We need to explain how the featureless blank disk surface is organized into tracks and sectors so particular 1s and 0s can be found (Diagram, formatted floppy disk with the tracks and sectors and gaps shaded in as if you could see them). We should mention how the head gets slewed around the surface of the disk (talk about tracks per (radial) inch or metric equivalent). We need to mention that the operating system looks after the messy business of deciding which tracks, sectors, and sides hold which parts of a file (Diagram, linear file mapped to sectors all over the disk). We should explain why there's a hole (or many holes) in a diskette - hard sectoring vs. soft sectoring. If we did all that, the rest of the article would make more sense. Only then should we get into "Binford Model 6100 vs Binford model 6110" Wikipedia-style trivia lists; let's put all the specific format information into "List of diskette formats". --Wtshymanski (talk) 16:22, 11 August 2010 (UTC)

No Index Performance

I have removed from the article what IMO is a speculative paragraph about the performance implications of FDs formatted without index sensing as in Apple DOS FDs. As I understand it:

1. In most controllers the FD's index has nothing to do with finding the target sector following a seek - after the seek the controller finds the first readable address mark regardless of the index - therefore no performance implications in single sector reads or in finding the first sector of a large block.
2. In large block reads, depending upon the format, there is the possibility of loosing one revolution at each track boundary; potentially reducing performance by as much as a factor of two. With sector skewing the performance could be improved but most format programs did not routinely skew. Formatting without an index more or less gives you a sector skew, but it is highly likely that it is not optimized. So performance may be better or worse than other FDs.

So it appears to me that the relative performance implications of "index-less" formatted FDs vis a vis "conventionally" formatted FDs is indeterminate. On the whole I think this is TMI for this article. Tom94022 (talk) 16:40, 30 September 2010 (UTC)

R/W performance

Read and write data rate are missed (but the 2" version). Interest may be to face common 3,5" FDs to actual USB 2 storage keys rates in its slower (or market average) mode.
Do emulators offer better rate? —Preceding unsigned comment added by 213.56.248.115 (talk) 08:16, 9 November 2010 (UTC)

3 1/2 inch floppy size

FWIW, I just hauled out my micrometer and measured the width of a few 3.5 inch floppies. All of them were right on 3.54 inches. And they were all (OK, sample size was only three) less than .04mm off from 90mm. Rwessel (talk) 19:48, 17 December 2010 (UTC)

At the risk of having this conversation by myself, while the ANSI spec *names* the floppy "90mm (3.5-in)", I'm pretty sure it defines the width unequivocally as 90mm (or ~3.54in). I can't find an online copy of the standard that can be viewed for free. FWIW, the 8" and 5.25 inch floppy standards are named "8-inch (200-mm)" and "5.25-inch (130mm)", and those *are* actually 8 or 5.25 inches in width. Rwessel (talk) 06:01, 18 December 2010 (UTC)

I would not be surprised if the specification did not have imperial measurements at all - just metric. I too am looking for a copy of the spec, so far without luck Tom94022 (talk) 08:00, 18 December 2010 (UTC)

I still haven't found the spec, but even if the internal specification is 90 mm (3.54 in) my understanding of dimensioning standards is that the nominal dimension is 90 mm and the parenthetical dimension is for reference. But this really doesn't matter since the point is that the specifying body, ANSI uses 3.5 in in the document title, not 3.54 in , conforming to the world's designation of the medium. Two editors seem to agree that this is the point to be made while Pantergraph doesn't. At this point the article should remain as is while this is discussed. Tom94022 (talk) 04:44, 20 December 2010 (UTC)

I'm under the impression that the idea was to accurately describe the size of the 3.5" floppy. As in "it's called a 3.5" floppy, but it's really 90mm (3.54") wide." That's at least somewhat noteworthy, since unlike 8" and 5.25" floppies, it's not actually the nominal size. Rwessel (talk) 05:34, 20 December 2010 (UTC)

As I recall history both the 8" and 5¼" FDs were originally dimensioned in imperial units and so named but the 3½" was originally dimensioned in metric units. So I think the more important point is that it is a metric dimensioned medium know by an approximate imperial dimension.

• I do have a copy of ECMA 135 (December 10, 1987) which is the standard for the 80 track/side FD. It does not have any imperial dimensions! It does not mention 3.5-inch. The cartridge is a rectangle 94,0 mm ± 0,3 mm by 90,0 mm + 0,4 mm -0,1 mm while the disk diameter is 85,8 mm ± 0,2 mm. So it turns out 3.5 inches is only the approximate dimension of one side of the cartridge. I suspect that the original dimensions in the ANSI spec were the same. Tom94022 (talk) 06:56, 20 December 2010 (UTC)

• ANSI INCITS 171 (Jan 1, 1989) is the current standard and is entitled, "Information Systems - One- and Two-Sided, High-Density, Unformatted, 90-mm (3.5 in), 5,3-tpmm (135-tpi), Flexible Disk Cartridge for 15 916 bpr Use - General, Physical, and Magnetic Requirements" Tom94022 (talk) 07:05, 20 December 2010 (UTC)

floppy?

