Talk:Peripheral Component Interconnect/Archive 1

This is an archive of past discussions about Peripheral Component Interconnect. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 1

Some More Specifications

Due to the fact, that the PCISIG is offering the specs only for its (paying ?) members it would be nice, if there could be some more information be included here. Of course such an article can not substitute for the lack of documentation. People serious about PCI will be PCISIG members anyway. But for some hobbyists and artists, anyway, people, who do not need full technical coverage, but a little more factual overview about the topic, it might be nice to get an extended information about the technology. Personally I would use the physical size specifications (I am designing a 3D model of a motherboard in a CAD program) and I guess, that some electronic-hackers might be interested in a link (or publication) of the signal paths on the bus.

Universal PCI

The page should definitely have some info on Universal PCI. I.e. PCI cards that accept both 5 V and 3.3 V. (June 22, 2005)

Specification table.

I think a table of specifications of the different pci versions would be nice.

PCI-X

Is it possible to put a "64 bit PCI-X" card in a "normal" 32 bit PCI slot? Will it work properly?

For the most part, yes Snickerdo 01:46, 8 October 2005 (UTC)

if it will fit it should work, the main issue is likely to be other components getting in the way of the overhanging part of the card. Plugwash 19:55, 11 January 2007 (UTC)

The 32bit-part of the slot has a 'REQ64#-signal line, which is used by the board to tell the card whether it's in a 32bit or 64bit slot' Ranma 16:09, 9 February 2007 (UTC)

PCI 64

Does anyone know anything about PCI 64? Is it a homonym for some other spec? I can't find any mention of it on the PCI SIG site, but Adaptec (amongst others) label some of their products as PCI 64. Mr. Jones 12:02, 17 Mar 2005 (UTC)

I'm guessing that means 64-bit PCI. 64-bit PCI slots are longer than 32-bit PCI slots, to hold 32 extra pins. --DavidCary 04:06, 21 Jun 2005 (UTC)

Forth, PCI and OpenFirmware

Perhaps someone could elaborate on the Forth boot code that was originally speced in the PCI standard?

I didn't think OpenFirmware, which is surely what you mean, was originally, or is, part of the PCI standard. It was common on Sun workstations and Apple Macs (and still is) and those machines also used PCI busses. But I don't think that either standard was included in the other. --drj

Well, according to 'Chapter 18: Expansion ROMs' of 'PCI System Architecture' (Third Edition, covering PCI2.1, 1997), the PCI standard defines Expansion ROMs to some extend, including a 'Code Type' field (Should be byte 16 after the "PCIR" signature): "This one-byte field identifeies the type of code contained in this image as either executable machine language for a particular processor/architecture or as interpretive code. A value of 00h indicates Intel ix86 (IBM PC-AT compatible) executable code, while 01h indicates interpretive code. The Open Firmware standard (reference IEEE standard 1275-1994) is used for interpretive code. The values from 02h through FFh are reserved. A basic description of the Open Firmware standard can be found at the end of this chapter". It is also possible to have "Multiple Code Images Contained In One Device ROM", which allows a device to contain _both_ x86 and openfirmware versions in one ROM. Ranma 16:06, 9 February 2007 (UTC)

Any thoughts on PCI 2.3?

http://www.pcisig.com/specifications/conventional/conventional_pci_23/

Somebody should also find out if PCI 2.3 motherboards DO support 5V cards, and then write about it. There seems to be a big confusion about the correct answer on that question. Not even PCI-SIG themselves writes clearly about this on their site. Or do they? At least the reality among PCI 2.3 motherboards, supporting or not supporting 5V cards, does not seem to be clear.

Afaict there is no such thing as a universal PCI slot (unlike with AGP), because all the cards must sit on one bus, 3.3V slots only seem to be seen on high end motherboards that support 66mhz 64 bit PCI and all cards i've seen are keyed for either 5V only or universal. ~~130.88.116.241 15:57, 5 February 2007 (UTC)~~ Plugwash 16:45, 5 February 2007 (UTC)

Dell - 64 bit?

I've just opened up a Dell server, and it has extended PCI slots. Are these 64 bit PCI slots, or something proprietry? If so a photo here would be nice. Rich Farmbrough 17:46, 8 Feb 2005 (UTC)

Almost certainly 64 bit slots Plugwash 15:46, 7 March 2007 (UTC)

Speed

So PCI 2.2 allows for a "peak transfer" of "533 MB/s"? say that you will be limited to 30 MB/s to 50 MB/s on a non X/Express PCI.--Jerryseinfeld 14:24, 18 Mar 2005 (UTC)

They're talking about 32-bit 33MHz slots. 64-bit 66MHz slots, even if not PCI-X standard, will still get considerably more bandwidth than a standard 32-bit 33MHz slot found in most desktop PCs. Snickerdo 01:48, 8 October 2005 (UTC)

It is incorrect to say speed here; it's capacity, something totally different. The whole template in the article is wrong. UPDATE: The template is now fixed.

PCI electrical power ratings

Can someone add this info? How much current can a PCI card draw ?

Aha, I found this two mail messages : http://www.pcisig.com/reflector/msg05240.html http://www.pcisig.com/reflector/msg05243.html

Can someone add this to the article ?

--195.250.201.212 13:07, 8 October 2005 (UTC)

Low Profile & Half-height PCI

- There is no mention of 'Low Profile' PCI cards. (anon) - could the half-heigh dimensions be added to 'Size of PCI extension card' ? -aug 25 '06. (anon)

I have attempted to cover Low Profile PCI specs, which I believe is the same thing as "half height". (It's not actually half, more like 2/3). I added links to the PCI specs I got this from, which are difficult to interpret. Had to figure this out for our own PCI card ... Aaron Lawrence 04:55, 12 July 2007 (UTC)

rates

This site gives rate of 528 decimal MB/s, not 533. Someone who knows these well double check. — Omegatron 15:46, 29 December 2005 (UTC)

I'm a bit confused here too. AFAIK, 1 KB = 1024 bytes, 1 MB = 1024 KB. In the paragraph "Conventional PCI bus specifications", the calculation is "33.33 MHz × 32 bits ÷ 8 bits/byte = 133 MB/s" but that's really just 133.32 million bytes. So, if you break this down to MB according to the rates above, shouldn't you get 127 MB/s?

ADude, 11 May 2006

Yeah, your math is correct. It's the same conspiracy again hard drive manufacturers saying 1.0 GB when they really mean 1,000,000,000 bytes. Rmcii 04:47, 12 May 2006 (UTC)

It's not a conspiracy. The only data capacity where it makes sense to measure in base-2 multiples, is RAM, when you're talking about row/column based binary addressing. From an engineering point of view it does NOT make sense to measure a hard disk in base-2 multiples, because an engineer will want to know the true number of sectors*bytes! This is because the natural, intrinsically sequential nature of that storage medium is NOTHING LIKE addressing RAM - it literally is just a big stream of bits. Likewise with serial communications specifications - it makes NO LESS SENSE to count the number of bits coming out in base-10, than it does to do so in base-2 multiples that software-types seem to cling to so much... in fact from an engineering point of view base-2 multiples would be very annoying (there are many interesting calculations that can be done with base-10 multiple specs, but base-2 multiples distorts the true number of "symbols" we're trying specify and have to be converted back to base-10 multiples every time!). 59.167.116.194 15:49, 2 September 2006 (UTC)

The 133MB/sec is a reflection of the nature of the medium for the engineer's benefit, not the consumer. It's not so clear-cut in this case: should they have expressed the bandwidth in Mbit instead of Mbyte, as most comms specs are? The answer is no, that wouldn't make sense, this is a parallel data bus 32-bits wide, not some 2-wire serial interface or radio transmission... so unfortunately, it's one of those grey areas where you want to talk about ordinarly communications bandwidth but the parallel nature of the data bus is designed from the point of view of getting whole words at a time from point A to point B. In this case, my engineering training would say that the solution is to express the bandwidth in Baud, aka "symbols per second" - each "symbol" being one 32-bit word, in which case you're left with 33.33 million baud. But then this doesn't help understand the data transfer problem any better - anyway, my point is there is no conspiracy ;-) 59.167.116.194 15:49, 2 September 2006 (UTC)

PCI Add-On

There should be a part in this page telling people what can be put on the PCI slots, That would make life easier for people wanting to know more about PCI. Alkady 16:44, 5 February 2006 (UTC)

I'll second that. I can think of Sound cards (for audifiles or back when AC'97 wasn't a standard MoBo feaure), Ethernet cards (again, before it was a standard), wireless cards, Raid controller, SCSI cards... But that's it. There's more right? 64.238.49.65 17:26, 25 September 2007 (UTC)

Ethernet and sound cards are still made and while not as common as they once were due to motherboard integrations are still pretty easy to get and come in a wide range of specifications. There is a wide variety of disk controller cards ranging from cheap and cheerfull IDE/SATA controllers (usually with fakeraid) to high end scsi and fiber channel controllers some of which have hardware raid. 56K modem PCI cards are extremely common (makers of desktop motherboards didn't want to bother with the beuracracy of certifying thier boards for connection to telephone lines, attempts to work arround this with special risers have been tried but not had much success). Some graphics cards are still PCI (and at one time in the pentium era almost all were). Then there are cards for most common interfaces (serial, paralell, USB, firewire), video capture cards, data aquisition cards and many many other more specialist cards. Plugwash 22:59, 26 September 2007 (UTC)

BDF/Bus enumeration

I've added some info pertaining to the bus/device/function concept and bus enumeration process to PCI Configuration Space, but I question if the information should remain there or be merged into this article. Does anyone have an opinion or preference? Rmcii 04:51, 12 May 2006 (UTC)

Versions compatibility ?

