Talk:Newton's Identities/Proof 2

Why do you think it is a good idea to, without any prior discussion and/or alignment, introduce a new format for proofs and "explanations" without first demonstrating some examples in your own user space?

Careful attention has been given to the exact structure and wording of all headings and their structure.

With a corpus of many thousands of proofs, it is probably worthwhile to discuss before editing in main space, because the inconsistency of editors just freewheeling around has caused considerable problems in the past. After all, if we are going to be a collective of idiosyncratic minds working on their own little pet projects, then we might as well write books or math blogs.

It is the integration that sets ProofWiki apart. And exactly this integration is endangered by having too many variants of style left and right.

Did you consider the nonlocal impact of your actions? Comments much appreciated. As it stands we might have to revert or set apart your work, which would be a shame. But the damage of isolated activity, the past has shown, does outweigh the added benefit of the new content. — Lord_Farin (talk) 14:30, 15 December 2019 (EST)

The Newton's Identities page was broken. The "new format" came from page Evolute of Ellipse. The copied Evolute of Ellipse idea seemed to match the linked subpage idea in Help:FAQ/General questions/Where is the proof?!. ProofWiki:Sandbox/Template is my method for testing. It would be a relief to contribute to ProofWiki without the pressure of editing an existing page or creating a subpage. Tell me how that works (Prague time zone here). --Gbgustafson (talk) 16:49, 15 December 2019 (EST)

To submit new ideas for structuring a page, or develop other ideas, you can start crafting your own sandbox at User:Gbgustafson/Sandbox which is of a more permanent nature than the general sandbox. You can also try multiple pages at once there.

I'll elaborate more later, but let me add that the issue I take with the "Explanation" is how you are placing it on the page in general; not as an outline considered part of the proof, but rather as a separate entity that in heading style and location seems erratic to me. But as said, we can discuss this further at some point. — Lord_Farin (talk) 16:57, 15 December 2019 (EST)

issue .. with "Explanation" : I agree with you, it is ugly (decided when created). Guilty: I chose a heading at random in frustration. I would like to craft something like Evolute of Ellipse. ProofWiki HELP and FAQ have been useful but fall short of explaining the structure of Evolute of Ellipse.

I'll elaborate more later - Please do give an example of a recent page that illustrates current House Style for pages with subpage links.--Gbgustafson (talk) 05:00, 16 December 2019 (EST)

Practically all of them. Try the "Random proof" link. Or read the house style guide. --prime mover (talk) 05:56, 16 December 2019 (EST)

Did both suggestions long ago (house rules and random proof). Repeated reading Help:Editing/House_Style today and absorbed exactly one bit of new information:

If, however, a page contains three paragraphs, each with such a reference, then at least one link to Definition:Ring (Abstract Algebra) should be included in each paragraph.

That's been deleted now because it was unsanctioned. --prime mover (talk) 17:38, 16 December 2019 (EST)

Spent 40 minutes clicking [[1]] only to find 6 new examples and many repeats, because the random generator returns duplicate hits (Knuth discusses the issues). In the past, [[2]] netted Evolute of Ellipse, the best example so far.--Gbgustafson (talk) 12:01, 16 December 2019 (EST)

Start here: Help:Editing --prime mover (talk) 17:40, 16 December 2019 (EST)

We had a contributor a while back who thought his ideas were better than ours, and went out of his way to rewrite the house rules. I'm still in the process of reversing out some of his inanities. --prime mover (talk) 17:43, 16 December 2019 (EST)

Whoever could that be, I wonder :-) Seriously: look, I know that I greatly upset you about a year ago, but on those help pages I honestly tried to write down the house rules as I understood them in a neutral way. These rules were not previously written down anywhere. That is, I did not "rewrite" anything and the rules that I wrote down are most certainly not according to my ideas, but rather according to yours, to the best of my understanding. I really do not know what is incorrect about the text that I wrote and that you deleted on the help page, here.

Anyway, if you feel that you must delete all my contributions to the help pages, this is easily done: my contributions are here. I made exactly three changes to the help pages; this was on 12 October 2018, between 9:38 and 9:58 (but the one at 9:58 has already been undone).

Hard as it may seem to believe, I was merely trying to help new users and keep them from falling into the same traps as I did. I still do not see what I wrote on the help pages that does not match your policies. KarlFrei (talk) 10:27, 19 December 2019 (EST)

As requested, another example: Cardinality of Cartesian Product of Finite Sets. I took some time to find one because of my reduced activity recently.

To create this structure, you will benefit from Template:Subpage and/or Template:Subproof.

I've now edited both proofs to be more in line with what I would generally expect from page structure. Note the changed position of the explanatory paragraph.

In general we have a tradition of full detail, so it might be that we decide that only transcluding the outline is not sufficient, and we transclude the whole proof section. This would also make it more in line with standard practice.

We have never just transcluded the outline. We always transclude the whole thing. --prime mover (talk) 17:44, 16 December 2019 (EST)

To get more familiar with the subpage construct, I can recommend doing web research on "mediawiki transclusion" and/or inspecting pages like Definition:Consistent (Logic). It should really not be hard to get the hang of it. As noted before, feel free to experiment with it in your user space (so subpages of User:Gbgustafson. — Lord_Farin (talk) 14:55, 16 December 2019 (EST)

Yes, I read everything from both of you (16 Dec, 8-hour time shift).

Wow! Splendid editing result from the two of you.

The advice to use Template:Subpage and/or Template:Subproof gave me something to try this week - thanks for the tip.

The topic Newton's Identities has maybe 10 different published proofs. So far this information is only a one-sentence list of authors and years in the historical notes of Newton's Identities. After tracing the history, a short 2017 paper was found which discusses about 7 different proofs, one of which is a curious one-paragraph handwaving proof in WikiPedia (could be an XKCD cartoon).--Gbgustafson (talk) 04:19, 17 December 2019 (EST)

The trouble with citing everything you find, just because you can, is that it requires a colossal amount of work to implement properly in $\mathsf{Pr} \infty \mathsf{fWiki}$. That person usually ends up being me. --prime mover (talk) 05:23, 17 December 2019 (EST)

Duly noted and already understood. It is my responsibility to provide citations. About 25 were not cited. The dates 1629, 1707, 1959 are important and the rest can be deleted, including the citation for Donald Knuth. Deletions affect the History section, which reduces the content there to one sentence about Girard and Newton. Please make it so.

For the record, nobody has any "responsibility" to do anything.

Citations are nice to have, but not essential. (EDIT: Except for definitions, for historical reasons. We are very keen on making sure all definitions are sourced.) My general rule of thumb is that if I am actually using a particular source work for content, then I will cite it. It will be a rare circumstance that I would scour the internet for every single reference to a particular entity, but if I just happened to chance upon a particularly interesting article, then indeed I would add it.

Similarly for proofs. Having found and entered a proof for a result, I would tend not to feel I had to search through for every single proof for that same result. Result entered, proof entered, I move on. And then if I happen upon a second proof, then I would enter a second proof. And so on. Otherwise maintenance of this site would become burdensome, and I would feel I was spending an inordinate amount of time on a small detail, and I would be keen on moving onto something new.

Finishing off unfinished work of other contributors takes up more of my time than I would like. --prime mover (talk) 09:11, 17 December 2019 (EST)

Should I care about too many citations? I do care about too few citations and citation errors. In this, I am thinking about Knuth's study at home in California with a single file cabinet for Volume 1 and a battery of cabinets for the other volumes, still in progress at age 81. Who can achieve Knuth's admirable citation standards?--Gbgustafson (talk) 08:51, 17 December 2019 (EST)