2 ms·
They are talking about the standard “Download zip”, which can also be done for commits individually. They aren’t precomputed, which causes the load.
by kassner 1y ago
They are talking about the standard “Download zip”, which can also be done for commits individually. They aren’t precomputed, which causes the load.
- bob1029 1y agoThey don't need to be pre-computed or stored. You can directly stream stdout from git archive to the HTTP response. This is the default behavior of git archive.
- loloquwowndueo 1y agoAnd git archive uses zero cpu therefore it can handle thousands of archive requests per second. Right? Right??
- bob1029 1y agoYes, your CPU should be nearly idle. Git archive is I/O bound.