Do you know what deja vu is? That same “it already happened somewhere”, but not about the last movie released in the cinema, but about what is happening to you.
For example, didn’t it seem to you that you had already seen any article on Habré before? In most cases, there is nothing wrong with this. Technical articles often contain previously published information - they are finalized, rechecked and simply improved. Well, the news - you just have to put up with them.
So, let's see if this really happened already by doing a little research.
Home
My deja vu started with this image. It is worth noting that I do not read most of the posts, but only look at the headlines and pictures before the tag <cut/> and I keep scrolling. That's why the harmless JPG caught my eye - it had already been somewhere recently.
Rice. 1. KDPV from (2019-11-20)
Plus, I've definitely seen translations of Robert C. Martin's posts quite recently. Find them was - all three. KDPV (2019-11-12) - the same image on .
"So what?" - asks everyone who has read up to this phrase. Of course, authors of posts on the same topic often use similar images to attract attention, and at the same time save space on hsto.org. To begin with, I will say that the authors of the mentioned posts are different ( и respectively). But the quality of their translation is similar, more precisely .
But back to the third post-translation, which I already mentioned - (2019-11-16).
KDPV
Rice. S1. KDPV from (2019-11-16)
Yes, KDPV is formally different - it calmed my paranoia, and I began to read the translation. We will not dwell on the quality of the translation, I will only note that after reading I saw the author's login and the paranoia returned. Post written . And it's not the same user .
Well, now I just couldn't stop. I had to find out how many of the same translations were published on Habré.
Investigation
The first step in our pseudo-detective investigation is to find common ground. A lot of things are published on Habré and it was necessary to narrow the circle of suspects. It all starts with “Hello, Habr! I bring to your attention the translation of the article "- the standard beginning. Surely, it is substituted by a UFO that releases posts from the Sandbox. Stop, sandbox - that's where all three posts come from. In fact, only two out of three, but I noticed this later and my initial oversight did not affect the investigation in any way. And they are all translations. For starters, this is enough for us.
In total, at the time of the first search (2019-11-20), I found 37 similar posts up to and including 2019-07-15. 14 of them were published in November. I may have missed some of the same during this period - the search was done manually, because as a real detective I do not trust anyone, not even a computer. It is worth noting that at the time of writing this part of the text (2019-11-23), 5 more translation posts from the sandbox were added. Wherein was sent into deep space by cons and hidden by the author.
So, we have 41 + 1 posts - what do we know about them and their authors. 10 authors have more than 1 post, maximum - 7 pieces from one. We will assume that these users are really active members of the Habra community. 12 of the remaining 32 registered on the site before 2019, and some of them as early as 2012. Two more users were registered 119 and 109 days before they published their first posts.
After such a rough filtering, 18 registered ones remained. Let's see the statistics of their transfers. Ratings range from -19 per (and he also received the most minuses - 24) to +33 for the already mentioned (with a maximum of 46 pluses). Fast was hidden until the collection of statistics from 2019-11-23 (but I saw a lot of minuses there, although you don’t have to believe me).
The statistics of the authors is slightly different - the user received the most votes for karma - 21 (which did not help the total karma in 6.2), and the lowest karmo rating for - 0 karma from 12 votes. Only one of the users commented on various posts on Habré, and therefore we will also exclude him from consideration. Five commented on their posts and another 12 did not write a single comment.
But something must be in common between these users? Finding the answer was not difficult. It was enough just to go to the tab subscribed to / subscribers. And then the answer to my paranoia awaited me - there is something in common between them. They are all signed to each other. Well, like everyone else, 11 accounts. Let's take a closer look.
Rice. 2. Mutual subscription of the accounts in question from 2019-11-20 to 2019-11-23
Same GIF animation (186 KB)
Rice. S2. Mutual subscription of the accounts in question from 2019-11-20 to 2019-11-23
Isn't it suspicious? Especially when you consider that they have practically no subscriptions to other accounts. An attentive reader may notice that in addition to the 11 users that I spoke about earlier (and which are listed under the following spoiler), there is a twelfth in the data - is the only one who wrote the original post . What he is doing here is a separate question.