Isn't a floppy disk floppy? —Preceding unsigned comment added by 69.230.85.122 (talk) 02:31, 23 December 2010 (UTC)

Yes, it a has flexible medium Tom94022 (talk) 01:13, 10 March 2011 (UTC)

Note that while the disk itself is always floppy, there is a difference between the actual products. A 5.25" floppy disk has a floppy plastic container, so it's very easy to bend or snap it (although this will destroy the disk), but a 3.5" floppy disk has a rigid plastic container, so in practice, in your hand, it isn't actually floppy. JIP | Talk 18:55, 22 January 2012 (UTC)

"popular and ubiquitous"

There is at least one article on the web that implies that in 2003 all HP and most Dell desktops had standard FDDs. The 1973 introduction of the IBM 3740 began the FD usage as a form of data storage and interchange. That sounds like at least 3 decades. Tom94022 (talk) 01:13, 10 March 2011 (UTC)

It's a bit of a stretch at either end. While floppies did ship on desktops fairly commonly into the early 2000s, the usage had fallen dramatically at that point. Nor can you limit that to desktops - laptops tended to make those optional a few years earlier (with many shipping without the usual external FDD). And small form factor desktops were losing floppies earlier than mainstream desktops. So by that point there were a significant number of systems without FDDs (so less than ubiquitous), and the usage was far thinner than that - I'd suspect that most FDDs shipped as built-ins in desktop systems after 2000 or 2001 never actually read a floppy while in user hands (so not so popular for data storage or exchange - which is the context). And even then, it would have been very limited usage (something to install a patch, for example). While there were certainly pockets of usage - someone has already mentioned systems needing disk drivers, but that mainly included personal systems from white box vendors, self builds, and servers, and *not* mainstream desktops and laptops.

On the flip side, while 8" floppies shipped from IBM from 1971, the were not initially used for data storage as such, which mostly started in about 1973 with the 3540 (diskette drive for the mainframe) and 3740 data entry system (basically intended to replace a keypunch machine and punch cards - the 3740 wrote to a diskette instead). That saw very limited use for data interchange (unless you count the stack of diskettes from the keypunch department at the end of the shift), and almost all mainframe-to-mainframe data exchange used tapes (as did most minis). The mid seventies saw some additional uses amongst minis (including IBM’s), to be sure, but there were many incompatible formats. The late seventies is when floppy usage started really picking up for micros (Apple introduced the Disk II for the Apple II in 1978, for example, although there were certainly diskettes used on micros before that). In 1976 if you wanted to send files from one data center to another, you did it on 1600bpi 9 track tape (with the more advanced sites able to deal with 6250bpi tapes). Floppies were *far* behind that in usage. So on that end, I don't think we can call floppies popular or ubiquitous, until the late seventies.

So most of the 80s and 90s. Plus (maybe) a few years on either side. Does that add up to a full three decades? It's a bit of a stretch, IMO. Rwessel (talk) 02:12, 10 March 2011 (UTC)

When i said we used them at college in 2002 (i left in that year), I meant it was really the only option. No one brought their assignments to class on a USB Drive. XP was still relatively new and unless the PC ran Win98SE or XP you couldn't use the USB drive until you installed the drivers. I ran an IT business until mid 2004 (SOHO) and nearly all customers were still actively using floppies. Why not CD's? To the masses it was a "pain in the ass to burn" (office documents - not photos or video). Regarding RAID and XP, I couldn't be rid of the thing until i went to Vista. Its ironic that early adopters of technology or enthusiasts were still chained to the thing until that time.

I would say certainly ubiquitous until mid 2005. When it happened, the change was practically overnight. (Ben, 11 March 2011) —Preceding unsigned comment added by 110.175.30.32 (talk) 15:05, 10 March 2011 (UTC)

I certain don't discount your experience, but mine is different. By the early 2000s it seemed that almost all class assignments (and mind you I've been out of school for a lot longer than that, but nieces, nephews, etc.) and business data transfer were happening by network (with CDs playing a lesser role, except for software distribution). And the disk driver issue, as I mentioned, affected only a tiny portion of workstation users (desktops and laptops, as opposed to servers).

Part of this revolves around the definitions of popular and ubiquitous, and how we count. For example, having a document shared on the network certainly displaces transferring it around the office via floppy when multiple people access it, but how do you count the shared accesses? You could easily have hundreds of instances of sharing and "transferring" in a day, which you would never do if you had to keep moving things around on a floppy (IOW, the much easier and faster network transfers actually change the workflow). And are they popular if 10% (or 1% or 90%) of transfers are done with floppies? And are floppies ubiquitous if they’re on 90% (or 75% or 99%) of PCs?

But even if we go to 2005, can we justify 1975 on the other end to get the full three decades? Rwessel (talk) 17:15, 10 March 2011 (UTC)

If the start of the ramp up was 1973 (3740 and founding of Shugart) and the beginning of the decline occurs in or after 2003 we can easily find exactly three decades. The 5¼-inch FDD was introduced in 1976 by which time I am pretty sure most data entry and minicomputers had some form of 8-inch FDD (e.g. Oct 1976 Datamation article on Small Business Computers has FDDs on 72 of the 151 systems) so 1976 to 2005 is 29 years which is close enough to three decades for me. Tom94022 (talk) 21:58, 10 March 2011 (UTC)

'70s, '80's, 90's, 00's - four decades, if you like. Though I haven't seen 5 1/4's at Staples for a couple of years now.--Wtshymanski (talk) 22:09, 10 March 2011 (UTC)

Obsolete (revisited)

The subject of obsolescence appears occasionally throughout these discussions, and I would just like to mention an article on the BBC website entitled 40 ways we still use floppy disks. This, along with some online articles stating that millions of disks are still being manufactured, suggests that the floppy is far from obsolete. Maybe future edits which touch upon this subject could be done with this in mind. Cheers. SP1R1TM4N (talk) 23:04, 16 March 2011 (UTC)

New section "Drive bay height"

Much of the information in this section is already covered in Drive bay, and some of the rest is incorrect (there were half height 8" floppy drives, although they were rare - and given my hazy memory of the details, it's possible they were two inseparable drives in a single full sized chassis), and some is incomplete in some ways (there is another smaller drive bay size for 3.5", ~.7", and drives small enough to fit were manufactured, although I think they were mostly used in laptops and in the 5.25+3.5 combo drives on desktops), and does not talk about bay widths at all - which are *not* the same as the nominal diskette sizes, nor the various vertical dimensions for 5.25 bays). Anyway, mostly minor issues needing a bit of cleanup or expansion.

