semi-voluntary moratorium and helping to edit.

[Antti Karttunen]

Alonso Del Arte wrote:

> Max,
>  
> Neil still opposes the idea of a Wiki engine for the OEIS, and for 
> what's it's worth, so do I. Wikipedia is very useful to me, but when 
> it comes to say, copying and pasting a 47-digit prime, I can be almost 
> 100% sure it's correct if I'm getting it from the OEIS. With 
> Wikipedia, even if I look at the edit history, I will still want to 
> have Mathematica doublecheck it for me.
>  
> There are several kinds of vandals on Wikipedia. The most obvious 
> vandal will put in irrelevant obscenities in an article, and that sort 
> of thing gets reverted almost instantaneously. But there's also a 
> subtler kind of vandal who might add an extra 0 to 6210001000 in the 
> article on self-descriptive numbers. That kind of vandalism could 
> stand for months.
>  

In any case, a potential future "WikiOEIS" would need its own custom 
Wikiengine
(or similar system). So it could be well customized to have various 
degrees on
"how easy it is to make changes", both "horizontally" across the entries 
(i.e. core-sequences would take
a more thorough reviewing round) and "vertically" across the fields, 
i.e. the sequence's terms (%S, %T, %U nowadays)
would be "harder" to change (or in some cases even "locked"/"frozen" 
from any further changes)
than say "Comment" or "See also" fields.

Also, many sequences could be generated easily (that is, 
"computationally easily")
on-the-fly, which would give some redundancy.


> Your proposal to have "trusted" and "untrusted" users might be one way 
> around this problem. But the group of "trusted" users would have to be 
> very small, e.g., those with Sloane number 0 or 1.
>
Yes, I agree.

Yours,

Antti

> Alonso
>  
> On 10/1/05, *Max* <relf at unn.ac.ru <mailto:relf at unn.ac.ru>> wrote:
>
>     Alexandre Wajnberg wrote:
>
>     > What about the possibility given to the submitters (with a special
>     > access code, or with a "recognition process" based on their IP
>     adresse,
>     > or something like that) to correct directly themselves their
>     submissions
>     > (and *their* submissions only), doubled by the automatic sending
>     of a
>     > copy of the job done to Neil?
>     > (Anyway, a "review" of our job (even the easiest ones) by other
>     > mathematicians will be allways necessary, but it would lighten
>     Neil's work).
>
>     Sorry to say that again but WiKi-driven engine would help a lot.
>
>     Basic idea is to have two databases: "release" (or "main") and
>     "development".
>
>     The main database will be editable by a set of trusted people
>     (maintainers) and will be read only for the others. The main
>     database will provide the same services as the current OEIS.
>
>     In opposite, entries in the development database will be allowed
>     to be modified by anybody. That will create a queue of
>     modifications that trusted people will review and accept/reject to
>     the main database.
>
>     Everything should be transparent and easy to use. For example,
>     entries in the main database will have [Edit] button but if it is
>     hit by "untrusted" person he will be automatically forwarded to
>     the corresponding "development" database entry.
>
>     Initially databases will be synchronized and a goal of the trusted
>     people will be to maintain this synchronization, i.e., every
>     modified entry in the development database sooner or later will be
>     either accepted into the main database, or rejected and its
>     original from the main database will be copied back to the
>     development one.
>
>     Max
>
>