Do PCI v2.2 cards work in PCI v2.0 slots ? Vice versa ?

xerces8 --213.253.102.145 13:03, 29 May 2006 (UTC)

afaict the important thing is the voltage, if a card supports the correct voltage (as indicated by its keying) for a slot then it should work in that slot. I'm pretty sure though that putting a slower card in a faster slot will bring the whole bus down to the speed of the slowest card (this is one reason why high end motherboards tend to have multiple PCI busses) Plugwash 20:30, 2 February 2007 (UTC)

are 5V only slots and cards still the norm?

initially i thought that the 5V comment in the infobox was wrong, but having looked at a box with (no its not mine, it just happened to be being fixed at the time i dropped by) i noticed that the 32 bit section of the 64 bit slots was notched the opposite way to the way the 32 bit slots (and every normal PCI card i've seen) were notched. This has lead me to belive that normal 32 bit PCI slots are still 5V (despite what the latest versions of the PCI standard say). Can anyone clarify and should this be mentioned in the article. —The preceding unsigned comment was added by Plugwash (talk • contribs) 00:52, 13 January 2007 (UTC).

photo request

IMO what would be really good is a picture showing multiple slot types (ideally all 4 but even three would be nice) in context with the case and/or the whole motherboard (because 3.3V and 5V slots are the reverse of each other keying wise) Plugwash 02:12, 17 February 2007 (UTC)

Add-in card dimensions from PCI Express Spec

From PCI Express Electromechanical Specification, Rev 1.0a

Full size: 111.15 x 312.00*

Half-length: 111.15 x 167.65

Low profile: 68.90 x 167.65

*It is strongly recommended that standard height add-in cards be designed with a 241.30 mm maximum length.

From table 6-1 on page 67. Dimensions are height x length in mm. Height is from the top of the card to the bottom of the fingers. Dimensions do not include the bracket. —Ryan 05:00, 8 June 2007 (UTC)

PCI Hotplugging

If PCI doesn't support hotplugging (as per this article), why does the Linux kernel include support for it? --CCFreak2K 00:57, 4 July 2007 (UTC)

The options is there in the spec though rarely implemented (i think some high end server hardware has it) and a very common derivitive of PCI (cardbus) also has hotplugging support. Plugwash 22:48, 28 July 2007 (UTC)

Edge v. level-triggered

The article says, "level-triggered ... was chosen over edge-triggering... However, this efficiency gain comes at the cost of flexibility, and one interrupting device can block all other devices on the same interrupt line."

However, that's not correct. There is no loss of flexibility from using level-triggering. Using edge-triggering for a shared interrupt line is simply an unambiguous design blunder. There's no trade-off.

Also, the statement that with level triggering one interrupting device can block all other devices on the same interrupt line is untrue. That's the case for edge-triggering, not level-triggering. With level-triggering, if ANY device requests service, the CPU will be interrupted. With edge triggering, if one badly broken device asserts the interrupt line continually, it will block all other devices on the same interrupt line, and NO interrupts will be seen. (Either way, the system is badly broken, of course, but it is wrong to call this a "cost" for level-triggering.)

In the case of a device that is so thoroughly broken that the interrupt request line is simply "stuck" in the active state, the effect depends on the interrupt triggering mode: With level triggering the result is continual interrupts, and with edge triggering the result is no interrupts at all, but either way the result is a broken system.

If two or more devices share an interrupt line, then the bus must be using "wired-OR" logic (e.g., active-low, open-collector TTL). That's simply a result of the fact that the interrupt line is shared, regardless of whether it is edge or level-triggered. So if one device asserts the interrupt line continually, there will be no "edges," and if it is an edge-triggered design there will be no interrupts.

The enormous advantage of level-triggered interrupts is that (unless some device is very severely malfunctioning) one device cannot interfere with another. Nothing that one properly working device does can cause another device's interrupt to be lost.

With edge-triggered interrupts, race conditions WILL cause interrupts to be missed, even if everything is functioning properly. If a second device should request service before the first device's interrupt pulse has ended, there will be only one edge, and only one interrupt will occur. The second device's interrupt is simply missed. So even after determining which device needs service, subsequent to servicing that request the OS must check for ALL other devices which could have requested service at (approximately) the same time, because their interrupt(s) might have been missed. This adds software complexity, and incurs considerable additional processing overhead, and lengthens interrupt processing latency.

But with level-triggered interrupts, interrupts can never be missed. After servicing a device, the ISR can simply exit, confident that if another device is also already requesting service, or is about to do so, then another interrupt will promptly occur.

There is no trade-off. There is no advantage whatsoever to using edge-triggered interrupts. Level-triggering is the ONLY rational design for a shared interrupt line. That's why edge triggering isn't used for shared interrupt lines. NCdave (talk) 11:09, 22 December 2007 (UTC)

I agree completely with this reasoning. There is no "cost" for using level triggering and no advantage to use edge triggered interrupts whatsoever. The latter could have emerged only in some Microsoft-uneducated kind of mind - which I think was behind the original edge triggered mode used with ISA. Vleo (talk) 05:21, 27 July 2008 (UTC)

Why are PCI slots 'backwards'

In 24 years of working with computers, mostly PCs and clones, I've often wondered why PCI, MCA, AGP and other slots are 'backwards' when compared to the ISA slot? Ie, when you compare an ISA board to the others, the components are on the opposite side of the board. The 'tower' PC case style predates the introduction of PCI and the later bus slots, so it was logical to tip the case up on its left end- with the motherboard on the right side. That would let convection easily remove heat from ISA peripheral board components. But with any non-ISA cards the components are on the bottom side in a tower case, trapping heat under the cards where it can't be quickly removed by natural convection.

In all these years I've seen exactly ONE sanely designed tower PC case for a board having no ISA slots. It had the motherboard on the left side so the components on the PCI and AGP cards were on the top side, the drives were close to the bottom of the case- and thus close to their interface connectors, and the power supply was located at the bottom where its cable could connect via as short and direct a path as possible without the thick bundle of wires being in the way of anything, since the ATX power connector was where it should be, right up close to the onboard port connectors. Some people commented that it 'looked weird' with the CD-ROM drive down so low, but I liked its logical design.

I think the reason was to allow for "shared slots" where a PCI and ISA connector would be placed hard up against each other allowing the user to choose either one for the plate position but i don't know for sure.. Plugwash 01:43, 1 January 2007 (UTC)

The BTX form factor flips the PCI cards back around again by reorienting everything in its mirror image. I believe this is still used in many of the Dell's, including their current (as of this date) Precision line of workstations. --InsufficientData (talk) 12:56, 14 October 2008 (UTC)

What is the expected life time of PCI?

The computer manufacture industry is forever moving from one standard to another e.g. ISA -> VESA -> PCI, what plans are there for the next generation? If so what sort of time-frame is planned?