But: how much of this discussion is appropriate here? Is an extensive discussion of sizes, and full, vs. half, vs. third height bays really useful? My inclination is to reduce this to a single sentence referring sizes to Drive bay. Any maybe adding a bit more detail to that article. If this section stays, it should probably be renamed “drive bay sizes,” or some such, and expanded as above. Rwessel (talk) 07:01, 26 March 2011 (UTC)

There were ½ height 8-inch FDDs by at least Tandon and Shugart Associates and two were intended to fit into one FH 8-inch bay. At Shugart we even went so far as to also offer a ½ height 8-inch HDD so that u could have one of each in a FH 8-inch bay. IMO, None of the discussion is appropriate; it should be deleted and the few appropriate facts therein incorporated into its parent section. Tom94022 (talk) 17:59, 26 March 2011 (UTC)

Arbitrary sector sizes

I think it's useful to designate the powers-of-two convention for sector size as "arbitrary" to identify to the reader that there's no particular physical or religious reason for this choice. Unlike DRAM packages, where every time you added a pin to the package you got 4 times the storage space , a disk sector has no physical limitation aside from how many bits you can cram on a track. The 128, 256, 512-byte convention wasn't handed down to us by the saucer people, it was a choice made (I suspect) by IBM for the floppy that has stuck due to floppy controller chips not being built for randomly defined sector lengths. (After all, the great-grandaddy of all disk drives had 100-character (not byte!) sectors [1]). And now we have non-power-of-two sector lengths in hard drives; this is a breakthrough discovery on the order of announcing we no longer have to go about with lead weights in our shoes. --Wtshymanski (talk) 18:31, 31 March 2011 (UTC)

I agree it is arbitrary. The early FDs were spec'ed as gross bits per track and the number and size of the sector was left to the arbitrary decision of the controller designer. A full track block almost by definition was not a binary number and those folks wishing to maximize capacity I seem to recall did in fact build such controllers. Controllers of the 8-inch ere for the most part constructed from MSI with a formatter/deformatter counting for the physical block size which was generally not a binary number so that a non-binary data block length at most represented an additional decode (end of data block) of the block counter. All of which means, arbitrary, albeit convenient in a computer with binary memory addressing. BTW, don't modern enterprise HDDs have 520 data byte sectors? And of course IBM CKD drives permitted record (sector) lengths from 1 byte to full track byte count and many applications were other than binary numbers, like 880 for example. Tom94022 (talk) 21:47, 31 March 2011 (UTC)

IIRC, the first IBM floppies were something like 319 bytes per sector.

The power-of-two is arbitrary in the sense that it could be any other size, but it has developed into a self-reinforcing standard at this point, and it is a convenient size for the OS to deal with. Many filesystems, including most of the popular ones, have the power-of-two sizes fairly well ingrained (and same even 512 bytes), and would require significant work to support other sizes. In fact the 512 byte standard has long been a significant burden to the hard drive manufacturers, and witness how slow the transition to 4KB sectors has been. The HD guys certainly want bigger sectors than that, but 4KB is at least reasonable for OSs and whatnot, so pushing any further is implausible at this point (and 4KB is a big improvement over 512 bytes).

A few* applications have supported other sizes, and many manufacturers of high end drives used to offer (at least for their high end drives), a considerable choice of sector sizes (for example, 512-4096 bytes, in increments of four). But that's largely gone away, except for a very small range (512-528 bytes or smaller, often just a couple of fixed sizes - similar for 4KB sectors). But most of these 512-plus-a-few-bytes drives are not seen by the OS or applications as that size, rather, the extra storage is for the RAID controllers - the system see the traditional 512 byte sector. IOW, the extra bytes are there *because* 512 bytes is so solid a standard (FSVO “standard”) at this point.

So yes – powers of two are arbitrary from a physical perspective, but impractical for external reasons. So I think that “Individual formatted sector lengths are almost always set to powers of 2” covers both the reality, and the (extremely) rare possibility of other sector sizes.

*ignoring mainframe applications that are aware of CKD - and note that no real CKD drives exist anymore anyway, they're all emulated on top of conventional fixed sector HDs. Rwessel (talk) 22:38, 6 April 2011 (UTC)

I agree with everything you say but disagree with your conclusion. I would point out that with headerless and split field recording technology there is no technical reason why any size sector could not be written. We have come full circle from when there were no physical sector marks so tracks were sub-divided into any number of arbitrary length sectors (as in CKD) to where it is again technically feasible but impractical because today's OSes are incapable of dealing with or exploiting such variability - that's why 4Ki sectors are having such a hard time.

So I think that the arbitrary nature of the powers of two block side needs to be conveyed in this article. Tom94022 (talk) 19:36, 7 April 2011 (UTC)

Well...yes, this is what I wanted to convey several days ago. There's no physical reason for power-of-two sector lengths, it's just convenient and customary. We serve our readers best when we reveal the man behind the curtain. --Wtshymanski (talk) 20:01, 7 April 2011 (UTC)

Since this article is meant to be comparing the byte sizes with what is advertised on the media I have removed the less than familiar IEC prefixes because they are not used by the media sources (WP:NOR) and replaced them with WP:MOSNUM approved disambiguation for the number of bytes. Glider87 (talk) 10:34, 8 April 2011 (UTC)

72.49.251.241 (talk) 15:46, 1 October 2011 (UTC) dont know where this fits but there was hard sectoring on floppys. the 8" had it on a lot of systems. the timing hole in the center has a lot of holes.. gives more precise sector position than assuming speed as a ratio between 360 degree rotation. usually more storage than soft sectoring.. micropolis was the good one for 5" 72.49.251.241 (talk) 15:46, 1 October 2011 (UTC)

