Flag Description

[japoco]

I thought it would be neat to add a field with a description of interesting/noteworthy parts of flags.
For example, the field could say that the bird depicted in the Papua New Guinea flag is a Raggiana bird-of-paradise.

[naveensd101]

Do we have a collection of interesting details that we can populated with?

[Joilence]

Personally I read Wikipedia and summarize my own notes of flag design and perhaps some fun fact. Would be glad to contribute if a PR is wanted. But then I guess there's a problem of fact checking on every note, to prevent spreading of false information.

[axelboc]

Hey that sounds super interesting. However, adding more fields to the main deck always feels a bit scary -- we have to consider translations, long term maintenance (updates/corrections), cluttering of templates (we already have flag similarities and country info), fact checking as mentioned by @Joilence, objectivity of what constitutes "interesting facts", etc.

We've long envisioned Ultimate Geography to become more of an ecosystem that would allow people to compose their own geography deck (fields, templates, translations, etc.) from multiple sources. This would fit right in. Unfortunately, I don't think any of us core maintainers have the time or motivation to champion this massive undertaking.

As a middle ground, we could very much create a new repo on the anki-geo organisation and add you and maybe others as maintainers. The repo could provide a CSV file with the content of this Flag Description field in English (and maybe other languages), one or more templates that make use of this field, and documentation on how to generate the UG deck with that field and those templates added. It'd very much be a DIY deck for power users, but better than nothing? Just a thought.

[Joilence]

Thanks for the rapid and detailed response! I would be glad to maintain a separate repo. I was wondering if you already have a collection of links to wikipedia page of existing flags in UG?

[axelboc]

Hmm, I think sources.csv has links to the image files, but not to the Wikipedia pages for the flags.

---

I forgot to quote this issue where @aplaice, @ohare93 and I have discussed the extensibility of UG. As you can see, adding more fields to the deck is only one of the use cases for extensibility. The hardcore-geography deck started as a way to address the removal of notes for small territories from the main deck. Perhaps it will increase in scope and will eventually become the main "extension" repo.

For now, however, since there are a lot of requests for new fields, we definitely need a way forward.

I propose that we create a new repo, perhaps called sandbox, and that we let people interested (like @Joilence) do what they please with it (as I said, at least: add a CSV file for each new field, a Brain Brew recipe, a template, and some documentation). What do you think @aplaice, @ohare93, et al? This way there's no pressure on long term maintainability, fact checking, etc. — people can just put their content out there without constraints. It might even help to move #603 forward by giving us content and use cases to work with.

[Joilence]

I feel the first step forward is a documentation of how to add a new field, which hopefully inspires more people making {field_name}.csv / media files. It is probably easier for them to do in their own repos, and share the link in relevant issue, or ask to be linked in UG README.

I already started working on my fork to get flag_desc field work, hopefully it could come out soon.

---

The hardcore-geography deck started as a way to address the removal of notes for small territories from the main deck

I think physical entity should definitely be separated from political entity, there's lots of empty card because of them.

It might even help to move #603 forward by giving us content and use cases to work with.

Very much agree. Definitely need enough custom fields / use cases to make the effort in architecture for extensibility worth.

[aplaice]

Hmm, I think sources.csv has links to the image files, but not to the Wikipedia pages for the flags.

I forgot to quote this issue where @aplaice, @ohare93 and I have discussed the extensibility of UG. As you can see, adding more fields to the deck is only one of the use cases for extensibility. The hardcore-geography deck started as a way to address the removal of notes for small territories from the main deck. Perhaps it will increase in scope and will eventually become the main "extension" repo.

Unfortunately (as you're probably aware, but just stating it explicitly), "additional-note extensibility" (what we have in hardcore-geography) and additional-field/additional-card type extensibility (what we'd like here) are not currently equally easy from a technical perspective.

For the current hardcore-geography, there's very little overlap between it and the main repo (the overlap is the note types (HTML/CSS for the notes)) and the country names (but none of the other fields) for a couple of notes (for the countries which are in AUG but without flag/capital), so just duplicating that overlap is not much of a maintainability risk

For flag descriptions, I'm not 100% sure what's exactly envisaged in terms of cards, but my best guess would be that the existing Country - Flag notes would be enhanced with the flag descriptions/fun facts on the back side. This would mean that it'd use the existing content (mainly flags) in AUG to a larger extent, making manual duplication more annoying. Cross-repo imports in BrainBrew would be one solution; there probably are others, but nothing that's particularly convenient atm.

For now, however, since there are a lot of requests for new fields, we definitely need a way forward.

I propose that we create a new repo, perhaps called sandbox, and that we let people interested (like @Joilence) do what they please with it (as I said, at least: add a CSV file for each new field, a Brain Brew recipe, a template, and some documentation). What do you think @aplaice, @ohare93, et al? This way there's no pressure on long term maintainability, fact checking, etc. — people can just put their content out there without constraints. It might even help to move #603 forward by giving us content and use cases to work with.

Yes, I have no objections. My main hesitation is the risk of low "long term maintainability" and hence the risk of the people's efforts being wasted, but naming it a sandbox should hopefully make it clear.

In the longer term, we'll hopefully have a proper solution!

---

I feel the first step forward is a documentation of how to add a new field, which hopefully inspires more people making {field_name}.csv / media files. It is probably easier for them to do in their own repos, and share the link in relevant issue, or ask to be linked in UG README.

Not at all proper documentation, but a brief summary for now, in case it's needed:

1. Add the text to src/data/{field_name}.csv.
2. Add any media to a suitable subfolder in src/media/.
3. Create a new note model in src/note_models/. Firstly, the HTML/CSS for the new cards in src/note_models/templates, and secondly a "description" of how it's supposed to be used in a new .yaml file.
4. A new BrainBrew recipe in recipes/.

Hopefully, looking at the difference between the "main" versions and the "experimental" (interactive) ones should help elucidate what each bit means.

---

Definitely need enough custom fields / use cases to make the effort in architecture for extensibility worth.

True!

[axelboc]

I've created the sandbox repo and added @Joilence as a contributor. Please let me know if anyone else is interested in becoming a maintainer/contributor.