It depends what you look at. For graphics cards PCI was replaced by AGP which in turn has been replaced by PCI express x16. For high end server expansion PCI was replaced by PCI-X which is being gradually replaced by PCI express x4 and x8. For normal PC expansion I belive the intention is that PCI will be replaced by PCI express x1 but the transision is extremely slow (and if you look carefully you find a lot of PCI express x1 expansion cards are just PCI ones with a bridge chip added). Plugwash (talk) 19:20, 23 October 2008 (UTC)

PCI bus speed

According to the article the PCI bus speed is: (33.3 MHz*32 bits)=133 MB/sec. Actually 33.3 MHz*32 bits yield ~127 MB/sec. —Preceding unsigned comment added by Arnob1 (talk • contribs) 01:44, 31 January 2008 (UTC)

IIRC for bus speeds normal practice is to use mega to mean 1000*1000 not 1024*1024. Plugwash (talk) 18:40, 26 August 2008 (UTC)

I/O port space

Why would the I/O port space be deprecated? There is no fundamental advantage or disadvantage over memory mapped I/O. In the absence of any specific information that confirms this status I have removed that comment. CrispMuncher (talk) 10:48, 10 July 2008 (UTC)

There's consistent pressure to move as much functionality as possible to MMIO space, which permits more concurrency (posted writes, write combining, and the like). See the AHCI SATA controller spec, or the MSI (message signaled interrupt) standards for examples. 71.41.210.146 (talk) 23:09, 5 October 2008 (UTC)

Section on signals and their functions list and outline of PCI transfer protocol

I can add a section on how particular transfers are happening on PCI bus - that is how signals like CLK, FRAME#, IRDY#, TRDY#, DEVSEL#, #STOP interact, as well as a table of signals and their functions as I understand them, after coding few working PCI 32 controllers with HDL. Vleo (talk) 05:47, 27 July 2008 (UTC)

I was thinking of doing the same thing. "PCI bus protocol" seems like a good section heading. 71.41.210.146 (talk) 23:10, 5 October 2008 (UTC)

Mini PCI cards have a 2 W

"Mini PCI cards have a 2 W maximum power consumption, which also limits the functionality that can be implemented in this form factor." In reality many vendors implement mini-pci modules which much higher power consumption, esp. radios. Also, some vendors apparently use sideband signaling for USB and digital I/O.. See 84.228.178.252 (talk) 08:33, 5 April 2009 (UTC)

5V PCI cards

Does anyone have a picture of a 5V PCI card? These must be quite rare. I assume they won't work in modern boards? Drutt (talk) 06:57, 5 June 2009 (UTC)

I have this back-to-front: it's 3.3V cards that seem to be rare (these have the notch on the side nearest the backplate). This standard is also newer than the 5V one. Drutt (talk) 07:05, 5 June 2009 (UTC)

Both 5V and universal (dual notch) cards are common, I don't think i've ever seen a 3.3V only card, it would be an niche market if it did exist since it wouldn't be compatibile with the vast majority of machines. I've only seen a 3.3V 32 bit slot once (on a mips devboard), 3.3V 64 bit slots are quite common because of the higher speeds possible (66 mhz PCI and higher PCI-X). Plugwash (talk) 00:52, 7 June 2009 (UTC)

I have a board here with two 3.3V PCI-X slots, two 5V PCI-X slots and two 5V PCI slots. However, the two 3.3V slots have removable notches which can be moved to the 5V position to enable the slot to take 5V cards as well. Strange arrangement. Danceswithzerglings (talk)

very strange since unless those removable notches are some clever electrical device there is afaict no way for a motherboard to detect the voltage of a card. Of course the manufacturer may be just assuming that even cards notched for 5V can in fact cope with 3.3V. Plugwash (talk) 11:15, 1 October 2009 (UTC)

Power ratings ?

Can someone please provide the power ratings for PCI slots ? Thanks ! --Xerces8 (talk) 07:18, 24 April 2009 (UTC)

From the PCI specification 3.0, section 4.3.4.1 (Power Requirements):

No limit on 3.3 or 5 V rails except for 25 W per card total. (That's 7.6A on 3.3V or 5A on 5V.)
At most 0.5 A from the +12 V rail
At most 0.1 A from the −12 V rail

Systems may impose lower limits on the 3.3V and 5V rails. It is recommended that cards use at most 10 W at boot time, and enable higher-power modes only when the device is in use. See also section 4.4.2 (Power Supply).

71.41.210.146 (talk) 09:30, 16 July 2009 (UTC)

Compatible ?

I see in the article that "PCI-X adapters and slots are backward-compatible with 32-bit PCI slots and adapters". Does this mean all PCI-X adapters will work in a PCI slot (at reduced speed obviously on the 32 bit bus)?

If the card supports the boards voltage (most will) and there isn't anything blocking the overhanging connector then it should work fine. 130.88.108.187 (talk) 13:13, 9 October 2009 (UTC)

Comments on PCI protocol sections recently added to article

This is a discussion area for beating the above sections into shape for inclusion in the article. NOTE that as concerns here are addressed, the text above will change and so the comments might be out of date. (The inclusion has since happened; the sections have been moved into the main article.)

Comments by User:JMiall (talk)

This was a solicited review of the section.
- Interleaved responses by User:71.41.210.146 (talk)

(please post these to the talkpage there if you wish):

You asked first of all if it read well to someone unfamiliar with the specs. I think that basically it does.
I'm a bit worried though that there is too much detail, I'm not sure where is a good place to draw the line though. Possibly all this information should be split off to (an)other article(s) and summarised within this one?
- Perhaps. Creating new articles is a pain for me, so maybe I'll leave that to someone else.
I've not looked at everything you have put on the talkpage yet, I'll come back with some more comments if you found these useful.

Specific questions/issues to be going on with:

Why is "asserted" in quotes?
- That was for emphasis, because that was where asserted was defined. Should be italics, thanks.
'controlled by 5 main control signals' etc - is signal the right word here? ie are FRAME# etc the name of a particular line or the signals that are present on that line?
- Yes, they are physical wires. You can see them in Conventional PCI#Connector pinout. (E.g. FRAME# is 34A.) Er... what is the distinction between "line" and "signal" you're making here? Generally, I think of them as synonymous, except that "signal" emphasizes information-carrying capacity. "+5V supply line", but "clock signal". I suppose that "signal" is more abstract, and e.g. two signals could be multiplexed onto the same wire...
Also the when the lines/signals are mentioned they sometimes are put in brackets, sometimes commas and sometimes nothing.
- I don't understand this one. [brackets]? Or (parentheses)? Can you give me an example?
'will remain high' - add 'voltage', also voltage needs linking in the main article
- H'm... probably.
'each bus line be undriven' - given the potential for confusion for someone not too familiar with English is it possible to reword?
- Okay. I tried to clarify.
Link Pull-up resistor
- Good point. Sorry.
'Signals nominally change on the falling edge of the clock' - should this be 'nominally' or 'normally'? Is the spec not specific about precisely when they should change?
- Nominally, and no the spec is not. To be precise, they must change within a specific time window relative to the leading edge of the clock, so the signal has time to arrive at the receiver. This window is close to the falling edge of the clock, and I felt that the details of setup and hold times and clock jitter was definitely something that didn't below in Wikipedia; see the spec if you care about that level of detail.
The wording rather implies at that all the PCI devices are concious intelligent beings. It is probably worth toning this down a bit. ie 'decides', 'promise', 'generally attempt', 'desire'
- H'm... this is because the actual function of the PCI device is not relevant here and the reason why it "wants to talk" to the PCI bus is not discussed. The language sort of follows naturally from that. I can try a bit, e.g. promise can be changed to "guarantee," but does anything particularly grate?
'(There are rules for how it is supposed to behave, but they will be mysterious to any single device.)' - is this necessary?
- In the middle of emphasizing that the GNT# signal might go away at any time for no obvious reason, it occurred to me that that isn't the same as no reason at all.
link 'bus master'?
'but may not start one unless it observes GNT# asserted the cycle before it begins a transaction' - unclear 'it's
'so there are actually three signals' - fairly colloqial language
'Actually, it has 2.5 cycles' - 'they' not 'it'?
- Perhaps "each target".
'On the fifth cycle...' - I'm unsure what this paragraph means.
- Okay, I'll work on it. See the figure below with numbered
'Access to PCI configuration space is a special case' - of what?
- Of address decoding; many address bits are ignored, and IDSEL matters. Thanks, I'll tweak it.
'enables' - link or explain?
- Don't the following sentences do it adequately?
'In case of a write' - missing 'the'
'more purely'?
- Arrgh.
Several times you use 'read' or 'write' when it would probably be clearer to use 'read signal' or something similar
- I can't find an example. Can you point one out?
'line turns around during' - what does turns around mean here?
- This is the undriven cycle discussed above. I expanded that section, because it becomes important later.

Main article:

'However, they are not wired in parallel as are the other traces' - this needs explaining. Traces have not been mentioned yet and it is unclear what this means any way due to the 'are not...as are...'
- Thanks, will fix.