Given the size of the holes on the ID it's not so clear that hard sectors gave more capacity than soft sectors. Be that as it may, hard sectors died early and are covered briefly in the article, so IMO its TMI. But if the IP wants to append a brief statement or a footnote to the already existing hard sector material, great. Tom94022 (talk) 16:33, 1 October 2011 (UTC)

"Floopy" drives

I remember this being kicked around in the trade rags, but unfortunately can't find any online cites. Not sure I have the energy to go through my archive of print BYTE magazines from that era. Is that helpful? Probably not, but... Rwessel (talk) 19:56, 30 May 2011 (UTC)

I for one recall a number of tape loop devices being proposed from time to time but none ever called "Floopy" so I suggest its subsection be deleted. Actually I think the whole Floppy_disk#Tape_based_floppy_alternatives section should be nuked or at least moved to a separate article. Tom94022 (talk) 20:27, 30 May 2011 (UTC)

I'd say leave the "stringy floppy" section here because at one time, hard to believe, a floppy disk drive was a rare and expensive piece of hardware. One of the later Sinclair computers had some kind of wafer tape, too, as I recall. I couldn't find any real references to "floopy" drives except for those lovable proofreaders at Tata-McGraw Hill (where the slogan might be "we skip the editing process and pass the savings on to you!") and their allies at other publications - even Google thinks "Floopy" is a type for "floppy". --Wtshymanski (talk) 15:22, 6 June 2011 (UTC)

There was a drive for the bbc micro called the phloopy. I quick google finds http://www.stardot.org.uk/forums/viewtopic.php?f=3&t=2913 with a copy of an advert that describes the basic principle of operation. Plugwash (talk) 17:28, 23 January 2012 (UTC)

Archived discussion

You can find about 100K of archived discussions from approximately November 2007 to August 2010 at Talk:Floppy disk/Archive 2, or the navigation link after "Archives" in the header box above. --Wtshymanski (talk) 15:38, 6 June 2011 (UTC)

Proposed split for section "non-standard formats" to new article "Floppy disk variants"

I presume that User:Thumperward is making the recommendation to split the article by moving out the section "Non-standard formats" to a new article "Floppy disk variants." I do believe this current article is very long, but should we also consider moving the data from the table above with all the other non-standard sizes into that new article as well? I think the current article is getting some very good updates and modification from User:Thumperward. Maybe with some additional changes we can get this article back to FA status. § Music Sorter § (talk) 07:09, 23 June 2011 (UTC)

• Approve § Music Sorter § (talk) 07:09, 23 June 2011 (UTC)

Thanks for your comments! The article still needs a massive amount of overhaul (if ever there were an article which showed how our FA standards have gone up, this is it). Simply moving all the weird minor formats out will help a lot with that; they can gradually be re-imported if need be. Anyway, I'll get on this on my next pass; I don't think we need a formal resolution for it. Chris Cunningham (user:thumperward) - talk 09:42, 23 June 2011 (UTC)

I just did a brief review this morning of the changes and I think this is going very well. I think Tom94022 and I (among others watching this page) have quite a bit of history between us in the computer industry, so we can also work on the background/expert elements that might need some work still. I was going to work on the sources to see if I can move the Biblio into inline references and then find the missing ones still present. § Music Sorter § (talk) 15:17, 23 June 2011 (UTC)

This might be unrelated, but I just put mergeto tags on IBM Extended Density Format and fdformat, with corresponding mergefrom tags in Distribution Media Format. The 2M variant also belongs to this zoo, but that article isn't too bad. –89.204.153.130 (talk) 08:06, 5 July 2011 (UTC)

There's really nothing bad in a zoo-like article, especially when dealing with floppies. The entire field can't be sinthetized in a simple way and there's a lot of detailed tech-info that we don't have to remove just for the sake of article-elegance. Blackvisionit (talk) 12:28, 5 July 2011 (UTC)

Intro section order

I think the right intro order should be

• Intro (History): ubiquity / decline / lagacy
• Tech details: from general to particolar.

Blackvisionit (talk) 16:54, 23 June 2011 (UTC)

Merge discussion: Flippy disk

I've suggested that we merge Flippy disk into this article. "Flippy disks" are basically just modified 5.25" floppy disks, representing a relatively limited portion of the market for all floppy disks. The Flippy article is unreferenced (and has been tagged as such since 2007), and its contents could easily be added to this article without bloating it too much. — KieferSkunk (talk) — 05:38, 7 July 2011 (UTC)

This article is too big as it is and there is some useful albeit obscure info in the Flippy Article. If the consensus is that the obscure information is not notable then we should just nuke the Flippy Article and keep the one sentence in this article that mentions Flippy. In summary, my recommendation is not to merge but if there is consensus we could nuke the Flippy article. Tom94022 (talk) 06:36, 7 July 2011 (UTC)

Yes on Floppy disk variants a good idea Tom94022 (talk) 18:22, 7 July 2011 (UTC)

Perhaps better to merge Flippy disk into floppy disk variants. Rwessel (talk) 07:36, 7 July 2011 (UTC)

No on Floppy disk, yes on Floppy disk variants. I agree with Tom94022 that the Floppy disk article is already too large. The small piece already in this Floppy disk article can link to the section on the floppy disk variants article that would include the information on the flippy. I also agree that if we move it we should also see what we can confirm with sources. I do remember most of what it says, so I am not doubting much of the technical elements. Here is one source so far. InfoWorld, January 25, 1982, page 12 talks about the basics of the technique. Then we can put source needed on the other points. § Music Sorter § (talk) 08:14, 7 July 2011 (UTC)

