Why do CT entries return a 'Content-Disposition: attachment; filename=f.txt' header?

[Miguel Jacq]

Hello,

I wanted to inquire as to whether there was a reason why requests to /ct/v1/get-entries, for instance https://ct.googleapis.com/pilot/ct/v1/get-entries?start=7157327&end=7157327, return a 'Content-Disposition: attachment; filename="f.txt"' header? Is there a reason this is forced?

For context, I am using the publicly available monitor https://ctwatch.net, which exposes results as an Atom feed, and I wanted to programatically parse the json/decode the MerkleTreeLeaf structure to obtain certain attributes. It would be so much nicer if the URL above just spat back the json like a typical API rather than force an attachment to download, which then requires more manual work to decode.

I scanned the IETF spec to see if there was something about attachments in there, but found nothing.

Thanks

[Miguel Jacq]

Please ignore this :) for various reasons I was testing on a version of curl that was too old. And distracted by the fact that the browser, too, always would download rather than display the JSON.

I see a normal curl request works fine, so I can make this work.

Thanks