bug-wget
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Bug-wget] Problem using GNU Wget 1.11.4 Windows version


From: JD
Subject: Re: [Bug-wget] Problem using GNU Wget 1.11.4 Windows version
Date: Mon, 19 Mar 2012 12:45:48 -0600

gnu does not distribute windows binaries.
So, I will resort to downloading it from from
http://code.google.com/p/mingw-and-ndk/downloads/detail?name=wget-1.13.4-static-mingw.7z


On Mon, Mar 19, 2012 at 9:33 AM, Micah Cowan <address@hidden> wrote:

> On 03/18/2012 03:24 PM, JD wrote:
> > When using wget with the -c option, it does recover and resume the
> download
> > after network failures. However, after it finishes the download (in my
> case
> > downloading
> > Fedora-16-i386-DVD.iso), I run the sha256sum on the downloaded ISO and
> it is
> > completely different to the value stored in the file of CHECKSUMS on the
> > same
> > page URL - http://mirrors.kernel.org/fedora/releases/16/Fedora/i386/iso/
> >
> > I downloaded this iso at least twice, with the same result - the
> sha256sum
> > performed on the file does not match the one at the above URL, and nor
> > does it match the result of sha256sum performed on the previous downloads
> > of the iso file.
> >
> > So, something is not right with wget!!
>
> As others have said, using a newer version is probably a good idea.
>
> However, it's probably also worth asking where you got your wget from,
> since we don't really provide official binaries for Wget. Perhaps it has
> a special case...
>
> It's also conceivable that it could be the server's issue, and isn't
> doing HTTP ranged requests correctly. Whether because of wget, or
> because of the server, the constantly varying sha256 sums are a clue
> that it's not happening correctly (assuming, of course, that all files
> are completely downloaded).
>
> With a partially-downloaded iso, I'd say, make a note of exactly how
> many bytes are in the partial download, and take a look at what the tail
> end looks like. Then, when you continue the download, take a look at
> that same spot, and see what you find. If HTTP headers suddenly appear
> there, or you see what appears to be the beginning of the file at the
> continuation point in the file, those are big clues. Also save a copy of
> the original partial download, so you can continue it again and see if
> you get different results, or if they're reproducible for the same-sized
> partial download being continued.
>
> And add the --debug flag to wget to get as much information about what's
> going on as possible. If you manage to find out what's happening, you
> may need these logs to know whether to blame wget, or kernel.org.
>
> Hope that helps,
> -mjc
>


reply via email to

[Prev in Thread] Current Thread [Next in Thread]