[PATCH 3 of 5] pager: skip uisetup if chg is detected

Jun Wu quark at fb.com
Mon Mar 14 16:49:52 EDT 2016

On 03/14/2016 08:37 PM, Sean Farley wrote:
> Would this be reason enough to put pager in core so we wouldn't need
> different implementations?

The reason that chg needs a different implementation is the pager must be
executed by the client. Otherwise it's likely to have trouble with
"terminal foreground job", Ctrl+Z etc.

To be more precise, the pager is currently using subprocess.Popen to create
process at server side. chg needs it to be "remotely" executed at client

I think Yuya has plans to unify the implementations somehow. See

More information about the Mercurial-devel mailing list