Absolutely yes on Floppy disk variants. I've just read the Flippy article and it only needs some reordering. Concepts are clear and interesting. Blackvisionit (talk) 11:37, 7 July 2011 (UTC)

I actually disagree on sending it to Variants. The majority of the variants in that article deal with either physical differences in disk design (shape, size, etc.), data formatting differences, or variants on the disk controller technology used. "Flippy disk", on the other hand, is really just a simple hack to a standard format that took off as a marketing force. The article on Flippy could be just as informative at about half the length since a lot of the info there has to do with how people could convert standard floppies into "flippies". It seems that "flippy" really has more to do with the base format than a specific company's variant. — KieferSkunk (talk) — 19:06, 7 July 2011 (UTC)

You're actually right. Flippy usage of floppies should be more floppy-related than floppy variants-related but a lot of people is complaining about the article size... merging it to floppy variants could avoid a new tranche of discussion. Blackvisionit (talk) 20:33, 7 July 2011 (UTC)

Why can't we just condense the Flippy article into a few encyclopedic sentences?

“ Since 5.25" floppy drives only had one read/write head and could only access one side of the disk at a time, users had to flip the disk over in order to utilize both sides of the media. Users and manufacturers found simple ways to "convert" their single-sided floppy disks to dual-sided, effectively doubling their capacity using simple tools, and distributors began to use the term "flippy disk" to promote software that was distributed in this fashion. ”

Just a draft, but I think you get the idea. — KieferSkunk (talk) — 03:11, 8 July 2011 (UTC)

Absolutely no on condensing. I read the flippy article and learned, at least, a few things about the index hole problem and the double version use. Don't cut/remove the original article and, if you wish, add only a brief sentence/reminder into the main floppy article. Blackvisionit (talk) 09:20, 8 July 2011 (UTC)

Also, I should point out that if the article size is a problem, it probably means that the article's scope is too large and it should be split up into smaller articles. I haven't had time to read the whole thing, so I can't speak to its scope at the moment. But there are plenty of Good Articles, even Featured Articles, that are HUGE, and yet stay within their scopes quite well - there just happens to be LOTS of encyclopedic information in them on their topics. If the Floppy disk article's contents are well-focused and encyclopedic, then size shouldn't be an issue, nor should merging in additional relevant information. — KieferSkunk (talk) — 03:18, 8 July 2011 (UTC)

All three article's titles refer to the disk and not the drive so I don't feel to bad about banishing Flippy to Floppy variants. Note the Flippy article describes at least one drive variant. I can live with the short sentence - but it need work along these lines:

Single sided 5.25" floppy drives with one read/write head could only access one side of the disk. Mechanically an upside down disk would fit but could not be read because of mechanical asymmetries. Users and manufacturers found simple ways to "convert" their single-sided floppy disks to dual-sided, effectively doubling their capacity using simple tools, and distributors began to use the term "flippy disk" to promote software that was distributed in this fashion.

Tom94022 (talk) 05:44, 8 July 2011 (UTC)

I would propose that wherever we end up putting the information we don't lose so much of the information. May I propose a more detailed version of Tom's proposal with

Flippy disk
Single sided 5.25" floppy drives with one read/write head could only access one side of the disk. Mechanically a single-sided disk would fit in a drive if it was flipped upside down, but drives which used the index sync hole could not read the disk, and no drives could write to the disk because the write-protect notch was not present on the opposite side.

Three solutions existed for this situation enabling a doubling of the storage capacity of the user's floppy storage. First, at least one manufacturer created a floppy drive for the TRS-80 system with two index sync sensors and two write protect sensors linked together to identify if the disk was write protected before trying to write data to the disk. The second option was to purchase the appropriate double-sided floppy disks (with or without the index sensor hole) at a higher cost than the single-sided disks. The third option was the most economical using a special disk notcher or disk doubler tool or a simple template to note where to cut or standard hole punch the disk sleeve creating the needed write protect notch so the single-sided disk could be flipped over. Drives with the index sync sensor also needed that hole punched, which was a much more complex process and less often done by users.

Software manufacturers were known to distribute software on both sides of these flippy disks with each side formatted for two different computer manufacturers. When the cost of the second head was low enough to add to all floppy drives, the flippy disk became obsolete.

If we are going with this we need to get it right. I don't understand the 2nd option. Doesn't a double sided (DS) disk has the same problems in a single sided (SS) FDD as a SS disk - the index hole is in the wrong location when the disk is flipped over. The only thing a DS disk does do is give a good side 1. Some early SS FDs had a defective surface on side 1 but very quickly it just was an untested side. Also it's not the cost of the head but the price difference between the DS FDD and the SS FDD that killed the advantage of a Flippy Tom94022 (talk) 05:51, 18 July 2011 (UTC)

2nd option - This is about disks manufactured to be a flippable disk. If the drive is hard sectored (requires the index hole and had multiple holes in the media) it had a second hole through the plastic covering. Soft sectored drives came in two variates with one requiring the index hole and one embedded what it needed on the media with the user data. If it was soft sectored and used the index hole it needed the second hole. If it was soft sectored and did not use the index hole any disk could be flipped over if the write protect notch was there. The way I wrote the text above was to indicate you could purchase "flippable" disks that did not require user modification. I can search for an image of a hard sectored flippable disk or the soft sectored that needed the index hole. On the cost of the head comment, it is actually the head and associated support mechanism that is added to a double sided drive which increases its cost and ultimately dictates the price difference to the user. Are you proposing an adjustment in that sentence? § Music Sorter § (talk) 15:30, 18 July 2011 (UTC)

