Message-ID: <419CAB58.3050905@panoramicfeedback.com>
Date: Thu, 18 Nov 2004 09:02:00 -0500
From: Aaron Bentley <abentley@panoramicfeedback.com>
User-Agent: Mozilla Thunderbird 0.5 (X11/20040309)
MIME-Version: 1.0
To: snogglethorpe@gmail.com, miles@gnu.org
Subject: Re: [Gnu-arch-users] cacherev policy script
References: <pan.2004.11.18.13.23.45.365921@progsoc.org>
	<fc339e4a04111805243d86056c@mail.gmail.com>
In-Reply-To: <fc339e4a04111805243d86056c@mail.gmail.com>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Cc: gnu-arch-users@gnu.org, Anand Kumria <wildfire@progsoc.org>
Precedence: list
Sender: gnu-arch-users-bounces+migo=homemail.com@gnu.org
Errors-To: gnu-arch-users-bounces+migo=homemail.com@gnu.org
Status: RO
Content-Length: 1065
Lines: 33
Miles Bader wrote:
>>it talks says an optimal cacherev policy would be "cached revisions such
>>that the sum of the sizes of changesets between any two cached revisions
>>is about the same as the size as the [...] cached revision."
> 
> 
> Not really "optimal" for all purposes -- e.g., in my emacs branch,
> there are currently over 700 revisions, but the total size of all
> those changesets is still less than a single emacs cacherev... 
> However for purposes of not driving people doing virgin checkouts
> nuts, more frequent cacherevs are called for.

Agreed: this varies by tree, internet connection properties and 
changeset application time.

For me, with the tla tree, downloading and untarring a cacherev costs 
about as much as downloading and applying 50 changesets.

Aaron

-- 
Aaron Bentley
Director of Technology
Panometrics, Inc.


_______________________________________________
Gnu-arch-users mailing list
Gnu-arch-users@gnu.org
http://lists.gnu.org/mailman/listinfo/gnu-arch-users

GNU arch home page:
http://savannah.gnu.org/projects/gnu-arch/