karma wrecking problem
hsitz
http://www.vim.org/scripts/script.php?script_id=3342
Earlier today it had a little over 1100 downloads -- maybe 1105 to
1110 -- with around 35 votes and 120 or 125 total karma. I just
checked the page and it now has jumped up to 1133 downloads, 57 votes
and 98 total karma.
It appears that there were 22 or 23 votes today, all for -1 karma.
Seems like someone has a campaign to lower the ratings of the plugin.
Is there any way to fix this? Some way to prevent this from
happening?
Thanks,
Herb Sitz
Steve Litt
Hi Herb,
I just read about this sort of thing in one of the trade mags. Turns
out for about 20 cents per review, you can have workers in China give
your site an excellent review (or perhaps a competitor a bad one).
Makes me a little nervous when I put such high reliance on ebay
sellers'reputations, and now I find out for a few hundred dollars they
can get thousands of positive reviews.
SteveT
Steve Litt
Author: The Key to Everyday Excellence
http://www.troubleshooters.com/bookstore/key_excellence.htm
Twitter: http://www.twitter.com/stevelitt
John Beckett
> I'm author of the VimOrganizer plugin.
> http://www.vim.org/scripts/script.php?script_id=3342
>
> Earlier today it had a little over 1100 downloads -- maybe
> 1105 to 1110 -- with around 35 votes and 120 or 125 total
> karma. I just checked the page and it now has jumped up to
> 1133 downloads, 57 votes and 98 total karma.
>
> It appears that there were 22 or 23 votes today, all for -1
> karma. Seems like someone has a campaign to lower the ratings
> of the plugin.
There has been a problem with spurious ratings and spam uploads
in the last few months. I deleted about ten spams nearly three
weeks ago and have watched for more spam since then. There has
been none, but I just checked again following your message and
sure enough script 3847 had recently been created and was spam
(i.e. it was some .rar download with non-Vim material).
I have deleted that script and its user. I will have a look at
the rating manipulations when I get a chance, but that may not
be for a while as I am going elsewhere soon.
See "rating manipulations on www.vim.org have taken place":
http://groups.google.com/group/vim_use/browse_thread/thread/889927c7afcea808
John
John Beckett
> I'm author of the VimOrganizer plugin.
> http://www.vim.org/scripts/script.php?script_id=3342
> It appears that there were 22 or 23 votes today,
> all for -1 karma.
Some preliminary findings for the ratings of script 3342:
Things look normal up to 2011-12-14 00:26.
In the next 12 seconds, there were 22 ratings of -1,
each from a different IP address.
Of those IPs, 19 also downloaded the spam script 3847.
I can delete those bad ratings, but I won't do that at the
moment in case someone like Bram wants to have a look at the
data. If nothing occurs, in a day or two I might delete the bad
ratings? It appears that might not fix the displayed karma, so
I'll have to think about what to do.
John
John Beckett
> Things look normal up to 2011-12-14 00:26.
> In the next 12 seconds, there were 22 ratings of -1, each
> from a different IP address.
I misread the timestamps: it was 15 minutes (not 12 seconds).
I also noticed that script 3025 had 231 ratings of -1 in 1h21m
starting 2011-12-12 02:43, each from a different IP.
Going to need some thought...
John
Bram Moolenaar
John Beckett wrote:
Sounds like the same thing what happened earlier. Although it's weird
that the script was actually downloaded from the same IP addresses.
That would not happen if the rating link was accidentally showing up
somewhere. It's more like a crawler that goes through a list of URLs.
--
Life would be so much easier if we could just look at the source code.
/// Bram Moolenaar -- Br...@Moolenaar.net -- http://www.Moolenaar.net \\\
/// sponsor Vim, vote for features -- http://www.Vim.org/sponsor/ \\\
\\\ an exciting new programming language -- http://www.Zimbu.org ///
\\\ help me help AIDS victims -- http://ICCF-Holland.org ///
hsitz
>
> I have deleted that script and its user. I will have a look at
> the rating manipulations when I get a chance, but that may not
> be for a while as I am going elsewhere soon.
>
John -- Thanks, I'd seen that thread so I wasn't entirely surprised to
get the mass of negative karma votes, but couldn't find the thread
again in a quick search yesterday. Strange to think anyone would
spend time/effort/money to do something silly like spam karma
ratings. . .
I don't know what all the potential solutions are on your end. Didn't
notice whether it was mentioned but possibly 'captcha' fields? I
would also put a vote in for requiring some comment from a user who
puts in a bad rating, explaining what they didn't like. Maybe just a
prominent message saying that if they're going to give a bad rating
they should let the plugin author know why. I know I've gotten a
couple bad ratings before and wondered what the complaint was and how
I could have fixed whatever caused them to dislike the plugin.
Regards,
Herb
Steve Litt
Sounds like the China connection to me.
Benjamin R. Haskell
> John Beckett wrote:
>
>> hsitz wrote:
>>> I'm author of the VimOrganizer plugin.
>>> http://www.vim.org/scripts/script.php?script_id=3342
>>> ...
>>> It appears that there were 22 or 23 votes today,
>>> all for -1 karma.
>>
>> Some preliminary findings for the ratings of script 3342:
>>
>> Things look normal up to 2011-12-14 00:26.
>> In the next 12 seconds, there were 22 ratings of -1, each from a
>> different IP address.
>> Of those IPs, 19 also downloaded the spam script 3847.
>>
>> I can delete those bad ratings, but I won't do that at the
>> moment in case someone like Bram wants to have a look at the
>> data. [...]
>
> Sounds like the same thing what happened earlier. Although it's weird
> that the script was actually downloaded from the same IP addresses.
> That would not happen if the rating link was accidentally showing up
> somewhere. It's more like a crawler that goes through a list of URLs.
Yes, it seems the same as before. Are we still close enough to the
incident that the ratings could be correlated with something that
indicates a User-Agent header (which might show that it's a
poorly-written crawler)?
--
Best,
Ben
John Beckett
Scripts with large negative ratings are shown at:
http://www.vim.org/scripts/script_search_results.php?order_by=rating&direction=ascending
Some old spam is shown in the above results. I will delete it in
a day or two. I was wondering if the rating attacks were aimed
at hiding spam (which are given negative ratings) by making lots
of scripts also have negative ratings. But that seems unlikely,
and nothing else I have investigated has given any
enlightenment, although I found one site which had a GET link
which, if clicked before Bram's recent fix, would have downrated
script 3464. To see it, search for "rating=unfulfilling" at:
http://lingr.com/room/vim/archives/2011/02/25
Vim-use discussions of rating attacks:
2011-09-01 "manpageview rating dive":
http://groups.google.com/group/vim_dev/browse_thread/thread/f9356dc653688bdf/84dc21ab976ca0dc
2011-11-19 "rating manipulations on www.vim.org have taken place":
http://groups.google.com/group/vim_use/browse_thread/thread/889927c7afcea808
2011-12-14 "karma wrecking problem" [this thread]:
http://groups.google.com/group/vim_use/browse_thread/thread/4194463f37323fef
The first thread above discusses script 489 (Charles Campbell's
ManPageView). It received 811 negative ratings in 56 hours
starting 2011-08-12. There are quite a lot of other scripts that
appear to have suffered from rating attacks ... a mystery.
John