|This talk page is automatically archived by lowercase sigmabot III. Any threads with no replies in 30 days may be automatically moved. Sections without timestamps are not archived.|
I do not know why there is a separate article for GUID and UUID. Probably it is for the proverbial "historical reasons", and if it were up to me, there would be only one article. I haven't worked on the GUID article, and in fact it has several problems, by the way.
The history here is that UUID's were invented by a team at Apollo in the 1980's for Apollo's Network Computing System. In the late 1980's and early 1990's, Paul Leach and others used a development of UUIDs in the Open Group's Distributed Computing Environment (DCE). A little later, Salz and others at Microsoft adopted the DCE UUIDs essentially without change for the Component Object Model and the Windows Registry, but decided to call them GUIDs. The only change made by Microsoft was to change the byte order from network byte order to little-endian byte order. Leach and Salz came together and wrote an Internet Draft spelling them out, including network byte order as "variant 1" and little-endian "native" byte order as "variant 2". Variant 2 in that Internet Draft was referred to as the "Microsoft GUID" variant, but apart from byte order and variant bits, Variant 1 DCE UUIDs and Variant 2 Microsoft GUIDs are identical. The Internet Draft eventually turned into RFC 4122.
Some years later, Microsoft changed its documentation, and started referring to variant 2 as the "legacy GUID format". Microsoft software now generates "GUIDs" which are just straight-up variant 1 UUID's, while continuing to use variant 2 as well, sometimes referring to those as "legacy" GUIDs. The current version of Microsoft's guidgen tool generates bog-standard Variant 1, Version 4 UUID's. Both variants are referred to as GUIDs in the Microsoft world. I tried to capture all this in the sentence which you reverted, which is that the term GUID is either a synonym for UUID, or sometimes refers to one particular variant of UUID. Apparently a legacy variant, though I didn't mention this, because I don't have good sources for it being entirely "legacy".
It isn't my fault that the GUID article exists and that it is confusing, and the existence of the GUID article shouldn't be a reason for anyone to think that GUIDs are somehow different from UUIDs. There is no real difference, other than terminological.
I am going to try to cover some of this in the article and I ask you kindly not to undo it again, because it is correct, and the wording which you apparently prefer does not capture the situation: "GUID" is not just a variant of UUID. Many of them, perhaps all the recently-generated ones, are in fact UUID's.
- I don't think you're wrong, but as long as GUID and UUID are separate articles, their introductions should not say that the two terms are synonymous. So rather than confusing the reader (which is what I just fixed) or lecturing me about it, the thing to do would be to propose a merger between the two articles. – voidxor 22:38, 15 January 2017 (UTC)
The Talk page for the GUID article discussed merger a couple of years ago, and that seemed to be the consensus. I am just an anonymous person interested in UUIDs and in making that article better. The article needs to say that UUIDs and GUIDs are the same thing, because that is the case, and it would be confusing not to say it. The GUID article needs to say the same, because what is true in the UUID article is true in the GUID article. It is not my problem that the two articles exist on Wikipedia, despite UUID and GUID being synonymous. If you want to propose the merger of the articles in some inner Wikipedia sanctum, and argue with people about it, have fun. — Preceding unsigned comment added by 18.104.22.168 (talk) 01:39, 16 January 2017 (UTC)
- Okay, as you've probably seen by now, I just closed that old merge discussion and performed the merge. Thank you for pointing out that discussion.
- I sense your frustration, and agree that GUID and UUID are more or less synonymous. However, rewording the introductions of those articles to say as much was the wrong way to go about it; I said so (above) and partially reverted your changes. I even advised you on the correct way to deal with two articles about the same thing (a merger), but you impatiently didn't wait for that discussion to play out, and reverted my revert.
