This site is supported by donations to The OEIS Foundation.

Talk:QandA For New OEIS

From OeisWiki
Jump to: navigation, search


This talk page might be a good place to discuss several topics without flooding the seqfan list. — M. F. Hasler 12:35, 15 November 2010 (UTC)

Suggested question: "Who is entitled to propose, to review, to approve?" — Klaus Brockhaus 11:33, 21 December 2010 (UTC)

User page for all contributors

I strongly support the idea of creating a user page for all contributors. Benefits are:

  • the possibility to avoid displaying all the e-mail addresses throughout the sequences
  • the possibility for any contributor to change his e-mail address "globally" whenever he wants without requiring any action from the OEIS maintainers.

I'd even suggest to run a bot to convert all "traditional signatures" to a link to a user page, to be created if it does not yet exist, also for all "" contributors. Below I will further disgress about ideas concerning the signatures. — M. F. Hasler 12:35, 15 November 2010 (UTC)

  • You don't need to convince us of the utility of user pages or signatures. The fact that they're not there now doesn't mean they won't come. One step at a time. — Russ Cox 14:25, 16 November 2010 (UTC)


On the seqfan list several people asked how signatures could be handled on the new OEIS. Suggestions included:

  • none at all, i.e. figure out the info about who contributed what from the editing history. I think this is awkward, on a sequence with many contributions and/or edits it will be almost impossible to find who added what.
  • continue "by hand" the traditional style, i.e. append "[From NNN (nnn(AT)xxx), MMM DD YYYY]"
  • automatically append "[From NNN (nnn(AT)xxx), MMM DD YYYY]" to any new contribution.
  • use one of the above for contributors who don't have a user account, and the following suggestion for users having an account
  • create a user page for all present and past contributors and use the following suggestion.

If a contributor has a user page, then

  • his email can be completely hidden and nevertheless there is a very comfortable way to send a message to that user
  • signatures can be appended to each contribution with minimally cluttering the appearance of the displayed sequence.

Indeed, one could then opt for (automatically or through some shortcut or macro) appending a minimalistic signature of the form (MFH, 15-Nov-10) with additional title="M. F. Hasler, Monday 15 November 2010, 09:48:00 EST. Click here to contact the contributor." to pop up on hovering the mouse over this link. That signature would take up very little space and yet contain the full information and possibility to contact the author. (The link might also go directly to the "send email" form, but maybe this is not better...)

Advantages are already listed under #User_page_for_all_contributors. Additional advantages include: automatic backlinks, list of User's contributions, possibility to elaborate one one's contributions in another (more detailed way) than the standardized format would allow, ... — M. F. Hasler 12:58, 15 November 2010 (UTC)

Thanks to Maximilan F. Hasler for starting this topic. I'd prefer very strongly to see each contributor's total list of contributions remain searchable, in some form or other. That's what's most important to me. It would be most convenient for me to have a single search mechanism capable of making all of these searches, ideally the same mechanism that we use to search for sequences.
I like the idea for minimizing the signatures and having additional information appear by hovering the mouse over the link. If this turns out not to be feasible for any reason, the signatures needed to make contributions easily identifiable and searchable should, of course, be as minimal as possible. I don't mind appending them myself if necessary.
This is a new format for me; I hope everything looks the way it should. —Matthew Vandermast 04:40, 16 November 2010 (UTC)

Linking from OEIS to OEIS Wiki?

What about adding links to OEIS wiki pages/sections from the main OEIS? Or is the main OEIS the only official part? What is the status of the OEIS Wiki relative to the main OEIS? — Daniel Forgues 04:09, 22 December 2010 (UTC)

JFTR almost seven years later: Some Wiki pages are obviously considered to be good enough for links from OEIS proper, notably the various Index to OEIS pages. Help:Page validation should explain how this works, but at the moment it is one of far too many not-quality pages:
  • What happens after users edited draft? (Guess: nothing, unless it is a critical page like the index.)
  • How can users suggest to review a draft waiting for some time? (Sadly, six years is not exaggerated.)
  • How can users suggest to demote a page to "no review needed"? (E.g., all p-rough pages are ordinary pages, but 11-rough is protected.)
  • How can users suggest to promote a page to the "permitted OEIS link" state? (E.g., IMHO abundant numbers is in relatively good shape after de-spamming by various users.)
  • How can users suggest to protect critical project pages? (E.g., Project:Privacy_policy has a draft, that should be reviewed and protected yesterday, because all Wiki pages link to it at the bottom.)
Frank Ellermann 12:49, 21 October 2017 (UTC)

