On Sun, Feb 14, 2016 at 3:37 PM, akhiezer <[email protected]> wrote:

> > From: Pierre Labastie <[email protected]>
> > Date: Sun, 14 Feb 2016 21:50:49 +0100
> > Subject: Re: [blfs-support] rsync-3.1.2
> >
> > On 14/02/2016 21:26, alex lupu wrote:
> > > On all my downloads, the MD5 sums come to
> > > 1be3ec9421f23e8eb7a06957992f6604
> > >
> > > However, the BLFS Dev. Book Version 2016-02-14 shows
> > > 0f758d7e000c0f7f7d3792610fad70cb
> > >
>


> > I just downloaded it (using the link in the book with firefox), and the
> md5sum
> > I get is that of the book.
> >
> > Using "wget -T 30 -t 5
> > https://www.samba.org/ftp/rsync/src/rsync-3.1.2.tar.gz";, I get the same
> md5sum
> > as the book, too.
>


> +1:
> $ wget https://www.samba.org/ftp/rsync/src/rsync-3.1.2.tar.gz
>         .
> $ md5sum rsync-3.1.2.tar.gz
> 0f758d7e000c0f7f7d3792610fad70cb  rsync-3.1.2.tar.gz

​----------------------------------------------------------------------------------​
​
Bruce, Pierre, Akhiezer,
​ ​
many thanks.

Curiouser and curiouser.
To summarize:
1.  On my main BLFS system, 32 bits

1.1. I normally download a file by clicking it with left button.
1.2. I normally download using my (32-bit) Chromium.
1.3. As far as RSYNC is concerned (however, see below for Ubuntu)
     this gets me the "terribly wrong", 3287040 byte file.
1.4. RSYNC downloaded through any other method know to man like,
     wget, Firefox, Win 10/IE (perish the thought!), etc.,
     gets me the "correct" file (892724 bytes).

% ls -og
 3287040 2016-02-14 15:15 rsync-3.1.2.tar.gz (chromium)
  892724 2015-12-21 15:23 rsync-3.1.2.tar.gz.1 (all the other)

% md5sum *
1be3ec9421f23e8eb7a06957992f6604  rsync-3.1.2.tar.gz
0f758d7e000c0f7f7d3792610fad70cb  rsync-3.1.2.tar.gz.1

(any surprize now?)

1.5. This is really a
​big ​
head scratcher.
     Once untarred, the two files create absolutely identical
     directory trees.  As a summary of the _many_ checks for identity:

% du -bd 1 rsync-3.1.2-3K
7810    rsync-3.1.2-3K/testhelp
497004  rsync-3.1.2-3K/zlib
84142   rsync-3.1.2-3K/testsuite
7270    rsync-3.1.2-3K/m4
18546   rsync-3.1.2-3K/doc
49886   rsync-3.1.2-3K/support
198981  rsync-3.1.2-3K/lib
62321   rsync-3.1.2-3K/packaging
99440   rsync-3.1.2-3K/popt
3162975 rsync-3.1.2-3K

% du -bd 1 rsync-3.1.2
7810    rsync-3.1.2/testhelp
497004  rsync-3.1.2/zlib
84142   rsync-3.1.2/testsuite
7270    rsync-3.1.2/m4
18546   rsync-3.1.2/doc
49886   rsync-3.1.2/support
198981  rsync-3.1.2/lib
62321   rsync-3.1.2/packaging
99440   rsync-3.1.2/popt
3162975 rsync-3.1.2

(after tar -xf ..., I named one directory ... -3K,
 so they can coexist)

​"Sparse" files?​
​(a cmp -l of the two tarballs crashes my machine :)​

--------
2.  For comparison, on my
ubuntu 15.10, 64-bit using Chrome 48.0.2564.109 (latest)

% ls -l
 320459924 Feb 14 18:25 qt-everywhere-opensource-src-5.5.1.tar.xz
   3287040 Feb 14 18:21 rsync-3.1.2.tar.gz

Note:  the BLFS book gives the qt file size=306MB -> OK

% md5sum *
c2a249e8795d45200d997d96860d0353  qt-everywhere-opensource-src-5.5.1.tar.xz
1be3ec9421f23e8E7a06957992f6604  rsync-3.1.2.tar.gz

PUZZLEMENTS
1.  Why the HUGE size difference on Chrom(ium) download of RSYNC?
2.  Why not
​with​
 other files (I picked qt-5.5.1, randomly, on a whim)?
2.1.  In all these years, always normally using Chromium, I
​_​
never
​_​

​have run into​
 a crazy size -> MD5 situation.

​Ho​ping
 you've had a good Valentine's day (to top it all off,
my red roses are already wilted - bad MD5, no doubt),
-- Alex
​
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Reply via email to