2nd Option - The terms Single Sided (SS) and Double Sided (DS) are terms of the floppy disk art so the use of DS in this context to describe a particular Flippy is confusing. I suspect that a Flippy FD cost at least as much if not more than a conventional DS FD for several reasons, in particular the much lower volumes at the same or slightly higher manufacturing cost as the DS FD (the manufacturer has to test and yield both sides). It is my understanding that virtually all FD controllers use the Index for formatting and most (if not all) do not use the Index for reading and writing and thus I am not sure of the relevance of this to the discussion (i.e., TMI). Likewise, I recall there were very few hard sectored FDs and to the best of my recollection they only had one hole in the jacket for Index/Sector. The index was between two sector pulses. Most drives had a jumper that put them in index or sector mode. Again maybe TMI

Cost vs Price - Yes, something like, "Initially the price of single sided Floppy Disk Drives was substantially higher than that of double sided Floppy Disk Drives but the price differential lowered to the point where there was little demand for the single sided drive and therefore ultimately little demand for Flippy disks." but getting a reference for this might be difficult. Tom94022 (talk) 15:53, 18 July 2011 (UTC)

I am good with Tom94022's proposal on the price/cost discussion. § Music Sorter § (talk) 03:38, 19 July 2011 (UTC)

"drives which used the index sync hole" - is another place that needs fixing. I believe all 5¼-inch drives use the index to measure up-to-speed and are pretty much indifferent to Index otherwise. Use of Index for formatting, reading and/or writing is a controller issue, not a drive issue nor an OS issue (other than perhaps Apple II and its Wozniak controller - the "IWM"). Tom94022 (talk) 16:45, 18 July 2011 (UTC)

I can verify that the index hole was completely ignored on the Apple II/IMN disks - the signal isn't even on the cable conntecting the drive to the controller.Rwessel (talk) 17:05, 18 July 2011 (UTC)

I haven't looked but I'll bet Apple II disks had an index hole. I think u mean Index was not used by the IWM but wasn't Index used in the drive for up-to-speed? Although it wouldn't surprise me if Woz just waited a long time and thereby saved the cost of the transducer in the drive. Tom94022 (talk) 21:36, 18 July 2011 (UTC)

Index holes were used on some hardware. The flippy required the index hole on both sides of the hub only when the hardware was using the index hole. Here is an image of a flippy diskette with the two index holes. I can confirm the 1980-84 Atari computer line did not use the index hole. We used standard single sided disks simply by cutting the write protect notch and flipping them over. For our SW distribution on both sides, someone already uploaded an example Tandy Flippy image [[File:Flippy_floppy.jpg]]. Somewhere I have a hard sectored floppy disk with the multiple holes in the media, but only one hole in the plastic housing, but that is a different discussion. § Music Sorter § (talk) 03:38, 19 July 2011 (UTC)

I just ran into this source working on another article that talks about Apple removing the hardware required with hard sectoring and did not using any timing hole. § Music Sorter § (talk) 05:10, 19 July 2011 (UTC)

I just checked the schematics of both the digital and analog boards in the Disk II (included in the Apple DOS reference manual), and there was no connection to the LED for the index hole. The hardware *might* have been present on some drives (since Apple bought standard mechanisms, at least in the early days), but it was unused. And as Music Sorter pointed out, we used the back sides of diskettes on Apple IIs without making a second index port (and inverted, the index hole port in the envelope was obviously in the wrong location to do anything). Since Apple Disk IIs used standard 5.25 diskettes, there *were* index holes in the diskette itself. Rwessel (talk) 19:28, 19 July 2011 (UTC)

And I propose we keep the graphic of the notcher to go with it. Now I also propose we collect some sources. I noted one above and I did see some adds referencing the different floppy formats (SSSD, SSDD, etc). § Music Sorter § (talk) 08:38, 8 July 2011 (UTC)

My assessment of the above is that a majority are supporting Floppy disk variants as the target article to hold the Flippy information. As long as we move it there I propose we keep the more detailed version of the content I proposed above. § Music Sorter § (talk) 04:02, 18 July 2011 (UTC)

Heavy editing

Don't want to start an edit war, but there has been a few heavy edits by Freywa (14 - 15 July 2011) that sounded (to me) too much disruptive and anti-WikiLove. Maybe some discussion could be clarifying. Blackvisionit (talk) 11:36, 15 July 2011 (UTC)

This article needs editing, and I found the previous version to be better organized. --Wtshymanski (talk) 13:44, 15 July 2011 (UTC)

The version we're talking about is not a simple editing or reorganizing because it also performs a lot of unnecessary cuts. You should also wait before restoring it, wait for a wider consensus. Blackvisionit (talk) 13:57, 15 July 2011 (UTC)

What (of value), specifically, is missing in the shorter version? --Wtshymanski (talk) 14:31, 15 July 2011 (UTC)

It was a pain restoring the missing info, due to the size of the editing diff. Next time it would be a kind thing to apply a few small changes instead of a single mass edit. Blackvisionit (talk) 17:35, 15 July 2011 (UTC)

Do we need to list all the microcomputer operating systems here? It's misleading to imply that a computer that loaded CP/M was loading a "more advanced operating system" - CP/M machines usually had *no* ROM "operating system", you were lucky if you had a hex monitor. And there was Apple DOS and TRS DOS and HDOS and MDOS and MSX and Amiga DOS and probably a dozen others; plus at least one very popular home system didn't load a DOS off diskette at all. Shorter, in this case, is also more accurate. --Wtshymanski (talk) 18:33, 15 July 2011 (UTC)

I don't agree at all with your analysis. A few links are a starting point to the jungle of OSs. Interested users will then go on exploring, without the need of a complete list... However this little edit shouldn't be worth a discussion. Blackvisionit (talk) 18:49, 15 July 2011 (UTC)

I believe Freywa is not editing in a civil manner as indicated my his massive edits here just a few hour ago. I have reverted them. I don't think he really is interested in editing civilly and if he keeps this up we will be forced to report this behavior. § Music Sorter § (talk) 08:53, 16 July 2011 (UTC)

I think you forgot to read the notice at the top of the article before reverting:

{{GOCEinuse}} and now is the copyediting drive. I am a copyeditor and I actually took this article for copyediting. Regarding the major changes, here is why I did what I did:

• The table has been removed because it looks like it does not regard the floppy disks themselves, but rather the types of floppy disks.
• The sections have been renamed because their former groupings do not seem to fit with their contents.