- Wikipedia has been around a while, and has lots of processes to keep things running smoothly. It's not sanctimonious; it's just experience. Process is important! Ignoring it, as you did, tends to waste everybody's time, including your own. For instance, you just spent hours working on an article that even you were arguing needed to go away. – voidxor 00:12, 17 January 2017 (UTC)
Regarding the paragraph which you reverted about randomly-generated node ids, as the paragraph itself stated, this is directly from RFC 4122 itself, as are almost all the rest of the details about UUIDs currently in the article. Having mentioned the RFC and provided links to it (wiki automatically makes any mention of "RFC 4122" a link) the article is not citing particular sections of the RFC for the various technical details. But if you would like to check, the section on randomly-generated node ids (including the detail about the multicast bit), is 4.1.6. This then refers to Section 4.5 which sets out how the random node ids can be generated. 4.1.6 says that random node ids should be used when there is no MAC address. 4.5 expands on this a bit by adding that random generation can be used when there is no IEEE 802 address (i.e. a MAC address), "or its use is not desired". 22.214.171.124 (talk) 02:29, 18 January 2017 (UTC)
- You're right; I'm sorry. I viewed voidxor 04:06, 18 January 2017 (UTC) in wikicode form, and thus didn't see the links. I'll self revert and add overt footnote-type citations to the RFC. –
I'm putting this here, because it seems like you don't read the UUID talk page. Regarding version 1 UUID rollover, the time is a 60-bit number in 100-nanosecond intervals since 15 Oct, 1582 00:00:00.0 UTC. 60 bits of 100 nanosecond intervals is 3653.3877 years. Add that to the 1582 AD and you get 5235.39 AD The uuid program on Linux, when displaying v1 UUIDs, rolls over at March 31, 5236 AD, which is consistent. Surely, doing simple calculations is not "original research". Is it? Oddly, RFC 1422 states that the time rolls over in 3400 AD, "depending on the algorithm". This is an odd statement because (1) there is only one algorithm for time in the RFC, so it is hard to see what other algorithms it could depend upon; and (2) according to the one algorithm mentioned in the RFC, the rollover is 5236 AD not 3400 AD. It looks like at some point there was a different, or more than one, time algorithm, and the RFC was never updated. 126.96.36.199 (talk) 14:30, 25 January 2017 (UTC)
In trying to reconcile 5236 AD versus 3400 AD, I realized that 3400 AD is what you get if the 60 bit timestamp is a signed number, rather than unsigned. 188.8.131.52 (talk) 14:45, 25 January 2017 (UTC)
- I don't know. Obviously Wikipedia has lots of worked examples, particularly in math-related articles, and those needn't be cited. But the way you'd written it looked more like a fact than a derived example when . Regardless, I like the way that you , attributing the rollover date to the RFC.
- And I do read the talk page; I'm just less of a subject-matter expert than you and Andy. So I'm letting you guys direct the evolution of the article. My interest is more that Wikipedia policies are adhered to, and you are now doing a much better job of that. Speaking of which, thank you for all of your hard work on the article. – voidxor 17:43, 25 January 2017 (UTC)
Hope to see you there! The Transhumanist 16:03, 12 May 2017 (UTC)
- @The Transhumanist: Thanks for reaching out, but I've been too busy in real life to do much editing recently. If that changes I may look into the WikiProject. – voidxor 04:02, 13 May 2017 (UTC)
Hello Voidxor, those lines are needed, because without them MediaWiki considers the description lists to be split there (as if there were a completely empty line instead of the lines with only a colon). So, after a nested list, there has to be such an additional line to get the result one would expect without them. Check it in the developer tools of your webbrowser (usually right-click, "inspect element" or similiar) ;-) --nenntmichruhigip (Diskussion) 17:39, 9 June 2017 (UTC)
- @Nenntmichruhigip: Really? That seems like a bug in the MediaWiki software. I hate having to do workaround hacks like that. Anyway, I've partial self reverted. – voidxor 01:08, 15 June 2017 (UTC)
- I tend to agree on it being a bug. Actually there’s already a ticket for it, but I don’t expect it to get fixed soon. I consider the workaround as tolerable to get valid markup in the meantime, because empty definitions are already silently dropped, so it probably won’t become an issue when the bug is fixed. --nenntmichruhigip (Diskussion) 11:37, 16 June 2017 (UTC)
