Update data via API vs update data via MediaWiki

[sacdallago]

(Sorry E if I use your space but this was where it made more sense to me)

Everyone @Rostlab/js_cs_sose_2016_students : There is a discussion going on here about where scraped data (!!! ONLY) needs to be updated. The options are:

1. Via API.
2. Via MediaWiki + running the scrapers again.

The implications are respectively:

1. The API opens for POST requests to directly manipulate scraped data.
2. The API doesn't allow modifying scraped data.

I'm all in for number 2 for the following reasons:

• People cloning A's repo want to have the same data as their friends who cloned the repo 2 months earlier
• Updating the data via GUI is the preferred method for normal human beings
• Databases can easily be rebuilt and updated by re-scraping the data off the wiki, without worrying about concurrency (someone updated the data via API call, now someone else wants to update the data via re-scraping... What is right?)
• A routine can be set up that will update the database periodically by re-scraping the data.

If you are against implementing option 2 please comment here within 2 days, otherwise I take it as everyone knows what's going on and is fine with that.

[sacdallago]

@kordianbruck @Adiolis OK. So it's final:

Updates to scraped data only via update of wikis. / Close POST calls for modifications on the scraped data

[kordianbruck]

Why are you closing this @sacdallago - did you remove the already existing update methods from the api?

[kordianbruck]

Ah, wupps, this is in Project E, not A...