aget [PSARC/2009/159 FastTrack timeout 03/13/2009]

[James Carlson]

Si-wei Louis Liu writes:
> I don't think puf is totally functioning equal to aget.  ;-)

I wasn't necessarily suggesting they were; just surprised to see the
somewhat obscure aget proposed with no mention of 'puf.'

> puf, just as its name, fetches bunch of URLs in parallel, more like 
> another wget with parallelism. However, it cannot simply download a 
> large file (i.e. a kernel archive like, an .iso image, etc.) in 
> parallel, esp. over a not so fast network. Aget can fill this gap by 
> dividing the large file downloaded into multiple parts, each of which is 
> handled by a pthread, and facilitates the falling over from download 
> failures.

OK ... though there are likely a few controversial issues buried in
there.

> Moreover, aget is based on BSD(-like) license, and its been ported to 
> OpenBSD, NetBSD, FreeBSD, and Linux, etc. long before. This is another 
> reason for selecting aget as multi-threaded HTTP file downloader on Solaris.

I don't think the license really matters, at least in this review.

The unfortunate thing here is the apparent long-standing disagreement
between the aget and wget folks on multiple streams, leading to
duplicate tools.  *sigh*

In any event, +1.

-- 
James Carlson, Solaris Networking              <james.d.carlson at sun.com>
Sun Microsystems / 35 Network Drive        71.232W   Vox +1 781 442 2084
MS UBUR02-212 / Burlington MA 01803-2757   42.496N   Fax +1 781 442 1677