Empowering Italian GLAMs/Tech evaluations

Technical documentation about the tool related to Empowering Italian GLAMs.

Please read also phabricator:T320369 ("Help evaluating Toolforge or VPS to empower Italian GLAMs").

Toolforge administrators

https://toolsadmin.wikimedia.org/tools/id/glams

Toolforge domain

Our Toolforge domain is very short. If you think this should not have happened, please contact us. Anyway this was more or less our attempts:

• https://musei-su-wikipedia.toolforge.org/

  Problem: this username is forbidden by Toolforge since it contains "wikipedia"

• https://musei-su-itwiki.toolforge.org/

  Opinion: too long and "itwiki" is too weird (by iopensa)

• https://empowering-italian-glams.toolforge.org

  Opinion: too long and weird (by iopensa)

• https://italian-glam.toolforge/

  Opinion: too long and weird (by iopensa)

• https://glams.toolforge.org/

  Opinion: weird but very short: wooow cute! (by iopensa)

  Opinion: ASD (by Valerio Bozzolan)

In short, we picked "glams" since "musei-su-wikipedia" was not available. If you're upset about "glams," I'm sorry. Domains have always been undemocratic.

Requirements

Monitorato su Phabricator
Task T320369

Considerations

• phabricator:T320369 - Help evaluating Toolforge or VPS to empower Italian GLAMs (Wikimedia IT)

Command line access

Should be not required.

User storage needed

40G (1000 musei, 20 foto, 2M/foto)

Dependencies

• PHP (executes Python commands)
• MySQL database
• Python

Available servers

• Cloud VPS - not needed

  m:wikitech:Wikitech:Cloud Services Terms of use

• server from current provider 💼 vh - average cost max. 400€ / server, automated backups already included
• lessema: enough storage but it's dedicated to CiviCRM - not enough separation
• beta: not owned by WMIT
• delta: not owned by WMIT
• epsilon: not owned by WMIT
• horror: dedicated to backups - not enough separation
• intreccio: not enough storage
• manoscritto: not enough storage - not enough separation with ZNUNY
• toolforge: not enough storage - https://t.me/wmcloudirc/44175 (this comment was super-seeded)
• Toolforge: OK - phabricator:T320369

FAQs

About unclear copyright status files on Wikimedia Cloud servers

<bozzy> there is a person who verifies a scanned piece of paper in which a museum manager authorizes the release in free license, before uploading on Commons.
<bozzy> This way the Commons community does not receive an avalanche of problematic data. We felt it was more respectful to them.
[29/08/22 17:10]
<taavi> you can't store material with unclear copyright status on WMCS either..

Answer: This has since been clarified, the materials are released by the museums in bona-fide with a free license, the problem is that the museum could be not so skilled about copyright, so it's better to double-check before uploading them to Wikimedia Commons.

About uploading directly on Commons

<bozzy> The tool will be a sort of "image uploader". Kind of "pre-Commons" for a GLAM-related project in a specific country.
[29/08/22 17:00]
<bd808_> Why not upload to commons and then mark the "unuseful" things for deletion? Commons will have more storage and better performance than almost any Cloud VPS project.
<bd808> https://commons.wikimedia.org/wiki/Commons:Volunteer_Response_Team is exactly for things like reviewing and validating image releases as well. I'm not saying I know enough to say that there is no need for a staging process, but this all sounds like things that the community knows how to deal with.

Answer: it's too difficult to train 3000 museums to upload something in Wikimedia Commons without doing mistakes. The community should not fix our import mistakes. It's better to do things in the good way, with a simple and dedicated wizard, oriented for our goals.

Repository

https://gitlab.wikimedia.org/repos/wikimedia-it/wmit-empowering-italian-glams

Tech Team

Original author

• Utente:Adert (consultant)

Code sprint and code review

• Utente:Mte90 (consultant)

Repository maintainer

• Utente:Valerio Bozzolan (volunteer)