Jump to content

Talk:Parallel ATA/Archives/2017/September

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


"AT" vs "Advanced Technology", yet again

The simple statement that the AT Attachment standards do not contain the words "advanced technology" is completely sufficiently sourced by reference to those standards - which are linked elsewhere in the article. This is a simple, easily verifiable claim of fact, no different from the statements of bandwidth and features that are also referenced to the same documents. Yes, it's a primary source, but this is an example of a primary source used correctly (as with the statements of bandwidth, etc.).

Meanwhile, the claim that "AT" is a "reference" to "IBM's "Advanced Technology", the interface standard Parallel ATA evolved from" is bogus - there was no "interface standard" called "advanced technology". Jeh (talk) 21:34, 3 September 2017 (UTC)

I tried to track down these claims and found no support for them. Many books state this is called "the ATA (Advanced Technology Attachment) standard" [1]. Please provide reference for the claims. Fountains of Bryn Mawr (talk) 21:58, 3 September 2017 (UTC)
Yes, it's an extremely common misconception. Nevertheless those words don't appear in the standards docs, which are linked from the article. And the only claim in dispute here is the claim that those words don't appear in the standards docs. It is not OR nor SYNTH to say that they don't appear when a simple text search proves that they don't appear. Perhaps we can use your book references as documentation for the misconception. Jeh (talk) 22:03, 3 September 2017 (UTC)
I agree with Jeh's analysis. Futhermore, it is my recollection that this was an issue in the very earliest days of the ANSI committee and it was explicitly decided to disassociate AT Attachment with Advanced Technology Attachment so as to avoid potential trademark issues with IBM. It's probably somewhere in an ANSI X3T10 meeting minutes circa 1989. Tom94022 (talk) 06:36, 4 September 2017 (UTC)
The "interface standard" (not the best term however) is ISA which ATA is directly derived from. ISA was introduced with IBM's Personal Computer AT where "AT" stood for "Advanced Technology". "AT Attachment" is - obviously, but not officially - a reference to this. I thought using a source where the author makes this connection is a viable way – apparently not.
The question is how do we express this here? Not at all? Without the "interface standard"? --Zac67 (talk) 11:02, 4 September 2017 (UTC)
That's incorrect. The 8-but bus for the original IBM PC was simply called "the PC bus". With the AT IBM widened it to 16 bits, otherwise nearly identical. The PC-compatible industry didn't want to give IBM free advertising or a trademark infringement case (sound familiar? ;) ) so instead of calling it the "PC-AT bus" or just the "AT bus" they came up with the name ISA, Industry Standard Architecture. Nobody called it the "advanced technology bus". But the definition of the ISA bus didn't say a dang thing about how you'd build a disk controller to plug into it. The host-controller interface defined by WD when they built those HDs was no more "directly derived" from the ISA bus than the HCI on various other interfaces available at or about the same time (ST506, ESDI, SCSI).
All that was new with WD's "IDE" drives was that the thing that plugged into the ISA slot was basically just a slot extension; the HCI was moved off to a card on the back of the HD (or in the first implementations, to another card that was mounted adjacent to the HD), and the cable from the ISA slot to this card was carrying "raw" ISA signals. Jeh (talk) 17:37, 4 September 2017 (UTC)
Whatever IBM called it doesn't matter. ISA is the slot standard IBM introduced with the AT, named by their competitors. ATA is a cut-down ISA slot, as you say. The controller behind it on the HDD PCB was initially an MFM/RLL controller equivalent with few improvements ("Integrated Drive Electronics", literally). But this doesn't answer the questions. --Zac67 (talk) 20:54, 4 September 2017 (UTC)
The history is a bit more complex. WD was privately working on the concept of "Integrated Drive Electronics (IDE)" or "Intelligent Drives (ID)" as early as 1985. Ken Hallam of WD made the concept public in 1986 by proposing at Buscon the "Direct connection” to PC AT bus using minimum 40 pin connector" - no mention of IDE. The first public announcement of what became ATA-1 was Conner's June 22, 1987 announcement of the CP342 using "an embedded IBM PC/AT controller" or "integrated IBM PC/AT controller; "it was subsequently described as a "40 pin host interface" (CP3022 spec, 2/1988). Miniscribe and Quantum were also an early suppliers both not using the term IDE but some form "Integrated drive and AT electronics" or "40 pin … AT partial bus connector." The CAM committee which led to the published standards started with “AT Controller Interface,” “AT Bus Interface” but by March 1989 adopted ATA as AT Attachment. IDE did not come into general usage until 1990 or 1991 well after ATA was in the public domain thru ANSI.
So the first "standard" is not IDE but in fact ATA and the best place to look for its definition is the American National Standards publications which from ATA-1 to (P)ATA-8 have always included a clause 3, Definitions which inludes this express limitation, "For the purpose of this Standard the following definitions apply:". Here are the relevant definitions
  • ATA-1: 3.1.1 ATA (AT attachment). ATA defines a compatible register set and a 40-pin connector and its associated signals
  • ATA-2: 3.1.1 ATA (AT attachment). ATA defines the physical, electrical, transport, and command protocols for the internal attachment of block storage devices.
  • ATA-4: 3.1.1 ATA (AT attachment). ATA defines the physical, electrical, transport, and command protocols for the internal attachment of storage devices
  • (P)ATA-8: 3.1.1 ATA (AT attachment). ATA defines the physical, electrical, transport, and command protocols for the internal attachment of storage devices to host systems.
