| layout |
|---|
default |
Once files have been added to this repository, the files you see on GitHub or when the repository is cloned will be LFS pointer files .
A question is, how do you specify a download link to the actual data files?
First, anyone cloning the repo will get only LFS pointer files and not the actual
data files. To get to the actual data files in a cloned repository, a
git lfs pull operation is needed. But, BE CAREFUL!! An unqualified git lfs pull
operation downloads every LFS file in the repo which could take hours. If you are
interested in only specific files, use the --include option to git lfs pull to
identify the specific file(s) you want to fully download.
When writing emails or documentation which includes links to files to be downloaded, cloning and git lfs is not used. Instead, we we provide HTTPs access to those files through the website that is hosted from this repository using GitHub magic URL to access their raw contents.
The download path for
aneurysm_tutorial_data.tar.gz
(not the LFS pointer file),
for example, will look something like...
https://github.com/visit-dav/largedata/blob/master/bindata/aneurysm_tutorial_data.tar.gz?raw=true
You can use this link anywhere including in email to give users the link to get that specific data file. Users do not require a GitHub account in order to access data through this link.
Better still, give them a link to the landing page for the file. For the aneurysm data, it will look something like...
https://visit-dav.github.io/largedata/datarchives/aneurysm
The last entry in the URL, aneurysm in this example, is whatever the name of the
markdown file hosting information about the data is without the extension.
The landing page link takes users to a page with more information about the file
including maybe images as well as all the download format options
(e.g. .tar.xz, .tar.gz, .zip, etc.) and their integrity checks.