Your message dated Wed, 12 Oct 2016 13:01:50 +0200
with message-id <20161012110150.ga4...@msgid.wurtel.net>
and subject line Re: Bug#84333: for speeding up partial downloads
has caused the Debian Bug report #84333,
regarding for speeding up partial downloads
to be marked as done.

This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.

(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact ow...@bugs.debian.org
immediately.)


-- 
84333: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=84333
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems
--- Begin Message ---
Package: rsync
Severity: wishlist

Many times, when one is downloading a large file, the transfer gets interrupted
for one reason or another. What if rsync could:

1) Compare file sizes between server and client, and for whichever one is
smaller (call this computer A), look at that much of the file on the end with
the larger file (call this computer B). In other words, take computer A's
file and compare the md5sum (or initially, Adler32) against the first however
many bytes of computer's B file (as determined by filesize on computer A). 
This treats the portion of the file already on computer A as one chunk, 
meaning only one checksum must be computed and sent over the wire. Possibly, 
rsync (on each end) could internally still break it apart, to take advantage 
of redundancies (but even with this, checksums of these internally-computed 
blocks do not have to be sent over the wire).

2) If they match, continue onwards (with the normal blocksize and whatnot) from
that point. This could save much time(/bandwidth) computing, sending, and 
comparing lots of small checksums. If they don't match, rsync can just work as 
normal.

-chris

-- 
"Meat. They're made out of meat."

"It's better than bad, it's good!"

(I subscribe to all lists that I post to; please do not Cc me)
Chris Danis                                 scre...@mad.scientist.com
                                                   screec...@home.com
                                                     dadan...@usa.net




--- End Message ---
--- Begin Message ---
On Thu 10 Jul 2014, Paul Slootman wrote:

> Sounds like --partial covers most of what this bug report is about, no?

Now closing this bug report after no feedback.


Paul

--- End Message ---

Reply via email to