As near as I can tell the term "Advanced Technology" does not appear in any of the ANSI published standards, draft standards or committee minutes; some early drafts of ATA-1 did define ATA as "AT Bus Attachment," but this never made it into the approval process. IBM uses AT as an acryonym for "Advanced Technology" but ANSI never did. So wouldn't it be impremissible synthesis on our part to say it is obvious that the ANSI committee members meant "Advanced Technology" when they used the term AT even though they never did - I think so. Therefore, such synthesis by a source makes the source unreliable for this fact and should neither be quoted nor referenced in Wikipedia. Perhaps there could be a note to the effect that some sources incorrectly assert AT in the context of (P)ATA or SATA interfaces stands for Advanced Technology, but I'd rather not go there. Tom94022 (talk) 22:53, 4 September 2017 (UTC)
This conversation is long on claims and very short on reliable sources (in fact there is only a few... all supplied by me):
  • "I agree with Jeh's analysis"--- and there lies the problem, Wikipedia articles do not rely on analysis, its a core Wikipedia policy, WP:OR
  • "Perhaps we can use your book references as documentation for the misconception" - err... no... we can't. We can only use a reliable secondary source stating there is a misconception.
  • Stating "AT" does not mean "Advanced Technology" based on primary sources is an interpretation of that primary source and "Any interpretation of primary source material requires a reliable secondary source for that interpretation."
So we are back to square one - Please provide reliable secondary sources for the claims. Fountains of Bryn Mawr (talk) 00:53, 5 September 2017 (UTC)
Ah, but you see, the claim you're disputing did not say "AT [in "Parallel ATA"] does not mean 'Advanced Technology'". The text you slapped a fact tag on merely stated "but the ATA specifications published by the several standards committees simply use the name 'AT Attachment' with no reference to advanced technology." This is simple fact. There is no "interpretation" (or "analysis" for that matter) of the specifications required to read or search the specs and see that no, the phrase "advanced technology" does not occur in them anywhere. Those words just aren't there in the specs and we don't need a secondary source to agree that they're not.
I cannot understand why you are insisting on a secondary source for a claim that the article was simply not making!
As for "jeh's analysis" I am certain that Tom94022 was referring to my analysis of the issues in the discussion, not of the sources themselves. Jeh (talk) 02:46, 5 September 2017 (UTC)
Would it help if I had said I agree with JEH's statements of facts and his conclusions therefrom?
Fountains of Bryn Mawr chooses to ignore the eight plus ANSI Standards which are reliable sources that define ATA as "AT Attachment." I have now referenced them in the article. Tom94022 (talk) 06:35, 5 September 2017 (UTC)
Darn - I was going to insist that FoBM do that. The URLs were right there in the standards section, so it's not as if it would have taken him much time to find them. (I'm really tired of tag-drops that amount to "you others, you need to do some work.")
Meanwhile, there's nothing in the references given that says that "AT" in "AT Attachment" was a reference to "advanced technology". To the IBM PC/AT, yes, but not to "advanced technology". Those words don't appear on the PC Guide page and the InfoWorld article doesn't mention "AT Attachment" or even make any particular mention of the HD's interface by any name at all. Nor was "advanced technology" even part of the PC/AT's name. It was just the "PC/AT". So there's that. Isn't it OR to connect those dots? Right back atcha, FoBM. Jeh (talk) 07:09, 5 September 2017 (UTC)
I do remember IBM referring to the AT as "Advanced Technology"; this seems to be accepted, the according article doesn't source this. However, as a kind of middle course, how about stating that "AT Attachment" refers to the IBM PC/AT (or is commonly conceived this way – we'd need a source either way) and leave the connection to "Advanced Technology" for the reader to discover in the linked articles? --Zac67 (talk) 10:58, 5 September 2017 (UTC)
The InfoWorld article does mention "Advanced Technology" but doesn't say whether it was an official IBM designation or just something they said a lot. In any case those words were not part of the actual product name.
I fully expect to be told that we'll need a RS for what anything the reader might "discover" or infer. Jeh (talk) 11:49, 5 September 2017 (UTC)
I too am tired of tag drops so at this point I suggest we see what FoBM does to the article. I did find in my files an explicit reference that IBM says AT == Advanced Technology so I changed the reference. I too started to list the ANSI standards pointing to the published versions available for a price at the IHS Standards store or the INCITS store where for example ATA-1 = INCITS 221-1994/ANSI X3.221-1994 but that was a bit tedious and they were already listed in Section 4.1. One very small issue is that the standards are for the most part properly named in the column but at least some of their links (and yours) point to draft documents and not the final published version. Tom94022 (talk) 16:37, 5 September 2017 (UTC)

We do not need to do excessive research trying to figure out what the wording in primary source documents mean. Secondary sources tell us what they mean. I think its fine to note what primary sources say but we have to reflect what reliable secondary sources say unless there is a very good reason not to. We do not need to know what the ANSI committee members meant when they came up with "AT Attachment", secondary sources tell us that. If a reliable secondary source says A and another reliable secondary source says B we have to "describe disputes, but not engage in them" per:WP:YESPOV) - but we are not even at that point because we are not confronting one secondary source with another.

