search

CollegeOfTheHolyCross / dta_sufia

1 stars 0 forks source link

Compound objects? (already fixed?) #102

Open kjrawson opened 7 years ago

kjrawson commented 7 years ago

Hey, Steven. Our lab has a number of two-sided images they are processing (mostly photos with writing on the back). At the moment, they are in the format of two separate JPEGs. To include both of those objects in the same record, they would need to combine them into a single PDF. Do you want to talk about the possibility of compound (i.e., multiple files attached to a single record) objects? I don't have a sense of how much work this would involve. Thanks!

scande3 commented 7 years ago

I'll have a response to this inquiry before Monday. If it isn't that complicated, will implement something for it. If it is more complicated, will outline what it would take.

Quick question: What is the scope of the use case being envisioned? Essentially should the scope of this be support for any number of images (ie. large books and such) or just allow an object to have up to a small X number of maximum files?

kjrawson commented 7 years ago

All of the examples we have now are two-sided objects, so there would be two files. Even if I am stretching my thinking, we won't need support for anything beyond 6 (somewhat arbitrary number) for now. I can imagine scrapbooks or objects down the line that would increase the scope, but we don't actually have anything like that right now. Thanks, Steven!

scande3 commented 7 years ago

After research, there isn't any easy way to support this. Options are:

  1. "The Dirty Approach". Essentially hard-code the maximum amount of files that can be on an object and shoehorn it in. Between the backend code and updating the front-end code for multiple files, this would be approximately ~5 hours of work.

  2. "The Better Approach": Allow for any number of files but they are unordered. This change would be about ~10 hours of work.

  3. "The Ordered Better Approach": Allow for any number of files but they are ordered. This change would be about ~14 hours of work.

  4. "The Upgrade Sufia Approach": Later versions of Sufia can handle compound objects. However, this would require a bunch of work of around ~40 hours.

Let me know if any of the above is worth it. For number 1, I could try to have it implemented by Friday morning. For others, it would take a few weeks (essentially around total number of hours / 5 hours per week).

kjrawson commented 7 years ago

Thank you for researching this and providing such thorough options––this is immensely helpful in making a decision about how to proceed. I am thinking that since I should hear back from grants within the next few months, we might use that to help guide our decisions. If we get an influx of funding and you are game, we'll go for option 4, which seems like the best approach. In the meantime, we'll put this on hold since it would impact so few items. Thank you!

scande3 commented 7 years ago

Alright, let me know if I should do any progress on this in the future then. Will move on to other issues!

kjrawson commented 6 years ago

Has this been addressed in the new version?