i intend on collecting all dumps in all languages, including images,
and don't want to break any laws (i'm in australia)
so far, i'm just experimenting with compression of the smaller dumps
a few things to note:
by uncompressing and renaming the text dump files (removing
"<wikiname>-<wikidumpdate>-*") and then using rdiffbackup, each
subsequent text dump is a fraction of a percent of the previous dump
for example, with enwikinews (without multistream or .7z):
original compression: 1.79GB
recompressed with xz -9e: 974MB
uncompressed: 22GB
uncompressed with rdiff-backup: 22GB, with only 200MB each subsequent
monthly dump (202???01 /* not 202???20/*)
rdiff-backup uses the rsync library
i intend on using the dumps for machine learning (when i study machine
learning in a few years) so i thought it'd be good to dealing with
huge amounts of data as a head start - also i'm concerned about the
dumps one day becoming corrupted, so i want copies now
i suggest that sha256sums be calculated for distribution (as well as
for the uncompressed files!) as apparently google has found potential
security flaws with sha1, and md5 is already completely insecure
anyway, i am just concerned about storing these dumps, especially the
images, but also the text
what are the chances of any of the images being of illegal content?
have there been cases of illegal images being stored in these dumps
before? what happened? what was the process?
if i find illegal images, do i just report to this dump list?
what happens if i'm unaware of illegal content in these dumps?
is there such a thing as text being illegal? can you elaborate?
also, the statistics below, are those sizes in page counts or bytes or
megabytes?
be awesome
griffin tucker
On Fri, 5 Mar 2021 at 23:00, <xmldatadumps-l-request(a)lists.wikimedia.org> wrote:
Send Xmldatadumps-l mailing list submissions to
xmldatadumps-l(a)lists.wikimedia.org
To subscribe or unsubscribe via the World Wide Web, visit
https://lists.wikimedia.org/mailman/listinfo/xmldatadumps-l
or, via email, send a message with subject or body 'help' to
xmldatadumps-l-request(a)lists.wikimedia.org
You can reach the person managing the list at
xmldatadumps-l-owner(a)lists.wikimedia.org
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Xmldatadumps-l digest..."
Today's Topics:
1. XML Dumps FAQ monthly update (noreply.xmldatadumps(a)wikimedia.org)
----------------------------------------------------------------------
Message: 1
Date: Thu, 04 Mar 2021 13:52:30 +0000
From: noreply.xmldatadumps(a)wikimedia.org
To: xmldatadumps-l(a)lists.wikimedia.org
Subject: [Xmldatadumps-l] XML Dumps FAQ monthly update
Message-ID: <20210304135230.bBj8c%noreply.xmldatadumps(a)wikimedia.org>
Greetings XML Dump users and contributors!
This is your automatic monthly Dumps FAQ update email. This update
contains figures for the 20210201 full revision history content run.
We are currently dumping 935 projects in total.
---------------------
Stats for bugwiki on date 20210201
Total size of page content dump files for articles, current content only:
19,778,622
Total size of page content dump files for all pages, current content only:
24,718,009
Total size of page content dump files for all pages, all revisions:
371,063,136
---------------------
Stats for enwiki on date 20210201
Total size of page content dump files for articles, current content only:
81,801,780,624
Total size of page content dump files for all pages, current content only:
181,026,335,781
Total size of page content dump files for all pages, all revisions:
22,229,491,552,833
---------------------
Sincerely,
Your friendly Wikimedia Dump Info Collector
------------------------------
Subject: Digest Footer
_______________________________________________
Xmldatadumps-l mailing list
Xmldatadumps-l(a)lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/xmldatadumps-l
------------------------------
End of Xmldatadumps-l Digest, Vol 127, Issue 1
**********************************************