Missing maps

[almontg]

Not all the maps from the American Revolutionary War-Era Maps Collection of Distinction made it over to argomaps.org; mostly those whose dates fall outside the 1740-1800 range but which should be included anyway. Happy to work up a full spreadsheet of effected maps, just need to know who to give it to!

I'd also like to bring in several other maps which are in the Leventhal/Digital Commonwealth collection and in the target time period but were not part of the collection of distinction.

[garrettdashnelson]

@almontg can you give the me commonwealth ID of one example? And are the maps affected in such a manner in the Airtable you were working with for initial tagging, or missing there too?

This is one clear example of what I mean when I describe why I am so wary of having multiple sources of truth and derived, unlinked datasets. The data pipeline from the repository ought to be more procedurally defined to avoid issues like this.

The issue of maps in LMEC digital collections but not yet identified as belonging to AmRev Collection of Distinction is a separate one that @lauremic and I have previously talked about. @almontg if you can draw up a list of identifiers of objects that you think should be included, we can handle that on the repository side.

[almontg]

@garrettdashnelson here are a couple examples below; seems to be related entirely to using dates to filter which in most cases was a good call (some of the filtered out maps were facsimilies that shouldn't be in ARGO anyway, or are way too early) but there are a couple of edge cases in the decade on either side of our cutoff I'd like to include; I have a full spreadsheet of all such effected maps with my call on whether or not they should be included that I will send over. All these maps were included in the airtable.

• commonwealth:hx11xz975
• commonwealth:hx11z459z
• commonwealth:6108vv70j
• commonwealth:hx11z457d

Currently working on a list of maps already in Digital Commonwealth which I'd like to see included; will send when it's done!

[almontg]

Full spreadsheet sent to Lauren and Garrett

[garrettdashnelson]

@almontg I think this is a @diedrick question, not a me and Lauren one. For instance, I don't see any of these Commonwealth IDs listed in maps_extended.json, which is ultimately what controls which maps get built into the ARGO site. I'm not sure what the derivation of that maps_extended.json data file is— @diedrick you'd be the one to answer this question.

[diedrick]

We skip maps that aren't in the date range, which at least the first two aren't -- can we confirm the maps are between 1740 and 1800?

[almontg]

@diedrick they are out of the date range (just by a few years) but should still be included for various reasons; only some of them need to be included. I can send you the spreadsheet of what needs to be brought in and what doesn't.

[diedrick]

Ok, easier to just paste the list here as solr IDs only.

[almontg]

Is the solr ID the same as the commonwealth id?

[diedrick]

Yep, solr ID is in the format of commonweath:xxxxxxxxx.

[garrettdashnelson]

@almontg What do you think about simply including everything from the AmRev Collection of Distinction in ARGO? This is what we'd said we would do in the grant proposal, and it's probably what my bias would be. It would also mean that we don't have to keep track of any data filtering since ARGO and the AmRev Collection of Distinction in the repository would be 1:1 matches of one another. I suppose there are a very small handful of items that might be slightly confusing in ARGO, but that seems to me like a relatively minor concern?

[almontg]

@garrettdashnelson we can, but I welcome the chance to do some pruning here — three of the maps, for example, are facsimilies from the Leventhal collection, which I really don't think should be part of the project, and two more are facsimilies from the Newberry. Most of the others are fantastic maps from other institutions but date as early as the seventeenth century and as late as the mid-nineteenth. Ultimately, however, we're just talking about a relatively small number of maps here; 44 total, with 22 I think we should definitely include. If you think it would be that much of a hassle to not include some of them I'm ok with including them, I just don't love the idea of, say, potentially throwing off the timeline feature by that much (would make for a span of 1670-1973).

[garrettdashnelson]

@diedrick is there a way of just filtering them out in the handling of the timeline module, but still have them included in the broader corpus? A more radical solution would be submitting a ticket to remove them from Digital Commonwealth, though I kind of lean against that. I do like the idea of the two collections having a 1:1 correspondence, particularly so that future collections managers aren't confused about why there is a slight mismatch.

[almontg]

@garrettdashnelson I would actually err on the side of just removing them from the AmRev collection of distinction if keeping that 1:1 is crucial; they can of course stay on Digital Commonwealth, maybe under a different collection. They just woudn't be part of the collection that is brought into ARGO.

I can see adding an additional filtering layer to the ARGO site causing issues down the road and I'd rather avoid having to treat specific maps with special rules in that way, especially if they're not contributing to the broader 1740-1800 story we're supporting.