Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 13 Jul 2000 00:42:29 +0100
From:      Ben Smithurst <ben@FreeBSD.org>
To:        Satoshi - Ports Wraith - Asami <asami@FreeBSD.org>
Cc:        Dag-Erling Smorgrav <des@flood.ping.uio.no>, Will Andrews <andrews@technologist.com>, ports@FreeBSD.org
Subject:   Re: fetch(1) timeout
Message-ID:  <20000713004229.Z11000@strontium.scientia.demon.co.uk>
In-Reply-To: <vqcaefnaqt8.fsf@silvia.hip.berkeley.edu>
References:  <20000712061459.D1690@argon.gryphonsoft.com> <xzpitub63in.fsf@flood.ping.uio.no> <20000712064405.E1690@argon.gryphonsoft.com> <xzpem4z62n0.fsf@flood.ping.uio.no> <20000712065724.F1690@argon.gryphonsoft.com> <20000712131725.A23242@mithrandr.moria.org> <20000712072043.H1690@argon.gryphonsoft.com> <20000712153329.I11000@strontium.scientia.demon.co.uk> <xzppuojl4by.fsf@flood.ping.uio.no> <vqcaefnaqt8.fsf@silvia.hip.berkeley.edu>

next in thread | previous in thread | raw e-mail | index | archive | help

--NJBzyuYYykgY2yjn
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

Satoshi - Ports Wraith - Asami wrote:

> Sorry, bsd.port.mk can't cope with -r.  If a file is left behind from
> an aborted transfer, bsd.port.mk will see it and won't call fetch --
> it will just go ahead and bomb in checksum.  This is a feature.
>=20
> The problem is that bsd.port.mk cannot tell if a file is a result of
> an aborted transfer.  (Humans can, but not a script.)

Hmm, I assume fetch(1) exits non-zero for an aborted transfer?  If so,
can't bsd.port.mk detect that and rename the file to whatever.ABORTED,
and check for that next time you fetch, and if it exists, try resuming?
It seems like a bit of an ugly hack though, and I haven't thought it
through in detail. :-( Although by your "this is a feature" I assume you
wouldn't want something like this, as you've thought about it before?

> Even if it somehow tried, you might get connected to a different
> master site the second time around.  What will happen if -r is
> specified?

It shouldn't matter, unless the distfiles are different on the different
master sites, which shouldn't happen, should it?  If it does, one of
them is out of date and you'd have got a checksum error from one site
or the other anyway.  Various Windows programs seem to take pride in
downloading chunks of files from different servers in an attempt to find
the fastest.

--=20
Ben Smithurst                 / ben@FreeBSD.org / PGP: 0x99392F7D
FreeBSD Documentation Project /

--NJBzyuYYykgY2yjn
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 5.0i for non-commercial use
MessageID: rFl6j6emQMlRML9Bhj7xYoyA0BwHcxFT

iQCVAwUBOW0CZCsPVtiZOS99AQE7BwQAll1GInesUeZumiDZC7dAXhekwwSYbImn
2TIthhGnP6brx2/1Cs4QHeYkIvYbeuqA57e2+y1Wp4Fd1oqtYC7dLcn9ytjdhnaS
J3p/PskbW7pjJ8VsJR/YbQDlXyxJxn7SU2zVTUPH8tVjW3OU/FMmoffqgEmFMp9c
Xf4umB8m3Lc=
=v6MK
-----END PGP SIGNATURE-----

--NJBzyuYYykgY2yjn--


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-ports" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20000713004229.Z11000>