List of signatories
Tab. S1. List of suspicious accounts sorted by date of receiving an invitation to Habr
| Account | Karma (votes) | Posts / comments | Register | Invitation | days to |
|---|---|---|---|---|---|
| 4.7 (14) | 1 / 2 | 20.09.2019 | 04.10.2019 | 14 | |
| 8 (8) | 1 / 0 | 11.10.2019 | 24.10.2019 | 13 | |
| 6.2 (21) | 1 / 4 | 30.10.2019 | 01.11.2019 | 2 | |
| 3 (3) | 1 / 0 | 28.10.2019 | 07.11.2019 | 10 | |
| 7 (10) | 1 / 1 | 02.11.2019 | 12.11.2019 | 10 | |
| 3 (13) | 1 / 0 | 05.11.2019 | 14.11.2019 | 9 | |
| 4 (8) | 1 / 0 | 01.10.2019 | 19.11.2019 | 49 | |
| 0 (12) | 1 / 0 | 19.11.2019 | 20.11.2019 | 1 | |
| 6 (10) | 1 / 0 | 10.11.2019 | 20.11.2019 | 10 | |
| 1 (3) 8 (10) |
0 / 0 1 / 1 |
17.11.2019 | 22.11.2019 | 5 | |
| 2 (8) | 1 / 0 | 08.11.2019 | 22.11.2019 | 14 | |
| 8 (8) | 1 / 0 | 20.11.2019 | 22.11.2019 | 2 |
*For the user data is presented as of 2019-11-23 and -24 in connection with the publication of a new post.
All accounts have been registered since 2019-09-20, and hacked since 2019-10-04. Between registration and publication took place from 1 to 14 days, with the exception of
(49 days).
While I was collecting my thoughts and making a beautiful subscription scheme, the account published a new translation dated 2019-11-23 on a completely different topic and no longer from the sandbox. Previously, a translation of another version of the same text by the original author has already been published on Habré 2017-12-16.
What do suspicious accounts do?
So the originals are taken from:
- 4x ;
- 3x (it should be noted that two out of three translations do not lead to the blog itself, but to the short link service clck(.)ru);
- ;
- ;
- ;
- ;
- .
Of the originals, 11 are in English, 1 is in German. 3 articles are written by one author and 2 more by another, the rest are different.
Let's talk about topics, and in the case of Habr, about hubs. So, out of 11 translations available for reading in the following hubs are presented:
- 6x ;
- 2x , , , и ;
- 1x , , , , , , , , , , , , , и ;
Habr users pointed out inaccuracies or translation errors in 7 posts out of 10 commented.
In addition to posting one translation from each of the 11 accounts, the surveyed users also favorited each other's posts 24 times (maximum 5 favorites per account).
Of the 12 accounts, 5 are subscribed to 26 different companies represented on Habré - a total of 45 subscriptions, 2 companies meet all 5 times and 3 more times 3 times.
Someone else?
So far, only accounts that have been seen in subscriptions a significant number of times have been mentioned. Are there others out there? Yes, three. Two read-only (registration 2019-11-09 and -22) and one full-fledged (registration 2019-11-07, invitation from UFO 2019-11-17). The latter probably got out of the sandbox and hid the post later, so there is no information about the post at the moment.
Rice. 3. Subscription of the accounts in question as of 2019-11-24, supplemented by accounts with similar behavior
An interesting observation immediately after registration added to subscriptions and was no longer active (as of 2019-11-24). Subscriptions и have not changed since the start of observation (last activity 2019-11-20 and -18, respectively).
Is it often possible to see subscriptions to a freshly registered read-only account? Perhaps I will leave this question unanswered.
Instead of a conclusion
"What to do with this information?" you ask, having read up to the end? The question is certainly interesting.
On the one hand, these are only 11 accounts (and 3 more no less suspicious), whose contribution to Habr is “translations by Google Translate”. 6 of them received karma above 5 and can affect the rating of other people's posts and the karma of accounts.
On the other hand, 6 of the accounts have appeared in the last few days and whether this proliferation of interconnected "users" will continue is unknown.
But still, why do they do it?
On this vague note, in the spirit of any standard straight-from-the-movie detective film, I will end. Thank you for your attention!
"Just because you're paranoid doesn't mean you're not being followed" Harold Finch
Update: In comments regarding this issue.
Quote from comment
We were also interested in this story and, of course, we conducted an investigation, but much more primitive than yours, for one reason only - we have the IP of all these users. General, yes.
But everything turned out to be prosaic (and this is not the first case) - the IP address belongs to one of the specialized state authorities. universities in Russia. That is, the guys were given the task to make technical translations in order to get a credit / exam. There is an invite - there is a jackdaw in the student's record book, probably. And since the translations are pretty good, and the audience accepted them positively, why not help the good guys? Let's hope that habraauthors will grow up of them, or at least good readers 🙂
Update-2: reason for their activity. Therefore, we wish them to achieve their goal.
Quote from comment
One very "good" teacher gave us the opportunity to get a credit automatically if the article is published and upvoted.) Thank you for posting our articles!
PS If you find any typos or errors in the text, please let me know. This can be done by selecting a part of the text and clicking "Ctrl + Enter"if you have Ctrl, either via . If both options are not available, write about the errors in the comments. Thank you!
Source: habr.com