Template talk:Molecular formula index

Enhancement for subdivision and improved navigation in & among Cat(s)

    I'm about to lock the template bcz i need to sleep, & bcz unless another editor with an idea at least as good comes along within about 24 hours, i'll add, at least as a straw man to solicit ideas for more the most useful form of the enhancements i have in mind, some conditional code so that subcats can be assigned instead via the template, and the addressability within Cats (and perhaps subcats) can be made useful. Someone else may quickly have better ideas, but i like to think that despite my info-theory-heavy, chem-experience-light status, i've gotten a little bit of a head start that the next designer will want to weigh before doing it better without incorporating any of my thots.
    I think my first cut, just suspended, is probably inefficient: the shorter the chain, the fewer the compounds, so putting the 3 single-carbon ones under "1", the couple dozen 2-Cs under "2", etc. is a waste; "0" 1-9 for Cs, 1 for 10-19 etc. would be more useful. I like the Inorganic subcat, and perhaps more serious chemists will prefer Category:1-to-9-carbon molecular formula disambiguation pages and similar sub-Cats. Or the template could parse the formulas so that C31H52N2O23 gets a sort key something like 31.52.02.23, and the template could be driven by a manually edited table, so that the starting points in the sequence at which each of 26 headings (A-Z) start could be adjusted (as trends in research create hot spots that overpopulate a few headings), keeping the breaks between sections easy to use in accessing by formula.
    All that said, proceed with your own approach if you think i'm, uh, tangled in my underwear; we'll sort it all out in the end.
--Jerzy•t 07:53, 23 -->    All that said, proceed with your own approach if you think i'm, uh, tangled in my underwear; we'll sort it all out in the end.
--Jerzy•t 07:53, 23 & 18:02 24 July 2011 (UTC)
July 2011 (UTC)

   The organization of pages within Category:Molecular formula disambiguation pages is currently in flux, while uses of the (enhanced) template are being edited to use the new |heading= switch.
   While real chemists may be able to do better yet, using the same new mechanism, this schema will probably work much better than the old one, where

every page was in the C portion of the Cat, with the exception of 3 inorganics each under its own letter,
the one indefinite-length polymer, which was, I kid you not, in the "(" portion, and
the two-carbon pages were between the 20-something-carbon ones and the 30-something-carbon ones, etc.

   The sorting scheme will at least organize organic compounds (or other organic structures) with fewer than 10 carbon atoms into 9 groups, each homegeneous as to number of carbons; the headings over those groups will be the digits 1 thru 9. Under A and B will be the low and high 'teens of carbons, respectively. C will be reserved, but may collect pages for organics that are added (contrary to the Cat page's request) "by hand" or without providing the needed "heading" parameter. D and E should comfortably accommodate those with 20-29 carbons and with 30 or more, respectively.
   There will remain some annoying anomalies: C10H18O... will precede C10H9O.... But -- since strict alpha-numeric order seems maintained within sections when single-letter heading params are used to impose non-default sort keys -- the dumb alpha-num sort does not too badly offend our intutive sense that an element name, whether 1 or 2 letters, is a unit that gets sorted relative only to other element names, and a number, of no matter how many digits, is a unit that gets independently sorted relative only to other numbers.
   (Hmm -- contrary to my wild musing before, i confess there's never been a Mediawiki mechanism whereby markup can decompose the constituent characters of a page name, or other text string, and base action on the individual characters, but -- in theory we could expect template users to decompose C10H18O2 into separate parameters C|10|H|18|O|2 and write a Wiki-markup procedure to recode that in a string like C 0AH 12O 02 that could convert such formulas into sortkeys that would simulate that intuition. But that would be a lot to expect of editors, even compared to the manual extraction of the carbon-atom count that I've coded for.
   (Specifically, i've counted on template users to examine the first 3 characters of the page name to extract the number, and then (when the "heading" parameter is not a single digit) to choose among the fairly arbitrarily named 4 ranges of two-digit values.)
--Jerzy•t 06:56, 25 July 2011 (UTC)

Category:Molecular formula disambiguation pages

See discussion at Wikipedia_talk:WikiProject_Disambiguation#Category:Molecular_formula_disambiguation_pages. PamD 18:55, 28 February 2012 (UTC)

Default sort?

The category sections and headings do not appear consistent or logical.

• Shouldn't the category headings simply be: C C2 C3 .. C10 C11, etc? (ie. not 1, 2, 3, 10-19, etc)
• Shouldn't each formula be sorted numerically by subscript, eg. C9H9, then C9H10?
• Can't we get the correct sort, by simply adding, for example,: DEFAULTSORT:C09,H09,x,y, and DEFAULTSORT:C10,H9,N,O2, etc?

Since we using a template, can't we just derive this information form the formula, ie. C10H9NO2 becomes DEFAULTSORT:C10,H9,N,O2? And I would think that a template could derive the correct DISPLAYTITLE and add the _.. in the same way, so we would need Molecular formula index to have any parameters? --Iantresman (talk) 13:34, 7 December 2013 (UTC)