1525283 - Please make it possible to resume failed downloads

Status: UNCONFIRMED

(Testing :: mozregression, enhancement)

Description

[lilydjwg]

User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:67.0) Gecko/20100101 Firefox/67.0

Steps to reproduce:

run mozregression to find a regression.

Actual results:

I'm in China and my connectivity to archive.mozilla.org is poor. Even with a proxy, downloading such big files (60+MB) usually stalls and then fails. mozregress will wait a long time and then either knows that it has failed and gives up, or thinks a partial downloaded file as complete and tries to unpack it and fails.

It takes a lot of time to re-download the first part of files several times, to wait a lot of time for data that will never come, and to re-test previously-tested versions in a new session.

Expected results:

Please try harder to resume the download when it stalls or fails. Don't treat partial downloaded files as complete. Wait less time before retry (maybe this is already configurable, but using a shorter timeout without resumption is a worse idea).

In short, I hope mozregression can download files like wget -T5 -t100 -c does.