Skip to main content
Bounty Awarded with 50 reputation awarded by Y     e     z
clarified per comments -- I mean one badge, not as many as can be cooked up, as a compromise between "no customization" and "wait, what are all those badges?" from users
Source Link
Monica Cellio
  • 185.2k
  • 66
  • 392
  • 708

This is a proposed implementation based on what I know of the badge system (which does not include knowledge of the source code).

Badges are currently awarded by scripts that run automatically -- sometimes nightly, sometimes every few minutes. I don't think it's going to be feasible to add 130+ special-purpose scripts to that set of jobs, but maybe one new nightly script could be added: the "award site-specific badge" script. What would that script do? It would execute a data query and award badges based on the results.

The moderators of the site would have control of the query, the same way they do with custom close reasons, certain help-center pages, and restricted meta tags like "faq". As with these other things, it would be up to the community to reach consensus on what the badge scripts should be; moderators are approvers, not sole authors. (Moderators are not required to be SEDE experts.) I am proposing a single custom badge per site, not "as many as you can somehow jam into that query".

I'm suggesting a query because, with SEDE, anybody (who learns SEDE) can develop and test candidate badge scripts. Part of putting a custom badge in place would be writing, testing, and showing preliminary results -- if we had this badge, what would its awards be?

Constraining the badge by what can be expressed through SEDE places reasonable bounds on it. It also prevents possible privacy violations; for example, no badges based on voting patterns would be possible. To the specific example in this question, you'll need to be able to detect the Balsamiq mockup in the post body; the post body is available, so you should be able to do that.

If a query is too complex and it times out, no badges for you -- it's up to sites to keep these to a reasonable load (whatever SE determines is reasonable).

To minimize UI impact (profile, notifications, badge pages, etc) and user confusion ("what the heck is this 'salad dressing' badge?"), I suggest naming the badge "(sitename) custom badge", with the description containing a link to the meta post that explains the badge. The submission from the moderators could include a proposed "short description" for the tooltip.

Finally, the badge would be bronze (no matter how easy or hard it is to earn) and there would only be one; this isn't an opening for site-specific badge hierarchies.

This is a proposed implementation based on what I know of the badge system (which does not include knowledge of the source code).

Badges are currently awarded by scripts that run automatically -- sometimes nightly, sometimes every few minutes. I don't think it's going to be feasible to add 130+ special-purpose scripts to that set of jobs, but maybe one new nightly script could be added: the "award site-specific badge" script. What would that script do? It would execute a data query and award badges based on the results.

The moderators of the site would have control of the query, the same way they do with custom close reasons, certain help-center pages, and restricted meta tags like "faq". As with these other things, it would be up to the community to reach consensus on what the badge scripts should be; moderators are approvers, not sole authors. (Moderators are not required to be SEDE experts.)

I'm suggesting a query because, with SEDE, anybody (who learns SEDE) can develop and test candidate badge scripts. Part of putting a custom badge in place would be writing, testing, and showing preliminary results -- if we had this badge, what would its awards be?

Constraining the badge by what can be expressed through SEDE places reasonable bounds on it. It also prevents possible privacy violations; for example, no badges based on voting patterns would be possible. To the specific example in this question, you'll need to be able to detect the Balsamiq mockup in the post body; the post body is available, so you should be able to do that.

If a query is too complex and it times out, no badges for you -- it's up to sites to keep these to a reasonable load (whatever SE determines is reasonable).

To minimize UI impact (profile, notifications, badge pages, etc) and user confusion ("what the heck is this 'salad dressing' badge?"), I suggest naming the badge "(sitename) custom badge", with the description containing a link to the meta post that explains the badge. The submission from the moderators could include a proposed "short description" for the tooltip.

Finally, the badge would be bronze (no matter how easy or hard it is to earn) and there would only be one; this isn't an opening for site-specific badge hierarchies.

This is a proposed implementation based on what I know of the badge system (which does not include knowledge of the source code).

Badges are currently awarded by scripts that run automatically -- sometimes nightly, sometimes every few minutes. I don't think it's going to be feasible to add 130+ special-purpose scripts to that set of jobs, but maybe one new nightly script could be added: the "award site-specific badge" script. What would that script do? It would execute a data query and award badges based on the results.

The moderators of the site would have control of the query, the same way they do with custom close reasons, certain help-center pages, and restricted meta tags like "faq". As with these other things, it would be up to the community to reach consensus on what the badge scripts should be; moderators are approvers, not sole authors. (Moderators are not required to be SEDE experts.) I am proposing a single custom badge per site, not "as many as you can somehow jam into that query".

I'm suggesting a query because, with SEDE, anybody (who learns SEDE) can develop and test candidate badge scripts. Part of putting a custom badge in place would be writing, testing, and showing preliminary results -- if we had this badge, what would its awards be?

Constraining the badge by what can be expressed through SEDE places reasonable bounds on it. It also prevents possible privacy violations; for example, no badges based on voting patterns would be possible. To the specific example in this question, you'll need to be able to detect the Balsamiq mockup in the post body; the post body is available, so you should be able to do that.

If a query is too complex and it times out, no badges for you -- it's up to sites to keep these to a reasonable load (whatever SE determines is reasonable).

To minimize UI impact (profile, notifications, badge pages, etc) and user confusion ("what the heck is this 'salad dressing' badge?"), I suggest naming the badge "(sitename) custom badge", with the description containing a link to the meta post that explains the badge. The submission from the moderators could include a proposed "short description" for the tooltip.

Finally, the badge would be bronze (no matter how easy or hard it is to earn) and there would only be one; this isn't an opening for site-specific badge hierarchies.

Source Link
Monica Cellio
  • 185.2k
  • 66
  • 392
  • 708

This is a proposed implementation based on what I know of the badge system (which does not include knowledge of the source code).

Badges are currently awarded by scripts that run automatically -- sometimes nightly, sometimes every few minutes. I don't think it's going to be feasible to add 130+ special-purpose scripts to that set of jobs, but maybe one new nightly script could be added: the "award site-specific badge" script. What would that script do? It would execute a data query and award badges based on the results.

The moderators of the site would have control of the query, the same way they do with custom close reasons, certain help-center pages, and restricted meta tags like "faq". As with these other things, it would be up to the community to reach consensus on what the badge scripts should be; moderators are approvers, not sole authors. (Moderators are not required to be SEDE experts.)

I'm suggesting a query because, with SEDE, anybody (who learns SEDE) can develop and test candidate badge scripts. Part of putting a custom badge in place would be writing, testing, and showing preliminary results -- if we had this badge, what would its awards be?

Constraining the badge by what can be expressed through SEDE places reasonable bounds on it. It also prevents possible privacy violations; for example, no badges based on voting patterns would be possible. To the specific example in this question, you'll need to be able to detect the Balsamiq mockup in the post body; the post body is available, so you should be able to do that.

If a query is too complex and it times out, no badges for you -- it's up to sites to keep these to a reasonable load (whatever SE determines is reasonable).

To minimize UI impact (profile, notifications, badge pages, etc) and user confusion ("what the heck is this 'salad dressing' badge?"), I suggest naming the badge "(sitename) custom badge", with the description containing a link to the meta post that explains the badge. The submission from the moderators could include a proposed "short description" for the tooltip.

Finally, the badge would be bronze (no matter how easy or hard it is to earn) and there would only be one; this isn't an opening for site-specific badge hierarchies.