Wikidata:Property proposal/number of pins, number of pin positions

number of pins, number of pin positions

Originally proposed at Wikidata:Property proposal/Generic

   Not done

Description	2 related properties: number of pins: number of contacts that an electrical connector has (excluding any grounding shroud not used for communication) number of pin positions: number of positions in an electrical connector, including empty or keyed positions
Represents	electrical contact (Q394001)
Data type	Quantity
Template parameter	"num_pins" in en:template:Infobox connector; "contacts" in w:en:Template:Infobox CPU socket
Domain	item: Subclasses of electrical connector (Q2119531). optical fiber connector (Q2296938) could be added later, pending discussion of how to handle optical module (Q48740842)s (which have electrical contacts on one end and optical connections on the other).
Allowed values	positive integers
Allowed units	none
Example 1	USB-C connector (Q58051489) → "number of pins" → 24 → "number of pin positions" → 24
Example 2	NEMA 5-15 (Q24288456) → "number of pins" → 3 → "number of pin positions" → 3
Example 3	6P2C modular connector (Q64831598) → "number of pins" → 2 → "number of pin positions" → 6
Example 4	Intel HD Audio connector (Q64764371) → "number of pins" → 9 → "number of pin positions" → 10 citation: https://www.intel.com/content/www/us/en/support/articles/000005512/boards-and-kits/desktop-boards.html
Example 5	Socket F (Q1475023) → "number of pins" → 1207 → "number of pin positions" → 1225
Planned use	Annotating items for electrical connectors. They could also be generalized to optical fiber connectors, since it's fundamentally the same concept.
Expected completeness	eventually complete (Q21873974)
Robot and gadget jobs	When only "number of pins" is specified, "number of pin positions" can be populated with the same value.
See also	Wikidata:Property proposal/contact area count
Type constraint – subclass of	electrical connector (Q2119531)

Motivation

For annotating items for electrical connectors (see connector (P2935)).

Why are 2 properties needed? Some connectors have additional positions which are not filled. For example, the RJ11 w:en:Modular connector has 6 physical positions, but only 2 of them are populated. Similarly, the Parallel ATA (Q230360) data connector has 40 positions but only 39 pins, since one position is keyed. The unfilled positions should be counted when the connector's conventional numbering includes them.

Aliases: replace "pin" with "contact" or "conductor"; same without the initial "number of" 2620:0:1000:3216:5413:4AF0:7532:8655 23:59, 21 June 2019 (UTC)

WikiProject Informatics has more than 50 participants and couldn't be pinged. Please post on the WikiProject's talk page instead. 169.234.25.175 20:28, 5 July 2019 (UTC)

Discussion

•  Support David (talk) 06:33, 22 June 2019 (UTC)
• Why not has part(s) of the class (P2670) along with quantity (P1114)? ChristianKl ❪✉❫ 18:40, 22 June 2019 (UTC)

• In principle, we could do it by combining qualifiers with an entity for a pin; there are just some practical issues:

• It's not totally obvious for an editor which combination to use, and I don't see a way to hint recommended values. For example, quantity (P1114) with applies to part (P518) looks just as plausible (even if you're not supposed to use it that way). Also, people might use other elements besides the intended "pins" and "pin positions", e.g. pogo pin (Q1400617), wire (Q551997), or lead (Q947546). This will make querying more difficult.
• It's easier to express a constraint that the "pins" and "pin positions" should be specified together when they are expressed as a property.
• We might want to generalize this to support w:en:optical connectors; we would then have to add some items for pretty abstract concepts that encompass "electrical contact", "termination of a single optical fiber" (is there even a term for that?), and positions thereof. These would be hard to find.

In any case, there are lots of existing "number of X" properties, where X is a domain-specific part that an item contains. 2620:0:1000:3216:5413:4AF0:7532:8655 20:39, 25 June 2019 (UTC)

•  Support Seems like a fine idea. —Scs (talk) 11:18, 30 June 2019 (UTC)
•  Split support and oppose  Oppose For the given examples, RJ11 is only valid for countries that only host 2 wires. You can find RJ11 with 4 wires in 6 positions. RJ45 can also be used to replace RJ11, so RJ11 and RJ45 will have several possibilities. In addition, you write "electrical connector" which is quite broad: you give an example of an IC, which enters the electronic domain. With an electronics training, you can find a multitude of cases (so number of pins variable) for a single component (therefore for a single manufacturer and a single function and sometimes with a single denomination). The differences are according to the use of the component: assembly, power, disposition, etc. For the multilingual side of WD and the complexity of the domain (electricity), I fear that this future property brings errors. Of course, in obvious cases, this property will work, but contributors will not stop there. —Eihel (talk) 17:15, 11 July 2019 (UTC)

• Thanks for the feedback!

• Regarding RJ11, the Wikipedia article explicitly says RJ11 requires a 2-wire connector (without citation); if you have better information please update that. I switched the example to 6P2C modular connector (Q64831598) for precision, which I forgot to do earlier.
• I'm not sure what IC you're referring to, but assume you're talking about the CPU socket. I'm aware that a given IC can be packaged in multiple ways, but as far as I've seen, a given CPU socket necessarily has a well-defined number and arrangement of pins, so I don't see the issue.
• Regarding the risk of misuse, would a property type constraint (should be used on connectors only and not ICs or other components) address this? I'll update the proposal.

