Download a full dataset #60

New issue

Closed

opened 2026-04-30 08:23:15 +00:00 by HenrikeW · 3 comments

HenrikeW commented

2026-04-30 08:23:15 +00:00

(Migrated from github.com)

Please describe the feature

As a medical expert I would like to download a full dataset in one click
so that I don't need to think about single files when getting access to a dataset

Detailed description

the button "Download dataset" on the dataset details page starts the download of a dataset
Nice to haves depending on how the browser handles the download.
- there is information on how long the download will take approximately
There is an informative message with number of files and total size to be downloaded where the user chooses to proceed or cancel.

Additional context

Questions:

Is there a specific API-endpoint that can be used for the download of a full dataset? Or do we use the downloadUrls provided for every single file?
A: The API provides single file downloads, so this will built upon #36.
Is an estimated download time needed/ possible?
A: See if this information is available before the user chooses to proceed and the browser takes over the download.

## Please describe the feature  As a medical expert I would like to download a full dataset in one click so that I don't need to think about single files when getting access to a dataset ## Detailed description  - [x] the button "Download dataset" on the dataset details page starts the download of a dataset - Nice to haves depending on how the browser handles the download. - [x] there is information on how long the download will take approximately - [x] There is an informative message with number of files and total size to be downloaded where the user chooses to proceed or cancel. ## Additional context Questions: - Is there a specific API-endpoint that can be used for the download of a full dataset? Or do we use the downloadUrls provided for every single file? A: The API provides single file downloads, so this will built upon #36. - Is an estimated download time needed/ possible? A: See if this information is available before the user chooses to proceed and the browser takes over the download.

pontus commented

2026-04-30 11:25:50 +00:00

(Migrated from github.com)

I'm fairly certain there is no API endpoint for a full dataset download.

In several way, I feel limited by not knowing what functionality modern browsers enable (without needing to manually enable features before loading).

That means I don't know what information can be provided to the user, but likely good candidates:

Downloaded number of files/total number of files. Downloaded number of bytes/total number of bytes.

If it's available: current download speed would be nice to show.

I'm fairly certain there is no API endpoint for a full dataset download. In several way, I feel limited by not knowing what functionality modern browsers enable (without needing to manually enable features before loading). That means I don't know what information can be provided to the user, but likely good candidates: Downloaded number of files/total number of files. Downloaded number of bytes/total number of bytes. If it's available: current download speed would be nice to show.

pontus commented

2026-04-30 11:32:36 +00:00

(Migrated from github.com)

I think this belongs in this user story but don't know if it's best in another or in the same card, putting here for now.

If the APIs available allows for it, it would be very desirable for a dataset download to be able to resume a failed download - e.g. in a dataset with 7000 objects where the download failed after 3000, it would be very nice if it could skip those after basic verification such as e.g. size matches and only download the remaining 4000.

Not necessarily MVP but desirable to keep in mind during design.

I think this belongs in this user story but don't know if it's best in another or in the same card, putting here for now. If the APIs available allows for it, it would be very desirable for a dataset download to be able to resume a failed download - e.g. in a dataset with 7000 objects where the download failed after 3000, it would be *very* nice if it could skip those after basic verification such as e.g. size matches and only download the remaining 4000. Not necessarily MVP but desirable to keep in mind during design.

aaperis commented

2026-09-24 12:18:33 +00:00

(Migrated from github.com)