<br><br><div class="gmail_quote">On Wed, Apr 15, 2009 at 4:59 PM, Matthew D. Fuller <span dir="ltr"><<a href="mailto:fullermd@over-yonder.net">fullermd@over-yonder.net</a>></span> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
On Thu, Apr 16, 2009 at 09:40:46AM +1000 I heard the voice of<br>
Robert Collins, and lo! it spake thus:<br>
<div class="im">><br>
> > 1. checkouts should be lightweight by default (ala svn)<br>
><br>
</div>> There are serious performance implications in this. [...]<br>
<div class="im">> lightweight checkouts are really only suitable for high bandwidth<br>
> low latency environments.<br>
<br>
</div>A possible solution is to default checkouts of file:/// to be light,<br>
and others heavy. The obvious downside is the added DWIMitude.</blockquote><div><br>I like your previous idea of calling them both checkouts (or whatever) with a with-or-without flag for ancestry.<br><br>But I do not think the meaning of the command or type of branch created should depend on the type of transport. <br>
That forces me to think too much for a simple operation.<br><br>-M<br></div></div>