Re: Re: [PATCH v5] builtin/clone.c: add --reject-shallow option

Hi,

On Mon, 1 Mar 2021, lilinchao@oschina.cn wrote:

quoted

quoted

@@ -1440,6 +1444,8 @@ static void receive_shallow_info(struct fetch_pack_args *args,

* shallow. In v0, remote refs that reach these objects are
* rejected (unless --update-shallow is set); do the same.
*/
+	if (args->remote_shallow)
+	die("source repository is shallow, reject to clone.");

I just found that Johannes Schindelin wrote a document 14 year ago
in Documentation/technical/shallow.txt:

"There are some unfinished ends of the whole shallow business:

A special handling of a shallow upstream is needed. At some stage,
upload-pack has to check if it sends a shallow commit, and it should
send that information early (or fail, if the client does not support 
shallow repositories). There is no support at all for this in this patch
series."

Oh wow, what a blast from the past.

I do agree that your patch is an improvement over the current situation. 

Thanks. Glad to hear the voice from you, hope I can get some suggestions
from you too, especially on the transport part.

Thanks,
Johannes

quoted

It seems that my patch can sovle his worry in some degree,
and maybe we could warn client in fetch-pack stage, if we don't
choose to reject shallow cloning.

if (args->remote_shallow)
die("source repository is shallow, reject to clone.");
else
warning("remote source repository is shallow.");