On Wed, 15 Oct 2025, [email protected] wrote:
What exactly what that be useful for except for your quite particular niche
use case?
I agree it's a niche use case and if the point of curl is to transfer file
(as opposed to troubleshoot HTTP services; although that's mostly what I use
it for), it might not be a suitable feature.
I would propose you do it like this, which uses already existing
functionality:
$ curl URL1 URL2 URL1
... and you have the server side delay the URL2 response a suitable time. If
the server kept the connection to the first URL1 alive, the second URL1
transfer will reuse the connection fine. If it didn't, the second URL1 will
create a new connection.
--
/ daniel.haxx.se || https://rock-solid.curl.dev
--
Unsubscribe: https://lists.haxx.se/mailman/listinfo/curl-users
Etiquette: https://curl.se/mail/etiquette.html