JMiall ₰ 23:16, 2 August 2009 (UTC)

71.41.210.146 (talk) 01:44, 3 August 2009 (UTC)

Conventional PCI vs. PCI Local Bus

The lead says "Conventional PCI (part of the PCI Local Bus standard...)". My questions are:

Is Conventional PCI an industry term or just something made up by Wikipedians to distinguish the original PCI from newer versions (PC-X, PCIe).
If Conventional PCI is part of the PCI Local Bus standard, what else is in that standard? --Kvng (talk) 21:47, 1 September 2010 (UTC)

PCI-SIG uses the terminology Conventional PCI (e.g. ). I assume PCI Express would be another part of the PCI Local Bus standard. —Ruud 13:04, 15 February 2011 (UTC)

New section: PCI bus latency

{{editsemiprotected}}

It's included verbatim below; sorry if this looks a bit messy on the talk page, but it seems easiest to cut & paste this way. It should go between "PCI bus transactions" and "PCI bus bridges" (the latter of which needs work). Thank you! 71.41.210.146 (talk) 21:33, 20 March 2011 (UTC)

More information Extended content ...

Extended content

PCI bus latency

Soon after promulgation of the PCI specification, it was discovered that lengthy transactions by some devices, due to slow acknowledgments, long data bursts, or some combination, could cause buffer underrun or overrun in other devices. Recommendations on the timing of individual phases in Revision 2.0 were made mandatory in revision 2.1:^[1]

A target must be able to complete the initial data phase (assert TRDY# and/or STOP#) within 16 cycles of the start of a transaction.
An initiator must complete each data phase (assert IRDY#) within 8 cycles.

Additionally, as of revision 2,1, all initiators capable of bursting more than 2 data phases must implement a programmable latency timer. The timer starts counting clock cycles when a transaction starts (initiator asserts FRAME#). If the timer has expired and the arbiter has removed GNT#, then the initiator must terminate the transaction at the next legal opportunity. This is usually the next data phase, but Memory Write and Invalidate transactions must continue to the end of the cache line.

References

[1]
PCI Local Bus Specification: Revision 2.1 vs. Revision 2.0 (PDF), Application Note AP-753, Intel Corporation, 1997 {{citation}}: Unknown parameter |month= ignored (help)

Delayed transactions

Devices unable to meet those timing restrictions must use a combination of posted writes (for memory writes) and delayed transactions (for other writes and all reads). In a delayed transaction, the target records the transaction (including the write data) internally and aborts (asserts STOP# rather than TRDY#) the first data phase. The initiator must retry the exact same transaction later. In the interim, the target internally performs the transaction, and waits for the retried transaction. When the retried transaction is seen, the buffered result is delivered.

A device may be the target of other transactions while completing one delayed transaction; it must remember the transaction type, address, byte selects and (if a write) data value, and only complete the correct transaction.

If the target has a limit on the number of delayed transactions that it can record internally (simple targets may impose a limit of 1), it will force those transactions to retry without recording them. They will be dealt with when the current delayed transaction is completed. If two initiators attempt the same transaction, a delayed transaction begun by one may have its result delivered to the other; this is harmless.

A target abandons a delayed transaction when a retry succeeds in delivering the buffered result, the bus is reset, or when 2¹⁵=32768 clock cycles (approximately 1 ms) elapse without seeing a retry. The latter should never happen in normal operation, but it prevents a deadlock of the whole bus if one initiator is reset or malfunctions.

Close

Done Stickee (talk) 11:10, 21 March 2011 (UTC)

Suggestion to make article more practical

I'm posting this suggestion with the hope that someone skilled in wikipedia editting might agree and add this to the article. I attempted to install a PCI to USB adapter card (a common piece of computer hardware) to an old computer motherboard in order to increase the number of USB slots (a very common thing to do). I assumed PCI cards are "one size fits all" but found out that the card that I was attempting to install did not fit. I then discovered that there are two standards of PCI (3.3 V and 5 V). One is new, the other old, and each is physically "keyed" so that you cannot insert a 3.3V card into a 5 V slot, and vice versa, unless the card is keyed for both. The solution for me was to purchase a card that was keyed for my older motherboard.

This article is what gave me this knowledge, but it was difficult to figure out. Yet, I am certain my situation is very common, and may be one of the primary reasons why a person would be seeking to read about the technical characteristics of PCI.

To accomodate these imagined "other people", I think this article could be improved by creating a special section describing the situation I experienced, with the intent to quickly helping people similarly situated understand their problem and find their solution.

It's a slightly delayed response, but I thought that the section "Conventional hardware specifications" covered physical hardware differences adequately. Heck it was useful for me last year when I was considering buying a pci-x card. Mike-s (talk)

Jonny Quick (talk)Jonny Quick —Preceding undated comment added 15:04, 22 May 2011 (UTC).

Unfortunately, Wikipedia is not an instruction manual. --Redrose64 (talk) 14:06, 23 May 2011 (UTC)

That instruction manual crack is unjustly snide. The gentlesapient isn't suggesting a step by step do this, do that guide. What cards fit what slots is appropriate and is discussed in this article as it stands at the moment I write this but not clearly. A summary with 4 simple sentences like the following would be totally appropriate: "32 bit (short), 3.3 volt PCI slots can hold 32 bit 3.3 volt PCI cards or 32 bit universal PCI cards only. 32 bit 5 volt slots can hold 32 bit 5 volt PCI cards or 32 bit universal PCI cards only. 64 bit (long) 3.3 volt PCI slots can hold 64 bit 3.3 volt PCI cards or 64 bit universal PCI cards only. 64 bit (long) 5 volt PCI slots can hold 64 bit 5 volt PCI cards or 64 bit universal PCI cards only." Alternately you could do six sentences, one for each kind of PCI card along the lines of "A 32 bit (short) 3.3 volt PCI card must go in a 32 bit (short) 3.3 volt PCI slot." Or a table might be clearer. I don't add it because 1-I'm not absolutely sure of the correctness of all 4 statements and 2-You silly farstards still have it locked. — Preceding unsigned comment added by 24.96.210.135 (talk) 04:55, 6 February 2012 (UTC)

Office Action

OFFICE action

The following were removed as a result of a DMCA takedown notice, per the office action policy. Please do not reintroduce these. The article remains open for editing, with the exception of any links to these specifications. If you'd like to appeal, please email your notice to me at philippewikimedia.org.

PCI Local Bus Specification, Revision 2.1

PCI Local Bus Specification, Revision 2.2

PCI Local Bus Specification, Revision 3.0

Office actions may not be reversed except through the Wikimedia Foundation office. The pertinent policy lives at WP:OFFICE. Philippe Beaudette, Wikimedia Foundation (talk) 01:42, 4 January 2011 (UTC)

This is quite pointless because you can still see it in the article history. --Hinata talk 15:46, 21 January 2011 (UTC)

Seconding the previous' posters finding. Also, are you <censored by myself>? The original Document is still online, publicly available and a top hit on google. I bet this DMCA Notice was automatically generated. Furthermore, completely removing the reference is just stupid - wouldn't it suffice to just remove the link? Effin' lawyers.. --Roeme (talk) 11:12, 23 January 2011 (UTC)

If it's not a valid DMCA compliant wouldn't it be possible to file a DMCA counter notice? --nn123645 (talk) 01:30, 2 February 2011 (UTC)

Since PCIe is quickly replacing Conventional PCI, and PCI-SIG should release the Conventional PCI spec for a reasonable price to non-members. • Sbmeirow • Talk • 15:15, 23 January 2011 (UTC)

For those who are interested, I've posted the DMCA take-down notice at this location on the Foundation's wiki. It specifically mentions the reference section and requires that we remove the specifications. At this point, until or unless we receive a DMCA counter-notice, those standards can not be reposted. If we were to allow that to happen, we threaten the immunity under which we operate through the Safe Harbor provisions of the Digital Millenium Copyright Act. I'd be happy to provide someone with instructions regarding where to send a counter-notice if someone has a valid couter-claim, but be aware that it would be a statement under penalty of perjury that you have a valid copyright counter-claim. Philippe Beaudette, Wikimedia Foundation (talk) 01:24, 24 January 2011 (UTC)

Don't you see that the one writing this notice didn't understand that it's not wikipedia that's hosting that PDF?! From my understanding of this notice and the DMCA in general, it would be completely okay to CITE the specification - without linking to it. Someone readded the reference in the mean time (even with a link to the pci sig), so further discussion is moot. Next time, I'd recommend just to break/remove the link. Removing the ref alltogether in my opionion is just excessive and feels like censorship. --Roeme (talk) 10:07, 24 January 2011 (UTC)

