[seqfan] Re: A suggestion regarding b-file uploads and edit quotas.

[Peter Munn]

Dear Neil, Antti and seqfans,

Regarding control of b-file uploading, might I suggest a keyword, for
instance "provisional" or "tentative", that is added automatically to new
sequences and removed by an editor or automatically on approval. This
would allow for automated checks in the future, such as not approving a
sequence that had a reference to a "tentative" sequence.

I personally welcome the ideas regarding the future of quotas. For
example, I know quota concerns contribute to my never having added a
sequence to the indexes.

I favour the longer-term outcome being an extra quota for minor edits.
This would allow control according to whether the feature was being used
in the interest of the community.

For instance, someone might be spending hours a week finding highly
relevant CROSSREFs to add. Another's 60 new CROSSREFs might show no more
rationale than all pointing to sequences to which they have just added
links to their latest publication.

I can see this being implemented by a user clicking on "minor edit"
instead of "edit" to start editing, with guidelines coming up about its
use. Automated controls, such as not allowing for core sequences, could be
added later.

Best Regards,

Peter

>> On Sat, Mar 9, 2019 at 12:46 PM Antti Karttunen
>> <antti.karttunen at gmail.com> wrote:

>>> BTW, the rule that one should not upload a b-file before the sequence
has been approved, is not (IMHO) only stupid, but also even
pernicious.
>>> What I suggest, how it should be changed:
>>> If at the editing process and after a few PinkBox-comments, none of
the editors is against the sequence (or nobody finds that it is an
erroneous duplicate of some existing sequence), then it should be
ALRIGHT for some of the editors (not the submitter himself) to upload a
b-file for that sequence. This especially if the sequence is from a
quality-submitter, who is known to submit approvable sequences. On the
other hand, with certain contributors, who often submit sequences
>>> with erroneous terms (because hand-computed?), but whose sequences
still are mostly sound and approved, the process of creating a PARI-
>>> or some other
>>> program, and then uploading a b-file (if the sequence is not wholly
off the wall) works as an essential error checking process, where at least
any erroneous terms will be caught, because the server software checks the
b-file data to the existing data section terms.
[...]
>>> One alternative I have played with, but which would require non-simple
changes at the server-software, would be that there would be two
separate quotas for each user: one, generally smaller one, for
creating new sequences (and maybe larger edits to existing sequences), and
one, larger or limitless, for uploading b-files and adding
look-keywords to existing sequences, and such. I guess that now people
with a very small quota, like 3 or 5, will often avoid uploading
b-files to their sequences later, because each such upload eats the quota,
which they prefer to use to submit their new ideas.
> A simpler idea: double everybody's quota, then make the creation of each
new sequence to consume 2 points from it, and an edit of any existing
sequence to consume just one point. That should be quite easy to
implement, right? (One could tune the ratio. In the de-luxe model it could
be even user-specific?)