73.202.12.249 23:22, 17 July 2019 (UTC)

1. w:fr:RJ11#Belgique. Indeed, 6P2C is normalized in the sense of the use of pins.
2. Yes, CPU sockets remain the same as far back as I can remember. But I suggest you look for the term 7805 on your search engine (only browse images to make your life easier). It is a simple CI used in voltage regulation and there is a multitude of form, manufacturer, etc. , so several possibilities. And it's the same for a multitude of CIs, from the simplest to the most complex. Sorry. There is conflicts-with constraint (Q21502838)
   

   relation (P2309) → instance or subclass of (Q30208840)
   

   class (P2308) → electronic component (Q11653)
   

   constraint status (P2316) → mandatory constraint (Q21502408)
   but this restriction will overshadow many items (In addition I do not know if it works!)

3. In the content of pages linked to the infobox on enwiki, it seems to be appropriate. 78xx does not contain a number of pins and the page does not contain this infobox. —Eihel (talk) 03:26, 28 July 2019 (UTC)

•  Support I had pretty much the same ideabut called it "contact area count. The reason is that eg. an LGA CPU has no pins. So I searched for a more general name that can be used for multiple things. number of pin positions is an interesting idea, however I'm not sure if this is really the way to go.

An example: A CPU socket has 1000 pin positions. CPU a) has 900 pins, because it's missing some workstation ECC RAM functions. CPU b) has also 900 pins, but not the pins for ECC RAM, but for the ones for CPU interconnects (so a single socket setup). CPU 3) has only 800 pins. It supports CPU interconnect and ECC RAM. He 200 missing pins are because it's a low end version that does not use as much power. How would you include that for the socket (and is the socket the right place to collect that information or should this rather be collected in the CPU item)

A second example: The Type 2 connector that is used to charge electric cars has 7 pins. The maximum output is ~43 kW with 7 pins. For slow charging out of eg. Schuko plugs it is about 3 kW and here only 5 pins are used. It is the same connector and the same layout. One charging type just misses 2 metal connectors. --D-Kuru (talk) 11:58, 2 August 2019 (UTC)

 Comment It sounds like maybe the item for the interconnection standard should list in some way the meaning of all the pins? Though that might require an item for each pin position that further describes it? Like, pin 3 carries this signal with this voltage, etc.?? And then items that conform to that standard would specify that conformance, and which specific pins they actually use? ArthurPSmith (talk) 18:08, 2 August 2019 (UTC)

If it's not an open standard you will never now why one pin is used and the other isn't. A description for every pin is just not usefull in my opinion. In my opinion a "number of used pins" is not really ncecessary since a configuration can change over time and with the used device. If eg. a CPU socket hast 4000 pins and a CPU that fits the socket has 3000 pins, the number of pins for the socket is still 4000 (even some are not used - for now - but could be used at some point) and the CPU also still has only 3000 pins. For me, this applies also for eg. USB or SATA where one pin isn't in use. If USB version 2.0 uses 10 out of 20 pins, the number of pins would be tagged as 10 on the USB connector side. If USB 3.0 uses 20 pins, the pin information could be tagged for version 2.0 and 3.0. But you would not have to set any information about how many pins are not used in the USB connector. What if the USB connector is used anywhere else for some other connection that actually uses all 20 pins right from the start. It's an open standard, so this could happen quite easily. --D-Kuru (talk) 22:07, 6 August 2019 (UTC)

•  Support Interesting idea. I'm curious to see how it goes. --- Jura 09:05, 4 August 2019 (UTC)
•  Comment So if 2 properties are actually needed here, could somebody split the template into 2 separate proposal templates with the appropriate examples etc, to make it a little easier on the property creator(s)? ArthurPSmith (talk) 17:09, 7 August 2019 (UTC)

•  Comment An idea: As far as I saw there is has part(s) of the class (P2670) which can be combined with an item and you can insert a value in that item. I saw this for CPUs where L1, L2 and L3 cache is listed in said property. Is this an OK thing to do or is a property prefered here? If it would be fine to use items instead of properties this could be solved quite easy. --D-Kuru (talk) 13:09, 15 August 2019 (UTC)
•  Oppose has part(s) of the class (P2670) lead (Q947546) with a qualifier of quantity (P1114) should suffice? Subclasses of lead (Q947546) could also be created to distuingish between connected lead (Q947546) and disconnected lead (Q947546). Dhx1 (talk) 15:01, 27 September 2019 (UTC)

•  Support轻语者 (talk) 07:53, 27 February 2020 (UTC)

•  Comment How many pins and number of pin positions do Mini-USB (Q16578670) or Micro-USB (Q1931429) have? Do they always have 5 pins and 5 pin positions? Or are there case where they have 4 pin or pin positions? What the number of pins / pin positions can apply to? Only connectors and similar things? In any case it's very useful to be able to represent the number of pins. It would also be even more useful to enable to have more in depth proprieties about each of the pins such as the voltage levels, the pin name, if it's connected or not, if it's analog or digital, and similar things. You could then potentially reuse that information in CAD software like KiCad (Q942363) if it's mentioned in components. It might be ultra useful for standard connectors or standards like USB connectors, as people often need just the pinout of such connectors for a given standard like USB 2.0. It could even be used to describe the peculiarities of some connector used on some computers or other hardware for instance. GNUtoo (talk) 15:24, 22 May 2020 (UTC)
• @ChristianKl, Scs, Eihel, D-Kuru: @Jura1, Dhx1, 轻语者, GNUtoo:  Not done Closing as stale, also proposal not formatted correctly, open issues not addressed. Any new proposal for this should reference this discussion. ArthurPSmith (talk) 16:13, 11 January 2021 (UTC)