I will execute a DMCA take-down exactly as instructed by legal. Anything else compromises our position. Your understanding may or may not be correct, but I will not follow your interpretation - I am duty bound to follow the interpretation of our legal team. Philippe Beaudette, Wikimedia Foundation (talk) 02:08, 25 January 2011 (UTC)

Roeme, see Wikipedia:OFFICE#DMCA_compliance, particularly "the Foundation is required to comply with validly formulated notices even if they are spurious". I'm not saying this one is spurious. The point is that the Foundation complies to keep its safe-harbor status, not because it has independently assessed the merits of the claim (although they may do that anyway).--Chaser (talk) 01:57, 1 February 2011 (UTC)

Well this kinda depends. It can be argued that linking to copyrighted material in this way is not a copyright violation, and thus usage of DMCA for this purpose (getting a website to remove a link) is abuse of the DMCA system. The question of linking to copyrighted material is FAR from settled. While what the foundation is doing, is acting on the safe side, it is (as with the caving in to another user who didn't understand copyright [alexander liptek]) a dangerious precedent that we should keep a close eye on. I don't want to arrive at a place where the foundation always acts on the safe side, and leaving this battles of DMCA abuse entirely to its editors. 93.111.234.204 (talk) 16:51, 1 February 2011 (UTC) (which would be User:TheDJ) 93.111.234.204 (talk) 16:52, 1 February 2011 (UTC)

I also think we should revisit wether we want these DMCA complaints marked in articles. A nice red meta template with "Specifications removed after DMCA complaint of copyright claimant" might be a useful addition to raise awareness -- TheDJ —Preceding unsigned comment added by 93.111.234.204 (talk) 16:55, 1 February 2011 (UTC)

Citations

Someone please file a DMCA counter notice with the Foundation. We need to be permitted to footnote the spec, without hotlinks. --Hroðulf (or Hrothulf) (Talk) 14:39, 3 February 2011 (UTC)

We can cite the spec in the same way that we cite a print source or a paywalled online source. --Damian Yerrick (talk | stalk) 15:32, 13 February 2011 (UTC)

"The article remains open for editing, with the exception of any links to these specifications." (emphasis mine) I interpret this as stating we can still reference the specs, as long as we don't add any hyperlinks to pirated copies. —Ruud 12:38, 15 February 2011 (UTC)

What makes no sense to me is that the notice cites only the standards, not hyperlinks thereto. The notice is "seeking the immediate removal of the Local Bus Specifications from the Infringing Entry" (emphasis added). The Specifications are defined quite precisely by the copyright registrations, and have never been part of any wikipedia entry. The notice does not mention links in any way, shape or form. Anyway, I'll replace them with Google "I'm Feeling Lucky" links. 71.41.210.146 (talk) 01:11, 7 March 2011 (UTC)

Please don't. Hyperlinking to infringing copies of the specs, whether directly or indirectly through Google's "I'm feeling lucky" link, is still assistance in copyright infringement. The DMCA notice did not explicitly mention the URL exactly because it wanted to be wide enough to cover your "trick" as well. —Ruud 16:28, 20 March 2011 (UTC)

Er, AFAICT, the original DMCA notice was not even wide enough to cover linking to the spec; I asked why Wikimedia legal made that extrapolation. (In my ignorance, I'm assuming it's, "they're pissed about something, so we'll assume they object to the status quo", but it also looked like a mass-mailed form letter, so maybe they didn't do careful review.) See the exact words I quoted (and I don't believe I took them out of context) from the notice above. They asked for removal of the specifications from the entry, and did not mention or even allude to pointers to the specifications or instructions how to find the specifications. (If they did, I'd wonder if we could not link to the PCI SIG's version!)

If Philippe or WMF legal thinks otherwise, they can say so. 71.41.210.146 (talk) 20:11, 20 March 2011 (UTC)

P.S. I have asked User talk:Philippe (WMF) for his input to arbitrate the point. —Preceding unsigned comment added by 71.41.210.146 (talk) 20:22, 20 March 2011 (UTC)

No, there is absolutely no question is this case that we're not allowed to link to unauthorized copies of documents copyrighted by the PCI-SIG. We shouldn't have been doing this in the first place, we certainly shouldn't be doing this after they sent an, in this case entirely reasonable, DMCA takedown notice. Now I don't know which part of Philippe's: "The article remains open for editing, with the exception of any links to these specifications" and "Under no circumstance are editors to remove this protection or edit in an attempt to circumvent the letter or spirit of it. Any attempts to provide such information will result in the reversion and suppression of the edits made, and the user in question may be blocked for an indeterminate length of time." you didn't understood the first time, but for your own good I'd suggest you try to read and understand them once more. —Ruud 21:06, 20 March 2011 (UTC)

So I included a link to a link to the document. ("All problems in computer science can be solved by another level of indirection.") I assume that Philippe is patrolling the page (there are exactly three office actions in progress, hardly a chore to keep track of), so I didn't notify him personally, but I announced it both in the edit summary and on the talk page here. Then when I disputed your opinion, I did notify him personally on his talk page. Because I might be wrong. I think this is reasonable evidence of good faith. (I am not trying to circumvent the spirit of the office action, but I am most assiduously trying to circumvent the spirit of Schwabe, Williamson, and Wyatt's "Take Down Notice Pursuant to 17 U.S.C. § 512(c)"; those who live by the letter may die thereby.) Have you actually read the DMCA notice itself? It lists specific documents. It does not list the PCI-to-PCI Bridge Architecture Specification, revision 1.1, so of course I happily included a link to that document. (I also included a link to the Revision 2.3 specification, which is not listed as part of the restrictions above, but was listed in the DMCA notice; this is perhaps questionable.) But the point is that it requested removal of the specifications. They're gone. They never were there. Despite the laywer's "reasonable belief that the Local Bus Specifications are being distributed and displayed as part of the Infringing Entry" (emphasis added), the entry never has included them.

There is mention in the DMCA about links. That's 17 U.S.C. § 512(d), not (c). It's definitely a hot topic of legislation (the DMCA does not establish liability for linking, only provides a safe harbor for whatever liability might exist), but in any case the letter, by specifying only (c), excludes (d). I think WMF legal's interpretation is already incorrectly overbroad, but I'm not going to argue that. I am going to go up to the line they've drawn. 71.41.210.146 (talk) 00:24, 21 March 2011 (UTC)

Now what ever flawed legal interpretation of this situation you happen to have is completely irrelevant. The WMF legal team has decided we should not link to any copy of these specifications. This obviously includes any other material infringing on PCI-SIG's, or anyone else's for that matter, copyright, not mentioned explicitly in the notice. Do not add such material to the article, this talk page, or anywhere else. I'm not going to warn you again. —Ruud 08:20, 21 March 2011 (UTC)

*Sigh*. I have not now, nor ever included any copyright-infringing material in a wikipedia article. I don't intend to.

Whether telling people where to find (i.e. linking to) copyright-infringing material is a form of copyright infringement is a hotly debated and unsettled point of law. Your statement "This obviously includes any other material infringing on PCI-SIG's, or anyone else's for that matter, copyright, not mentioned explicitly in the notice." is, IMHO, completely and utterly unwarranted. The notice, and subsequent office action, was narrow and specific. There's no need to generalize to suppressing all material that might annoy someone. If wikipedia removed everything that someone somewhere wished it didn't include, it would turn into Newegg product reviews.

Find me one word, in the lawyer letter, or any office action notice, or anywhere else on wikipedia, that suggests that anyone except you has any problem with including a link to the PCI bridge architecture specification. The Office Action notice is at the top of this section, if you want to re-read it. (Also, find me one word that suggests that adding the link anywhere except in the Conventional PCI article is problematic.) As I said, my interpretation disagrees with WMF legal's, but I'm not challenging that. I'm disagreeing with your interpretation, and I am willing to challenge that. Unless you have some authority to speak for WMF legal that I'm got aware of.

I've explicitly asked Philippe for a ruling. Can we just wait until we get one? 71.41.210.146 (talk) 14:59, 21 March 2011 (UTC)

Let me quote Philippe (again):

The article remains open for editing, with the exception of any links to these specifications.

