This site is supported by donations to The OEIS Foundation.

User:Russ Cox/OEIS Server Features

From OeisWiki
Jump to: navigation, search

This page collects planned or suggested features for the OEIS server software.

Using — ~~~~ for signing makes the signature stand out (the — for increased readability, the ~~~~ for the linked signature with time stamp) like just at the end of this very comment. — Daniel Forgues 16:35, 4 September 2011 (UTC)

Planned

  • Warn about dangling cross-references when a sequence gets recycled.
  • Warn about long sequence data lines during sequence editing and suggest b-file instead.
  • Bring back yellow highlights (suggested by User:David Scambler)
  • Bring back recent.txt, add recent changes RSS feed, add recent new sequences RSS feed.
  • "Subscribe" to specific sequences and receive email for updates (suggested by User:Alexander R. Povolotsky)
    • Auto-subscribe as a result of editing sequence, possibly.
    • Send pink box comments to all subscribers.
  • Make text form of entire OEIS available for download.
More generally, how about making the "text" format of a search result available for d/l ? — M. F. Hasler 18:54, 5 November 2011 (UTC)
  • Merged login status between sequence editing and wiki editing. Merged search.

Suggested

  • Various suggestions, in entirely random order D. S. McNeil:
    • Regarding "long sequence data lines": given the terms and an offset, we can automatically generate a b-file (if one doesn't already exist). Make this a one-click option.
      strong "+1" (and replace "one-click option" by "default behaviour"). This could (almost(?)) replace my suggestion of a special page / form to generate a b-file directly by/on the server, from pasted sequence data. — M. F. Hasler 19:00, 5 November 2011 (UTC)
    • My hobbyhorse: institute a "perform sequence analysis" button which runs a small script scanning the sequence for common problems. [Am willing to code in any needed language to help get this done.]
    • This one's probably not worth the trouble of implementing and would slow down page generation, but I mention it anyway: colour or otherwise flag sequences on the page by whether or not you've made an edit or commented on them, so you can follow the sequences you're helping work on.
    • We should either make a "null edit" possible, to bring sequences back up into the mix when one gets approved when we're not sure it was ready yet (sometimes people are making minor punctuation changes to obtain this effect), or agree on a convention (say temporarily adding "uned").
  • The current behaviour to show the diffs is often really useful. Sometimes it's a bit troublesome, though, like when you want to cut and paste to see if a new code works, or to compare the proposed version of a sequence with your own. A switch to toggle diff/non-diff (via the usual CSS tricks used for line numbers in code, etc.) would be really handy. D. S. McNeil 09:12, 25 February 2011 (UTC)
    I think the CSS tricks won't work: a text/HTML portion with display:none will make its way into the clipboard. — M. F. Hasler 19:08, 5 November 2011 (UTC)
  • Suggestions regarding the oeis.org/draft page David Scambler 18 March 2011
    • A new column with attributes summarizing what has changed e.g. "NCF" means name, comments and formula section changed.
      • also "!" to indicate pink comments (T. D. Noe)
      • also "*" to indicate new sequence (i.e. not a currently published sequence) (Charles Greathouse)
    • also hover-over A number shows proposed name of sequence (franktaw at netscape.net)
  • From Olivier Gerard:
    • Sort draft page by proposer, edit count.
    • Display edit times in the sequence review list and in the draft page according to the user local time zone (useful for Europeans and Oceanians)
    • Send pink box comments to everyone having entered a pink box comment for this sequence in the current draft phase
    • (not important but cool) a few dynamic graphs of editing/reviewing activity day by day for the whole OEIS (such as number of edits mades, number of allocated sequences, number of accepted edits, size of the edit queue, ...)