I hope you listen to this. FREYWA 09:27, 16 July 2011 (UTC)

User:Freywa, I think there may be some mistake.

• The articles needing copyediting under F does not appear to list Floppy disk, so I don't know why you said it did.
• Regardless of the article's status in a copyediting guild list, major changes to sections of an article that have been present for some time should be reviewed with other highly interested parties in the article before massive edits are made.
• If you would like to make various grouping changes, please consider doing them separately and not as one massive edit that would be considered controversial.
• Please consider re-reviewing WP:COPYEDIT to better understand the GCOE intended scope of edits. The goal of the GCOE is to improve the grammar and prose of the article. It does not provide unbridled editing without consensus.
• As a copyeditor I would expect a better explanation than (Man overboard!) to a major deletion of 9,000 bytes of article.
• The table you deleted is about floppy disk formats in a section titled Floppy disk#Formats. I don't think it can be any more relevant. In either case it is your opinion that it did not belong, but I disagree. Now if you proposed we move the section to a different article we can discuss that back here where I originally proposed. That is how all other articles are reviewed and updated with so many interested editors.

I appreciate your understanding. § Music Sorter § (talk) 09:57, 16 July 2011 (UTC)

I'm going to think that User:Freywa's edits are deletionist and non-cooperative. He has to accept our right to edit his edits. I'll repeat it twice: please don't do mass edits and discuss relevant changes before doing them. Blackvisionit (talk) 11:43, 16 July 2011 (UTC)

Another little thougt : last mass edits are, as Wtshymanski pointed out, too wordy. Floppy disk is a technical article, shouldn't be edited like prose. Blackvisionit (talk) 12:10, 16 July 2011 (UTC)

Fine, you editors win. Revert everything back to where it originally was (here). But I leave you one little request: join the guild and copyedit this article for me, as you wish. Signing off. FREYWA 13:41, 16 July 2011 (UTC)

Nobody said there's a war or you're not allowed to cooperate. I think that latest joined revision could be seen as 'clearer' than before your editing. Blackvisionit (talk) 14:36, 16 July 2011 (UTC)

Discussion of constructive edits

As long as we are talking about edits, we might as well review a few to come to consensus.

• The Expert needed template seems out of place with so many active editors who appear to be experts in this article. Does anyone know if that was added for something in particular?
• We have a number of Citation required notes. Should we go through and either find a citation or remove the copy?
• There are probably a few of the changes Freywa was making that would enhance the article, but I think there were too many major changes to sort through the good mods. Maybe we can each review a paragraph or section and make a few edits at a time for what we find compelling. In fact it might be good to review the article against the GA criteria to see if we can get it back to that status.

§ Music Sorter § (talk) 16:08, 16 July 2011 (UTC)

Well OK, I got the idea. Regarding the template

I don't have a floppy and I don't know much about floppys. Anyone out there?

Citations: I think that citations may be found using Google Scholar and Google News. Anything where no citation is found, remove it!

For the sections, I think that the last content section (Impact and legacy) should be moved to the History section, as several of its details pertain to history. By the way, I would say that I was here to fix the prose; extraneous words like also and however should be removed, for example. We must finish this by July 31, or else I won't get credit for this article. FREYWA 02:18, 17 July 2011 (UTC)

Freywa, I sincerely appreciate your constructive feedback on the talk page. Regarding your prose edit comments, I don't have a problem with changes like that since they are in general the idea behind improving the article and the GCOE. If you can focus on those types of adjustments, many of us (myself included) are working on finding the necessary citations/sources for each necessary paragraph. I ask that you do not delete those sections at this time. Please give us an opportunity now that we are doing per the expert-subject template. Specifically, your comment to remove also is generally a good removal, but please be careful of changing the meaning of a paragraph by removing too much text because it matches a word you often remove. If you are following WP:COPYEDIT, I don't see where it says to remove also and however. Maybe you are reading a different recommendation. I like staying up on these copy edit guidelines and would love to see that reference if you can help me.

For your idea to move the Impact and legacy sections, can you create a new talk page section with that proposal and we can give support and oppose comments to that big move. I have not looked at the details, but you may be spot on.

Thanks again for this very constructive interaction. § Music Sorter § (talk) 05:42, 17 July 2011 (UTC)

I am referring here. Part of it says:

1. Simplify the language and make it more direct. The article will typically be a little shorter when you are done.
2. Remove peacock terms and flowery language unless backed up by the sources.
3. Remove extraneous words. The words "also" and "however" can almost always be removed. Phrases like "due to the fact that" or "in the year" can typically be modified or omitted.
4. Make corrections to grammar and spelling. FREYWA 10:04, 17 July 2011 (UTC)

Freywa, I Ihope that's writing ' We must finish this by July 31, or else I won't get credit for this article. ' is only a joke... we're here to share free knowledge not win any price. We've got all the time in the world. Blackvisionit (talk) 10:52, 17 July 2011 (UTC)

I also hope getting credit for the article is secondary to improving it. Thanks for the link to your source for the guidelines you are following, that is helpful. I am also fine with your last change. It is nice to be able to see the edits is small chunks to more easily recognize what is different. § Music Sorter § (talk) 16:30, 17 July 2011 (UTC)