(Emphasis mine). That shouldn't be too hard to comprehend, right? If you link to material infringing on someone's copyright one more time, you will be blocked. This is Wikipedia, not the Pirate Bay. —Ruud 15:36, 21 March 2011 (UTC)

And I read that, and obeyed that. To the letter. I included links to other specifications (not the listed ones), and links to google searches for the listed specifications. As I said, up to the line. I was very careful not to cross it. If WMF wants to tell me the line has moved, okay. But I don't see why your personal interpretation deserves more weight than mine. 71.41.210.146 (talk) 16:19, 21 March 2011 (UTC)

I'm lazy to read the whole case and it probably isn't necessary anyway but is there actually any doubt the link above is copyright infringing? If so, I don't think any of us give a damn whether we were specifically asked to take it down. Our policy is clear and has been since long before this case, we don't link to copyright violations. The fact that this time it required a DMCA notice before we removed it, after they already asked us nicely is rather scary but if my understanding of what's going on is correct, it doesn't change the fact we are majorly in the wrong here and should have removed it a long time ago before anyone even asked nicely. If it's claimed the work is not copyright infringing, can someone briefly explain why? Is it that the work is not copyrighted or that the place we are linking to has permission from the copyright holder to host it? Or is it something complex like the place hosting it is located in a country where it's not copyrighted but it is in the US? Nil Einne (talk) 18:03, 21 March 2011 (UTC)

Nil, thank you. I had never read that section. Or if I had, I forgot it. I'll have to study it in more detail. In particular, I'll have to read the discussion archives for limits on the threshold of "knowing". I have many times installed links to freely available versions of academic papers (e.g. on a preprint server or the author's own web site, as opposed to behind a journal paywall) or newspaper articles, in simple ignorance of the copyright status of the copies I'm linking to. I'll do a bit of reading. (14 pages of talk archives, here I come...) 71.41.210.146 (talk) 18:52, 21 March 2011 (UTC)

I think linking to an authors website for journal articles would generally be fine because journals often do allow authors to host the article themselves subject to some conditions so it's reasonable for us to presume these aren't copyvios. To be fair, I wouldn't say many authors actually know what they're doing (except perhaps for legal ones) so some of them are probably copyvios but particularly if they're on a university website there are people running them who do care about copyright and the journals themselves know that so it would usually still be fair for us to just presume there's no copyvios unless there is actually evidence something is a copyvio. In cases when it's not reasonable to presume the content we're linking to has the copyright holders permission (i.e. there's a reasonable chance it is a copyvio) we don't usually allow it. Youtube is the classic example when the content is apparently not uploaded by the copyright holder (nowadays the copyright issues is particularly for music is not always as clear cut but we usually still don't allow it) but it applies to basically anything. While there are potential legal issues relating to contributory infringement, I don't believe this is really the primary reason for our policy, it's more of wanting to respect copyright. Nil Einne (talk) 16:04, 31 March 2011 (UTC)

Maybe references can be simple "<ref>PCI Local Bus Specification revision 2.3 2002-03-29</ref>" instead of direct link with google?, any person can lookup these things anyway. The important point must be to point to what specification and what section that is referenced? Electron9 (talk) 04:20, 20 May 2011 (UTC)

This whole thing sounds way overblown. The standards holder wants to keep access to its standards behind a paywall. Fine, so cite the standards holder with a link to its paywall. Now, was that so hard? DMahalko (talk) 02:44, 8 October 2011 (UTC)

With Professional Communicators on three levels (PCI-SIG attorneys, Wikimedia Foundation attorneys or Legal team, and Wikimedia Foundation's Head of Reader Relations), the PAID work they have done is embarrassing. Incomplete and unclear statements, followed by clarifications that don't. It is still unclarified, but us dumb editors SEEM to have figured it out what's really the issue -- it is kind of rude to link to bootleg copies of someone's $5000 spec, even if those bootleg copies are hosted on university web sites. Currently the article cites those specs (without linking to them anywhere), and Legal hasn't come to smash it, so that must be okay. However, if we are afraid to deep link to PCI-SIG's paywall page, then we are lying too low. Deep linking is used everywhere on Wikipedia and the web, so we must have a de-facto policy of allowing deep linking, whether the target sites like it or are stupid enough to dislike it. A876 (talk) 21:05, 20 December 2011 (UTC)

They are lawyers, they like to keep everything as vague as possible. The PCI-SIG to make there takedown notice as broad as possible, the WMF to avoid losing their DMCA safe harbor status. To anyone with a bit of common-sense it should have been pretty obvious that what both parties where trying to make clear to us was that referencing the spec and (deep)linking to the PCI-SIG's website is okay, but linking to pirated copies is not. The article has been stable for months, so I'm going to prod Phillipe to see if the banner/protection can be removed from the article now. —Ruud 20:54, 21 December 2011 (UTC)

I think that at least mentioning that it exists on the internet, somewhere is not a link. If it were, we should also be rev-deleting all comments on this talk page saying anything along the lines of "it is possible to link to it," since saying that a link exists means that it is on the internet. Between "speech" and "linking," I think this falls on the "speech" side of the line.--New questions? 09:09, 10 April 2012 (UTC)

Completely unencyclopaedic information. Please, don't reinsert. —Ruud 22:00, 10 April 2012 (UTC)

Why would you say it is "completely unencylopedic information"? I would say it is completely encyclopedic.--New questions? 00:12, 11 April 2012 (UTC)

In particular, I think that saying something along the lines of "there exists a genuine PDF of it, but it is not linked to avoid infringing copyright" is appropriate for an encyclopedia since it provides integrity to the connection to a possible source (i. e. if someone comes across the PDF, they would know that it is the genuine article), which is at least something that could be done even if it cannot be linked to.--New questions? 00:26, 11 April 2012 (UTC)

I don't particularly feel like arguing with you (it gets a bit tiresome to have the same discussion every other month, just read the rest of the talk page and use some common sense). If you insert such material into the article again I'll block you indefinitely and without further warning. —Ruud 00:42, 11 April 2012 (UTC)

I don't really like that threat, and I have read the talk page and I am using common sense. I would like an explanation as to why it would be problematic to insert such material that is not links, nor even aiding anyone on how to find it.--New questions? 01:19, 11 April 2012 (UTC)

In particular, I believe you are making an inappropriate threat to block. I do not believe adding a simple statement, "link removed due to copyright complaint," does any more than the very visible notice on the very top of the page.--New questions? 01:36, 11 April 2012 (UTC)

It's possible to link a official pdf document that explain differences between PCI spec 2.0 and 2.1? Is hosted on intel.com and intel is founder member of PCI-SIG... or the "smart guys" will try to sue also Intel? :) 88.149.209.147 (talk) 16:27, 11 June 2012 (UTC)

That's fine. In fact that document is already linked to from the article (Reference #12). —Ruud 18:41, 11 June 2012 (UTC)

Alternative WP:OFFICE template

Hey Philippe. Posting here as I'm guessing you're keeping an eye on this page. I realize that the language of the current {{pp-office}} template is fairly vague about time period for protection. Still, in my experience, pages that have WP:OFFICE restrictions are often temporary (in the several months sense of temporary). For an article such as this, where it is essentially under indefinite protection (until someone properly files a counter-claim), I'm wondering if a template that indicates the indefinite period of the Office Action protection might be appropriate.

The simple reason for a separate template would be to alert users to the fact that they cannot simply "check back later" and expect the office action to be resolved. I doubt all users dive through the talk page. The extended reason for a different notice is that the DMCA notice, even if entirely valid, serves as on-going censorship of this page (in that it limits the content in a certain way). Just to be clear, assuming PCI-SIG's claim is valid, they could very well have the legal right to censor this article in the way they have. Still, since WP is not normally censored, I'm suggesting bringing to more prominence on the page itself that this page has been censored.

I'm suggesting nothing more extensive than something similar to the DMCA notice Google drops at the bottom of your search results when they've censored the results. New template could read:

The Wikimedia Foundation Office has placed [[ref/link to specific restrictions|certain indefinite restrictions]] on this page as the result of a [[link to legal notice|legal request]] by party making request under the DMCA. For more details, please see Wikipedia:Office actions, [[page?|how you can respond to this DMCA notice]], or the article's talk page. Do not remove this template from the article while these restrictions are in place.

I could do up the template, but as this is an Office matter, wanted to check in first. Again, even if a revised notice doesn't look like the suggestion above, I really think that this page should have a clearer notice about the indefinite nature of the restrictions on this page, as it does make it a bit of an oddity on the project.