Format of Search Results

  • Q. The format of results from searches, such as my contributions, has been changed [1] from a brief listing (of A names and sequence terms) to the whole contents of each entry. How do I now get the brief listing to appear? — Jason Kimberley, Dec 23 2010
  • A. At the moment you can't; it's something I need to return to. Russ. (from seqfan reply) —Jason Kimberley 05:40, 6 January 2011 (UTC)
This is a very minor problem in my opinion. In contrast, it is miraculous how well the new OEIS is working! Russ, thank you again. Many people have told me how much they like the new system. N.J.A.S. (from seqfan reply) —Jason Kimberley 23:48, 6 January 2011 (UTC)
Yes, very good work. I was merely thinking there might have been an easy "hack" to get the other format. —Jason Kimberley 23:48, 6 January 2011 (UTC)
This has now been implemented, very nicely! --Jason Kimberley 04:38, 17 September 2011 (UTC)

Helping people to help themselves

I am suffering from the login problem mentioned on this page. I spent some time wandering around the wiki searching for a place to ask about the problem - eventually I found this page linked to from Instructions For Associate Editors, which I started browsing after I gave up my search.

I'd suggest this page should at least be linked to from Help:Contents and OeisWiki:Community Portal, and probably also from Instructions For General Users and/or Main Page.

There are also a number of infrastructure pages with pending edits, many of them among the first pages a user is likely to see - most of them seem most recently to have been approved more than a year ago. Is there anything I can do to help speed up vetting and approval of such edits? Hv 07:42, 29 January 2011 (UTC)

Further to this - I tried the recommended approach for fixing the login problem, and it didn't work for me. I note that if I try to use the new password supplied to log in as "hv" I get "An unexpected error happened while authenticating. Sorry.", whereas if I try to log in as "Hv" I get "The user name and password combination was not recognized.". I wonder if that could be a clue to the underlying problem.
I'm guessing also that this may be related to getting my wiki user set up a long time ago, well before the "brain of two halves" split occurred. Hv 07:58, 29 January 2011 (UTC)

The Unicode/ASCII section doesn't make sense isn't really clear. First it says that non-ASCII characters shouldn't be used, then it is said that UTF-8 is used. But UTF-8 supports all of Unicode! And UTF-8 without non-ASCII characters *is* ASCII, so why even say you're using UTF-8 if you're only using ASCII.

I suspect reason Unicode use is verboten is fear of bad font support..? Neven Sajko 23:25, 22 April 2016 (UTC)

Yes, that's rather obsc. There are actually only six plausible possibilities:
  1. US-ASCII (RFC 20)
  2. ASCII+ISO Latin-1 (adds 96 characters 160..255)
  3. Latin-9 a.k.a Latin-0, eight differences from Latin-1
  4. windows-1252 (adds 27 of 128..159), because HTML5 and WhatWG consider ASCII or Latin-1 with garbage as windows-1252
  5. UTF-8, because it will be the charset soon, 1998+50=2048 predicted in RFC 2277
  6. UTF-4, because it's the same as Latin-1 encoding the rest of Unicode with sequences of 128..159 octets
Latin-9 is rather unlikely, maybe a case of UTF-4 is totally impossible, I "control"  the spec. The HTML5 approach is for consumers (browser developers), not for providers (incl. databases), and the EIS wasn't plagued by Windows idiosyncrasies. US-ASCII and Latin-1 are also out, e.g., oeis:A061091/internal contains "Erdős" with an u+0151.
Therefore the winner is UTF-8, unless there is some plausibility check for undefined codepoints based on an ancient Unicode standard. And colored emojis might be a really bad idea in the OEIS. –Frank Ellermann 00:35, 10 October 2017 (UTC)

Ugh, this issue was already fixed in the FAQ, the Wiki needs some {{resolved}} or {{done}} template. –Frank Ellermann 00:56, 10 October 2017 (UTC)