You reverted the discussion of Forms D, K, X, Y and Z contacts on the basis that they were uncited. In fact, the citation at the head of the enclosing Contact Form section (to the Relay Handbook by the National Association of Relay Manufacturers and its successor, the Relay and Switch Industry Association) is the source for all of the contact forms. The citation you restored (to a data sheet from Matsushita) is one relay manufacturer's summary of the material relevant to their small-signal products. I don't see a need for Wikipedia to describe all of the relay and switch contact forms defined by the NARM/RSIA, but A, B and C are not enough. I want Form D described because the C/D comparison illustrates the issue of Make-Break order. Center-off switches (Form K) are common, while center-off relays are rare. Forms X and Y are common in toggle switches with ratings over 1/4 HP, and form X is extremely common in power contactors, for example, the contactor linking a thermostat to an air conditioiner compressor. Douglas W. Jones (talk) 15:52, 25 June 2017 (UTC)
- @Douglas W. Jones: Since you duplicated the above comment in two different places, I am responding at Talk:Electrical contacts#Contact forms. – voidxor 19:27, 25 June 2017 (UTC)
"Fixing" other people's comments
- Please familiarize yourself with the policy you are linking to before going around lecturing others: "Some examples of appropriately editing others' comments: ... Fixing format errors that render material difficult to read. ... Examples include fixing indentation levels,..." – voidxor 03:54, 6 July 2017 (UTC)
New railway glossary subpage?
Hi voidxor, thanks for the welcome message. I see that you are very much involved with the rail glossary page. I am working on a major European rail cooperation project which includes some development of topic specific glossaries. I think it would be great to make that content available on Wikipedia and possibly even lure the experts to work directly here. To be honest, I think many of the people in the industry already search for terms and content in Wikipedia because the relevant technical standards are typically locked away somewhere and it would take too long to find them or request them from someone. Anyway, I am thinking on how this could be done, as some of the terms are very topic specific, e.g. Automatic Train Operation. Do you think a sub-glossary for ATO would be a way to go? I think it could be more useful in this way, as people could read terms about a certain topic in one place and not swamp the general glossaries with such specific words.
Usually we reference technical standards or the EU railway authority. The glossaries are published as part of subsets, so new terms coined within the project will always reference the subsets. Out of curiosity: is there any case you know of, of experts using Wikipedia as their working tool? It is probably a silly question. For very conservative railway sector this would be a huge step towards openness and an improvement on current working methods. (Renatoar (talk) 08:36, 11 July 2017 (UTC))
- @Renatoar: Sorry for the delay in getting back to you. For full transparency, I should start by saying that I'm more of a Wikipedia expert than a rail expert. While I certainly appreciate rail technology, I don't work in that industry and am by no means an expert. My interest in the rail glossaries has been more in getting them up to Wikipedia's standards (everything cited, for example).
- With that in mind, I'm not going to be to "talk shop" in your area of expertise. I can, however, provide guidance on Wikipedia's policies and procedures. My biggest concern with your proposal is the issue of notability. Wikipedia's notability policy requires topics to meet a certain threshold of mainstream attention before they are deserving of their own article, and I'm not sure a sub-glossary of ATO would meet that threshold.
- Many experts do edit Wikipedia (I've actually worked with some librarians, professors, and scholars to do exactly that), but they need to do so from a perspective of helping to build an encyclopedia, rather than using Wikipedia as a venue to publish their original research. If you and your colleagues would like to build a repository of specific knowledge without constantly running into Wikipedia's requirements, I might suggest starting your own blog or wiki outside of Wikipedia (e.g. on Wikia).
- Alternatively, if your goals are smaller scale to start, I'd just create a glossary section on the automatic train operation article. Just be sure to cite your sources. Happy editing! – voidxor 21:20, 16 July 2017 (UTC)
- That is precisely the type of advice that I need(ed). I have been busy with other stuff but I will get back to this in the next weeks. As long as things are on Wikipedia so everyone can consult them, it is fine really. Thanks! Renatoar (talk) 13:03, 26 July 2017 (UTC)