P.S. Just as I was finishing this, I thought, "Maybe the article page doesn't point to the notice because the notice was less than absolute?" (I'm referring to the exceptions to the take-down mentioned earlier on this page by other editors.) That is, since they haven't come in guns blazing, maybe it's best not to be pointing editors directly to the means by which they could "pester" PCI-SIG, for fear of things escalating. (I realize escalation of any matter would be a first and entirely unheard of for the project.) I'm not sure if this played into WMF's thinking at all, but if it did, I think that moving the DMCA response instructions from talk to article page wouldn't have much effect given it would be the editor making the claim and it really is just one level removed from where the instructions are already located. Now that I've taken too much of your time, I'm off. Please hit me with a {{tb}} or something similar on my talk page if you respond as I'm not really expecting one soon. Thanks for reading!--Policy Reformer(c) 20:39, 31 December 2011 (UTC)

This looks reasonable to me. I'll ask legal what they think. :) -Philippe (talk) 21:22, 31 December 2011 (UTC)

What about showing a small lock icon in the corner only and giving the full explanation only when editing the page using an edit notice? —Ruud 21:31, 31 December 2011 (UTC)

My thought was that this protection, even more than most office protections, should be highlighted rather than diminished. Here, an article is being indefinitely censored, and given that consensus typically beats censorship on the project, this instance where consensus cannot beat censorship (without going through the DMCA counterclaim procedures) should be highlighted both to alert the reader (99.5% of whom aren't editors) that they are looking at a DMCA-censored page and to serve the purpose that (I entirely agree) would be well served by an editnotice warning (i.e. alerting editors to situation and/or specific things they can't include).

Since this would be a change/addition to the pp-meta template, I'm sure there's some channels or proposal procedures this might need to go through anyway. Initially posting here to both get comments from WMF as well initial comments to see if this is a non-starter from the beginning.

If legal says, "Yes, do it like that right now period end of story," maybe public comments/consideration might be brief. If legal says, "Yeah, let's see what the community comes up with," then we toss it through the normal process, and probably ask for approval on whatever we come up with. Legal could also say, "That's absolutely foul language. What made you think we'd go with that? Although we do like the idea of a DMCA notice as opposed to other office actions. We'll give you some language." And finally, they could say "No, let sleeping dogs lie."

Before we go to any trouble (other than me writing this junk that you're kind enough to read), thought it best to check it to see where we fall in the spectrum. Thanks again for reading my novels. Happy New Year!!--Policy Reformer(c) 05:55, 1 January 2012 (UTC)

If the article really was censored, okay, I'd agree there definitely should be a big and clear disclaimer targeted at the readers. But in the current version of the article nothing has been removed for which there wouldn't have been a strong consensus to remove it anyway based on our own policies. Cheers, —Ruud 15:43, 1 January 2012 (UTC)

I agree that links to copies of the specifications that are in violation of copyright laws also contravene WP policy and undoubtedly consensus. It's more the ongoing limitation (the need for at least a padlock). In order to avoid violations of the safe-harbor provisions of the DMCA, links to a legitimate alternative/proper copyright holder (ACH) are also impermissible based on the same need to protect WMF under DMCA. Before posting links to the ACH site, there needs to be a DMCA counterclaim. An editor who stumbles upon this ACH site likely does not have standing to file a counter-claim. In this situation (and perhaps others), the editor would be "censored" in that he is prevented from adding material to the site on the basis of the notice. "Prevented from" might be better. I was trying censored in the sense that when one is prevented from expressing a statement or idea by a third party especially under the auspices of a legal or governmental threat, one is censored. For example, when a radio DJ uses language to dance around the seven dirty words, I'd argue she's censored, even though it's not a case where they "bleep" her.

A side effect of this DMCA notice is that PCI-SIG makes their site the only source listed on the page, and in some way may funnel business to themselves through that circumstance. Not saying their claim is improper or that's not how it should be. Still, had they made an account here and raised the copyright flag, we probably would've resolved it ourselves in favor of them anyway and then it would be the community upholding the protection rather than WMF under legal threat. Again, something just feels different about this one. Most Office Actions: "Don't use that one picture, or say that one thing, or include that one page." This action, and perhaps broad DMCA notices in general: "Remove now and don't ever at any time possible in the future post a link to anyone other than us or we'll bring the hammer." I feel this one's an oddity due to it's indefinite nature. Office Actions are already pink elephants (thankfully - Philippe & rest of WMF, thank you for the freedom) on the project. Just thought it's worth pointing out this is a pink three-legged elephant. I'm probably running all over WP:MOUNTAIN right now, but thanks for entertaining my diatribes.--Policy Reformer(c) 20:29, 1 January 2012 (UTC)

According to the takedown, they first requested removal through e-mail. No idea why the Foundation failed to respond to this in time. The ongoing protection of the article, also seems to be a choice of the Foundation (it's a technological measure the MediaWiki software happens to provide, but I doubt it's required to prevent losing the safe harbor status.)

You rightly point out that there will be some indefinite restrictions on this article, but they are mostly theoretical and only affect editors. We're still presenting readers the article we would have wanted to present to them if the PCI-SIG wouldn't have sent a takedown notice. —Ruud 21:04, 1 January 2012 (UTC)

Could be that they did respond to it and the response was insufficient, or that the response wasn't received. Thanks for pointing out the prior email as I didn't pickup on that part of the take down (the prior email). Based on that email, I'm wondering if there's more to this in OTRS than we know.

If this was just a matter of being as "polite" a follow-up as PCI-SIG could make based on a non-response or unreceived response to their initial email, and given that, as far as I know, they haven't tried to further affect the article, I agree that the indefinite restrictions are on editors rather than community created content. I do believe the restrictions on editors are real, but regardless, they're editor restrictions which could be served by an {{editnotice}} as you pointed out.

When I looked at this restriction, I assumed PCI-SIG was playing hardball, as I imagined that WMF would have lifted as soon as possible just like normal. Just knowing that PCI-SIG reached out first makes me rethink that premise. Moving forward, I'm wondering if what should happen here is the move of the WP:OFFICE notice to an {{editnotice}}. Saves all the readers that never hit edit from seeing the warning, since as Ruud pointed out they're not seeing different content anyway as of right now. If it is moved to an editnotice, I think it's just a matter of pointing out that these are WMF required restrictions instead of community based. Something like "WMF has placed the following indefinite restrictions on this article per a [[link to DMCA explanation and counter-claim instructions|DMCA takedown notice]] which you can find [[link to actual notice|here]]."

Normally office actions result in full protection of an article, but I don't think that's the case here anymore. If the article is left indefinitely semi'd though, maybe just included that in the editnotice as well. Could also maybe go with a small=yes protect (either office or semi) template, as you suggested, to avoid the issues of constantly revisiting to see if we could remove the large {{pp-meta}} template.

Finally, I still think there could be a circumstance where someone or some company could try to censor content through the liberal use of DMCA valid or invalid notices including threat a future action if the notice is not enforced on an ongoing basis. I'm guessing WMF doesn't want to do that enforcing nor spend time fighting them, and I'm sure we don't want to either. It looked like this was a case where that might be occurring on a small scale. It's probably not. Even in the face of legal threat, I shouldn't have forgotten WP:AGF. I also suppose it's a good sign that this issue hasn't come up before. As long as we continue to make clear the restrictions, and that the restrictions are based on something outside WMF and the project, doesn't matter where it is as long as it works, right?

Philippe - Thanks again for looking into this. Whatever legal wants to do goes, of course. I still think we should figure out a way to change this to differentiate it from the temporary actions. However, you can see that I'm questioning my approach of declaring the malevolence of all those bearing DMCA notices through a new notice. For me right now, I think it just gets down to effectuating the requirements by WMF without having this as a "Current Office Action" from now until the end of time. (you guys have the servers leased through then, right?)

Ruud - Thanks for making me rethink this. Take care.--Policy Reformer(c) 01:37, 4 January 2012 (UTC)

Hi folks - Ruud rightly pinged me again today about this. We have an alternative that Geoff has cleared... how's this?

Pursuant to a rights owner notice under the Digital Millennium Copyright Act (DMCA), the Wikimedia Foundation acted under the law and took down and restricted the content in question. A copy of the received notice can be found here: xxxxx. For more information, including websites discussing how to file a counter-notice, please see Wikipedia: Office actions and the article's talk page. Do not remove this template from the article until the restrictions are withdrawn.