Went ahead and edited problems re: those and other Wikipedia policy. I'll take them in the order they appear:

I have removed the note containing "It was never written out as "Advanced Technology Attachment", except in error." That statement of fact needs to be in (many) reliable secondary sources. Even if there is a reliable source secondary stating that, there are many that do not describe it as any kind of error and, in fact, state AT = "Advanced Technology" so we can not make any kind of direct statements one way or the other (WP:YESPOV #2).

Edited the lead: the lead of this article should define the term "Parallel ATA" (MOS:BEGIN). We can't leave the reader in the lurch and just ignore two letters in the term and just say "we don't know where they came from, they must have dropped from outer space or something". Since we have many many many sources[2] including dictionaries[3] and entire books on the topic [4] stating that it is "AT Attachment", "AT = Advanced Technology", and it is also refered to as "Advanced Technology Attachment", we must mention that per WP:YESPOV #5. "Advanced Technology Attachment" is also an incoming redirect (MOS:BOLDSYN).

The problem I ((fact)) tagged in the first body paragraph was an un-referenced statement. It has gotten a little bit worse[5], two different statements with two references connected with "but" that "implies a conclusion C that is not mentioned by either of the sources" (WP:SYNTH) with the added problem of the second source(s) not even being reliable secondary sources making a statement. I have separated the statements as a partial fix and cleaned up some WP:OVERREF, it was too much. Fountains of Bryn Mawr (talk) 14:14, 7 September 2017 (UTC)

Per WP:BRD, you should not be editing your preferred changes back into the article while discussion is still in progress. Your pronouncements here are not the last word. And your arrogance in positioning yourself as the final arbiter of Wikipedia policy here is not appreciated.
We will be remiss and misleading if we do not state clearly that the "horse's mouth" documents that define this interface do not use the words "advanced technology" in any of their multiple versions. And no, we don't need a secondary source for that.
If a secondary source flatly contradicts a primary source, with no "interpretation" of the primary required for that evaluation, then we are obligated to at least consider whether that "secondary source" is unreliable.
The fact that this interface is called "Advanced Technology Attachment" by many is indisputable. But the conclusion the reader may draw from the present text, that that name is officially supported by the defining committee, is false. The absence of the phrase "Advanced Technology" in the primary sources (and I do not support the POINTy removal of some of them) at the very least introduces an issue to which we must call attention. As it is, you are erroneously insisting on wording that implies that "Advanced Technology" is part of the official name when there is no support for that conclusion. Jeh (talk) 16:48, 7 September 2017 (UTC)


Again I find myself in agreement with the facts and conclusions stated by JEH and since we clearly have a dipute I reverted to the last version before the dispute. Regardless of whether the ANSI documents are primary or secondary sources, by virtue of their development process they are clearly the most reliable source as to the "definition" of ATA (one could argue the drive specs themselves are primary and ANSI is secondary, it certainly has a rigorous development and publication process, well beyond any peer reviewed journal). There are sources that say the earth is flat but one would not include them in Earth. In the spirit of consensus I would agree to a note in the lede to the effect that "Many sources use the phrase 'Advanced Technology Attachment' for what is described by the 'AT Attachment' standards issued by ANSI and other standards bodies." Tom94022 (talk) 17:18, 7 September 2017 (UTC)
Despite seeing a posting titled "attempt to gain consensus" all I am really seeing is allot of WP:OWN (and someone seriously needs to read WP:TPYES and maybe WP:TALKNO). Leads are for explaining something for an average reader, and should not read like a technical document (which Wikipedia is not). With a request to supply any secondary sources being ignored, a constant claim that (most?) of the secondary sources on this are unreliable, and equating what seems to be a majority of sources with trying to push a flat Earth theory, I think we are done here. Unless I am missing something this needs to be settled somewhere else. Fountains of Bryn Mawr (talk) 14:21, 8 September 2017 (UTC)

I for one won't miss FoBM but he does have a point that there are many references that incorrectly use "Advanced Technolgy" for what should be only "AT" so maybe to help the reader we should put in a note in the lede such as, "ATA is frequently called Advanced Technology Attachment but such a defintion does not appear in any ATA standard." or words to that effect? Any one of FoBM cites would support the note. I think the note is as far as we should go. Tom94022 (talk) 02:35, 10 September 2017 (UTC)

I'm good with that note, see my proposal above. --Zac67 (talk) 08:22, 10 September 2017 (UTC)