I think Freywa's idea for the Impact and legacy section was so obvious I made the change already. I even questioned if it should be added to the intro section, but thought this was probably just as appropriate rather than at the bottom of the article seemingly out of place. § Music Sorter § (talk) 17:25, 17 July 2011 (UTC)

I've moved it back down. Structurally this is not part of the "current use" section, nor of the history: legacy and history are separate things. The section needs significantly expanded to explain the floppy's place in history as the default storage medium for almost all mainstream computer use covering several decades. Some of that can perhaps come from the "ubiquity" section, but undoubtedly we'll need some new material as well. Chris Cunningham (user:thumperward) - talk 13:42, 19 July 2011 (UTC)

Reverting again a piece of Freywa's edit: the floppy materials list has to be complete and detailed. (Freywa, please, don't be so in a hurry...) Blackvisionit (talk) 10:12, 18 July 2011 (UTC)

Reverting again some deletions from Freywa. Please can we slow a bit down? This forced editing process is a pain! Blackvisionit (talk) 11:02, 21 July 2011 (UTC)

All those trivial bits in parentheses do nothing but confuse the reader and bog down the prose. Freywa was right to remove them. Margin-scribbling for edge cases and random observations should be kept to a minimum and attributed to secondary sources where really required. Chris Cunningham (user:thumperward) - talk 11:11, 29 July 2011 (UTC)

Sorry, disagreed. This is encyclopedia, not a prose. Devil is in details. If you want citation, please place tag. Outright deletion is for clear nonsense. The bits you deleted was written by someone of certain expertise. This person may be mistaken if he wrote off his head (so you may want to request an independent verification), but deletion of detail just because you don't like without discussion is not an option. "Trivial bits" is your personal derogatory judgement; other people call it "techical details". Muslim lo Juheu (talk) 15:32, 29 July 2011 (UTC)

You're right. There seems to be users (Freywa docet) that don't understand that simplifying a tech-article is something very difficult. Cutting off details is not simplifying, is vandalism. Had the same problem with thumperward in this other floppy page - Floppy disk hardware emulator. Blackvisionit (talk) 16:26, 29 July 2011 (UTC)

Consider this a second warning for labelling good faith edits as vandalism. This article is a collossal mess primarily because the "experts" who frequent it have been far more interested in dumping their knowledge into it en masse than making an accessible article for readers to obtain information from. The capacity section is a perfect example of this: it prattles on so much that a reader who wants to do something so vulgar as, I don't know, find out how floppy disk capacity is measured will give up about halfway through the paragraph. And then there are all the elementary typographic problems, such as using "3.5" and "3½" arbitrarily throughout. Chris Cunningham (user:thumperward) - talk 22:04, 29 July 2011 (UTC)

Colleague, please consider giving more respect to those who are "dumping their knowledge" into wikipedia. Please don't forget that we are all volunteers here, all are doing what they can and wish to, no more, no less. Handling "3.5" vs "3½" issue is commendable. However many people don't have the ½ button on my keyboard (just as I don't have any key :-), and I don't think they are going to jump over the hoops to find it. I agree that the capacity section is a perfect example. However in order to fix the problem you duly noted you don't have to decimate the text. You have to rearrange it per wikipedia:summary style and per "inverted pyramid" metaphor. 00:44, 30 July 2011 (UTC) — Preceding unsigned comment added by Muslim lo Juheu (talk • contribs)

Using your system administrator role to warn people expressing a different editing opinion is an abuse. I'll report it if you go on like this. If vandalism sounds too rude to you (WP:ACCUSE) then excuse me. Nevertheless more than a single user felt your edits as non-constructive or whatever term you like. I think that this is an excellent article, perfectly readable. Floppy-unexperienced users have a real chance to discover an almost forgotten part of pc history. Blackvisionit (talk) 22:38, 29 July 2011 (UTC)

Guys I feel uneasy seeing you in this section instead of discussing my suggestion about a major change/improvement of the article presentation structure, see #"Formats" section must be dismantled. Of course I could have done this without asking you, but I feel this task needs discussion and cooperation. Muslim lo Juheu (talk) 00:49, 30 July 2011 (UTC)

CAV image discussion

In the Floppy disk#Sizes, performance and capacity section the second and third paragraphs speak about constant angular velocity so I added the image that gives a visual to the two paragraphs.

Tom94022 disagreed with my addition on the basis of it being TMI and not relevant to the article. Maybe there might be a more simplified image that would be more appropriate, but I have not seen one. I thought we could get some additional opinions on the image being in or not. § Music Sorter § (talk) 03:08, 18 July 2011 (UTC)

Way, way too much information IMO. Diagrams should be simple, attractive and directly related to the material in question. This ons is trying too hard to cover history, geometry and capacity at once, and the layout is suboptimal at best. Chris Cunningham (user:thumperward) - talk 13:44, 19 July 2011 (UTC)

I removed it because it covers all disk storage; therefore belonging in that article only and is TMI in this an FD article. Tom94022 (talk) 17:08, 19 July 2011 (UTC)

Non-representative pictures

I feel that these are not representative images and should be replaced by pictures showing more typical hardware; cf. the reasons given here:

http://en.wikipedia.org/wiki/File:1.2_MB_2.88_MB_floppy_disk_drive.jpg

http://en.wikipedia.org/wiki/File:2.88_MB_floppy_disk_drive.jpg

http://en.wikipedia.org/wiki/File:1.2_MB_floppy_disk_drive.jpg

31.16.112.242 (talk) 13:10, 24 July 2011 (UTC)

I don't know what pictures u think are not representative, but the ones you link to are pretty ugly so I would oppose using any of them. Tom94022 (talk) 19:06, 24 July 2011 (UTC)