Looking forward to feedback. Philippe Beaudette, Wikimedia Foundation (talk) 05:22, 11 April 2012 (UTC)

Looks great to me! Thanks so much for the follow-up!--Policy Reformer (c) 05:28, 11 April 2012 (UTC)

Hi, Philippe. Thanks for following up on this. The proposed text still implies the article is being "censored" in some sense. Yes, technically speaking content was removed, but this should have been removed in any case if we would have followed our community's content guidelines. I do think the proposed text is at least an improvement over the current banner, but I'm not sure if it will stop the occasional recalcitrant editor trying to insert a reference to the illegal copies of the specifications precisely because the banner tells him not to. Cheers, —Ruud 09:40, 11 April 2012 (UTC)

Bearing in mind that it's not perfect, but we're willing to give it a shot... I'll get someone to update the template. Philippe Beaudette, Wikimedia Foundation (talk) 19:02, 16 April 2012 (UTC)

I created {{pp-office-dmca}} and added it to this page. It substitutes notice into "notice can be found here: ..." Cheers, - Stephen LaPorte (WMF) 19:42, 16 April 2012 (UTC)

miniPCI power

Intel or PCI-SIG for "created by" field, chose one + correct date

phased out

shouldn't it be mentioned that they are starting to phase out PCI ports on motherboards, the same way they have with IDE ports? — Preceding unsigned comment added by 174.44.124.123 (talk) 01:44, 21 April 2012 (UTC)

From the very first paragraph:

PCI is being replaced by PCI-X and PCI Express, but as of 2011[update], most motherboards are still made with one or more PCI slots, which are sufficient for many uses.

I suggest further coverage would probably be undue weight. The total elimination of PCI slots is not about to occur anytime soon. Crispmuncher (talk) 02:07, 21 April 2012 (UTC).

What i'm tending to notice is that lower end ATX boards still have plenty of PCI slots because they are cheap (once you put the PCIe to PCI bridge on you can hang multiple slots on it) while higher end or smaller form factor boards tend to have few to no PCI slots. Plugwash (talk) 14:33, 12 June 2012 (UTC)

I don't think, that PCI-slots are cheaper in the overall cost for the motherboard manufacturer than PCIe x1 slots, because they need much more leads than PCIe x1 slots, which actually makes the board itself more expensive and from what I know the price of the board itself makes is a big part of the whole product manufacturing costs. imho the main reason why more expensive boards usually have less PCI slots is just because they use 2 to 4 slots for PCIe x16 slots, which are needed for SLI and/or Crossfire and the number of slots on an ATX-board is limited to 7, but the cooling solution can reduce the aavailable number of slots to a value as low as 4, so there are just not many slosts available for PCIe x1 and PCI. But most ATX-boards still have 6 or 7 slots overall and at least 1 PCI-slot, boards which use a very big cooling solution and have 4 slots for graophics cards might not have any slots left for anything else, so they don't have PCIe x1 or PCI. Of course for smaller form factors you might not find a PCI-slot. Some very expensive boards might have more than 4 PCIe x16 slots, but this is not standard because almost nobody needs that, because you cannot have more than 4 graphics cards in SLI or Crossfire (you could use a 5th slot for PhysX, but thats not good for gaming performance, because you will need to use single slot cards, with 4 cards you can use dual-slot cards if you have the right case and almost all high-end-cads are dual-slot nowadays). So boards with more than 4 x16-slots only make sense for people who use other x16 cards than graphics cards, but consumers normally don't do that (and I think we are not talking about server motherboards here). --MrBurns (talk) 21:26, 14 October 2012 (UTC)

Should delete the artitcle, and start up again.

Anyone willing to do this? --Hinata talk 23:34, 15 August 2012 (UTC)

No, we shouldn't. Why do you think we should do that? —Ruud 00:11, 16 August 2012 (UTC)

No, we can still edit it, thus not an issue. • Sbmeirow • Talk • 06:35, 16 August 2012 (UTC)

Wikipedia is not complying with the DMCA and should delete all revision history of the infringing material, because all of those past history must be deleted to remove the infringing history. Wikipedia apparently has not removed it, and, judging by the context of the complaint, I can see they can sue Wikipedia for violating the law. I also looked up and found these PCI specifications were pirated - and if you want to find the specifications, you must pay a lot of money from them and be a member. This is blatantly illegal and I see it justified. I am not bothering to look at all the history of this article, because I do not have time - however, I suggest someone do it. The best way is to delete the article, remove all revisions, and start over. That would be the best way.

Note I'm not a lawyer , and I'm not in there interest or have ties to them in anyway. Just concerned about the article. --Hinata talk 00:41, 9 September 2012 (UTC)

The PCI-SIG did not explicitly ask the WMF to also remove the the links from the history of this page and it is their policy to only remove it from the current version in those cases. I'm pretty sure it was the WMF's lawyers that came up with this scheme. If the the PCI-SIG did ask us to remove the links from the history (which they won't, they will be perfectly happy if the infringing copies of the specification can't be easily found anymore) it would be easier for the system operators to remove the link from the text in each revision in the database. It is also unclear if they would have any footing to sue as linking to infringing materials has only been found to be infringement itself in exceptional cases (i.e. thousands of links to clearly infringing material collected for the sole purpose of financial gain by the linker, also see DMCA#Linking to infringing content). —Ruud 12:12, 9 September 2012 (UTC)

Seriously? The current article's content/linkages clearly satisfy PCI-SIG (ie, no new complaints). Rename the current article, create a new article using the original name, copy the original article contents to a new article, and lock the original article (under it's new name). Voila! New (clean) history. Original article (under it's new name) can be hidden away for posterity. End of discussion. No changes in other articles referencing the original article. No offending links to be dug out of histerical references. Everybody happy. In fact, I bet it could be accomplished in less time than it took me to type in this (IMHO obvious) course of action. Rudy, I agree with you that the devil is in the details but the subject of this article is more than a decade old -- a lifetime in the personal computer industry -- thus it is unlikely there will be any new content other than link maintenance ergo no need for the old history, don't you think? Just because the lawyers haven't gotten around to complaining yet doesn't mean they won't when they figure it out and then y'all be having this discussion again. IMHO, a few simple actions now just heads off all that nonsense down the road. Just my 2¢ on the question. Y'all have a great day. JimScott (talk) 16:52, 12 October 2012 (UTC)

We cannot do that with the GFDL. This is why any cut-and-paste move that does not retain edit histories is frowned upon, because we are obliged to restore all history as much as we can. What you want is, perhaps, Oversight on infringing edits. - Penwhale | ^{dance in the air and follow his steps} 08:13, 31 October 2012 (UTC)

confusing

"PCI is [...] often shortened to PCI"

Lead photo

Edits made by IP

Office Action counter-notice

PC2 vs. full

Yes, I read most of the arb case about PC2/Kww. But I couldn't find anything that indicated why PC2 was chosen rather than full and why changing to full at this time wouldn't be acceptable. Hobit (talk) 17:29, 5 February 2014 (UTC)

This diff might help --Hroðulf (or Hrothulf) (Talk) 19:39, 5 February 2014 (UTC)

Yep. The basic idea is apparently to leave some way for the article to be open to editing by as many users as possible while still being partially protected. Full protection would severely limit who could edit. I know we don't use PC2 here and I have never supported deploying it more generally but the office believes it is the best fit here and unless there is some sort of emergency there is no reason to be altering it. Quite frankly I fail to see why it should be a big deal that the office has chosen to do this exactly one time on one single article out of four and a half million.

I had another look here because of a post at the arb noticeboard that seemed to indicate there were some sort of pressing issues here that needed our attention. I don't believe that is actually the case and that this obscure article on a technical topic of limited interest to anyone who doesn't build computers by hand doesn't really merit all the attention it is getting as a result of what was really a very minor dispute that should never have happened in the first place. Beeblebrox (talk) 20:00, 5 February 2014 (UTC)

One of the questions I've seen repeatedly asked is why semi-protection and/or PC1 wouldn't have sufficed, and we still haven't gotten an answer out of the WMF. As Kww recently mentioned on Philippe's talk page, the spam blacklist, and/or an edit filter may also have done the job just fine. All of those options have been used widely across the rest of Wikipedia's nearly 4.5 million mainspace articles, and compared to PC2, their use generally isn't controversial.

I agree with you that this entire dispute should have never happened. Even PCI-SIG's legal group overreacted in sending a DMCA takedown notice to the WMF. I disagree with you that the article is of limited interest. While the article as currently written would probably appeal more to technical readers, PCI isn't just related to board-level hardware. It also relates to software (including the way some low-level software communicates with hardware) as well as the way some hardware communicates with other hardware (even at times within a single chip). To put it simply, PCI has become extremely important within the computing industry, and continues to be important even as PCI Express becomes more and more popular as a hardware interface. --Tothwolf (talk) 04:00, 6 February 2014 (UTC)

This diff may also help. --Tothwolf (talk) 04:00, 6 February 2014 (UTC)

Thanks for the comments and links. I'd seen the first but had missed Risker's comments the first time and found them helpful. I still think Kww's moving protection to full was reasonable, but things make a bit more sense. Hobit (talk) 17:26, 6 February 2014 (UTC)

revert of edit