  • On the submissions page, when a long submission entry is entered, the title and submit button center on that entry and stray off to the right. I think it would be better if they were horizontally centered on the OEIS logo at the top of the page. User:David W. Wilson
  • Remove the %STU lines from the database and use the b-files as the sole repository for sequence terms. In the submission form, whether the user enters explicit terms or uploads a sequence file, the terms are stored in the b-file. All displays of the terms (sequence pages, internal format %STU lines, list, table, and graph values) are gotten from the b-files. This eliminates %STU line formatting issues and %STU line/b-file consistency issues. User:David W. Wilson.
    Note: Under this proposal, a formatted sequence would normally show the standard amount of sequence data. Neil was concerned that some sequences need to display a nonstandard amount of data. The %S line (now available) or the %O line could be used to specify a nonstandard number of terms to display (or 0 for standard number of terms). David W. Wilson 01:27, 18 June 2012 (UTC)
  • Display a popup with links to (i) Recent contributions (ii) User home page (iii) User Talk page (iv) Send message to user [(v) ...], whenever the mouse pointer hovers over a user name (e.g. in AUTHOR line, ideally all occurences of user names [e.g. at the end of any line, preceding a date] should be "hyperlinked" in that way). — M. F. Hasler 19:40, 5 November 2011 (UTC)
  • It would be nice to have categories for sequences on the main OEIS, as was the case when the sequences had their pages on the wiki. — Daniel Forgues 04:17, 25 October 2011 (UTC)
    I wanted to add the same suggestion. This would really help to (re)organize the whole content in a simple and efficient manner. — M. F. Hasler 03:55, 26 October 2011 (UTC)
    I have developed an idea for an extremely simple yet probably sufficiently complete and powerful implementation of categories (as user-definable keywords and associated ("pseudo-sequence") entries, at User talk:M. F. Hasler/Categories. Please feel free to add comments and suggestions there or recycle the idea in any other useful way. — M. F. Hasler 13:12, 27 January 2012 (UTC)
  • Have a button "Add note to discussion" so we don't need to do some edit just to add note to discussion (pink boxes.) — Daniel Forgues 16:59, 8 November 2011 (UTC)
  • Support index sequences. If a sequence is defined on, say, squarefree integers, allow it to specify A005117 as its index sequence, perhaps by some extension of the %O line. Allow it to be displayed in b-file or graph format against its index sequence. David W. Wilson 01:41, 29 February 2012 (UTC)
    I've often considered something like this, though in the more general sense of listing subsequences and supersequences in a standard manner, something that allows searches of the form allin:A000040 (though I'm not tied to the syntax). (It would be even better is the searcher was smart enough to respect the inherent DAG structure, so that (for example) listing a sequence as a subsequence of A40 would make it automatically a subsequence of A27, assuming A40 listed itself as a subsequence of A27.) Obviously in either case there's a lot of populating that would need to be done. Charles R Greathouse IV 08:08, 2 March 2012 (UTC)
  • On the sequence edit page, add a section showing the sequence in its post-edit form, without any markup (as it will look once your edits have been approved). David W. Wilson 16:57, 4 March 2012 (UTC)
  • I had the idea of multi-sequence submissions. It would be convenient to enter a comment, formula, or crossref that mentions two or more sequences and have the comment copied to all mentioned sequences, and perhaps other specified sequences. David W. Wilson 00:58, 21 March 2012 (UTC)
  • It would be nice to have a third Plot in the button “Graph”. The appearance of the third Plot should be the same as "Plot 2" or “Plot pairs of sequences in the OEIS” with the following configuration. Sequence 1: A?????? untransformed (our sequence). Sequence 2: A000004 (always) untransformed. Contents of Plot: “Plot Seq2(n+shift) vs n and Seq1(n) vs n”. Appearance of Plot: “Draw line segments” (only). For example see the third Plot of A005132 Recaman’s sequence. See also A064413 EKG sequence. Many other sequences may show a nice graph using this Plot. Omar E. Pol 11:43, 14 April 2012(UTC)
  • Bugfix: When I edit A124162 the line-initial tilde in the formula line disappears. I imagine this is related to the use of a tilde to represent an empty input box. Charles R Greathouse IV 04:24, 6 August 2012 (UTC)
  • Feature request: when clicking "Plot 2" from a sequence, pre-fill one of the fields with the sequence number. This is useful because you can store the A-number for the other sequence in your clipboard. (Possibly the "Plot 2" link could be moved to, or added to, the line with "Graph"; it's not obvious that this is a good idea, though.) Charles R Greathouse IV 04:24, 6 August 2012 (UTC)
  • It would be nice to have a link from the Draft page of every user to the page https://oeis.org/draft . Thanks in advance! - Omar E. Pol 00:47, 14 December 2014 (UTC)
  • It would de nice to have the total number of entries in the pages of https://oeis.org/draft, for example, "Total number of entries: 256" below of the title "Proposed drafts" or above of "Showing entries 1-100 | older changes". The same for the pages of https://oeis.org/editing. Omar E. Pol 14:14, 10 October 2015 (UTC)

[Please sign your edits.]

Likely Not to Happen

  • I strongly request implementing emails sent from OEIS to authors/editors, confirming the fact of the sequence's submission User:Alexander R. Povolotsky
    • That is completely redundant; submissions/proposals are immediately confirmed on the Pending Changes page. - Klaus Brockhaus
      • Some people at times may not have access to web but still could receive and read emails so the above "redundancy" claim by Klaus Brockhaus is not useful. User:Alexander R. Povolotsky
  • Turn off the nearby-word scan for searches which are in quotes. If I'm searching for a typo, say "mentionned", I don't want it to autocorrect to "mentioned". (suggested by D. S. McNeil)
    • The indexing applies the Porter word stemming algorithm to all input when generating the index. It can't be turned off, at least without significant rework. Doesn't seem worth the effort. We should make the text form available again soon. User:Russ_Cox
  • Add the column "Author" in the page http://oeis.org/draft, for example: "Sequence Author Status Time Draft by Reviewed or edited by". [Omar E. Pol]
  • Add the column "Author" in the page http://oeis.org/draft/editing, for example: "Sequence Author Status Time Draft by Reviewed or edited by". [Omar E. Pol]
    • Not sure this helps. The Author does not seem relevant to the edit being made. --Russ Cox 18:21, 29 March 2012 (UTC)


  • Cf. ToDo#Known Problems: When hovering over the link to (the wiki version of) a sequence, the pop-up text should include the sequence definition (the %N line). A000040 should be automatically like this: A000040 (manually done with the {{abbr}} template, e.g. {{abbr|The prime numbers.|A000040}}) — Daniel Forgues 18:16, 28 October 2011 (UTC)
    • The wiki doesn't have fast enough access to that information. Making the current links fast was already quite challenging. --Russ Cox 18:24, 29 March 2012 (UTC)

Completed

http://oeis.org/A000001/internal
https://oeis.org/edit?seq=A000001&internal=1
but both send more HTML than needed and the latter requires a login. This would be useful for reducing server load from bots checking sequences, and can be seen as a limited return of the eisBTfry00000.txt files.
Charles R Greathouse IV 03:09, 10 July 2011 (UTC)
Completed: [1] Charles R Greathouse IV 22:02, 27 July 2012 (UTC)
  • Simple one: change
<span dir="LTR">Search:</span> <b>search string here</b>
to
<div style="overflow:scroll"><span dir="LTR">Search:</span> <b>search string here</b></div>
in the search results so that long queries don't cause horizontal scrolling.
Charles R Greathouse IV 21:59, 31 July 2011 (UTC)
I tried this but it didn't work: for many-word searches it already wraps, and for searches of very long numbers (no spaces) the div does not change the behavior. What search did you have in mind? Russ Cox 18:15, 29 March 2012 (UTC)