Want to take part in these discussions? Sign in if you have an account, or apply for one below
Vanilla 1.1.10 is a product of Lussumo. More Information: Documentation, Community Support.
The other two are from 91.200.12.42 which is a known Ukrainian spam address. (Can someone more knowledgeable please check this out. It seems to be a bot / spider.) I will clear the two pages and send them to ‘empty’. (143 and 144)
The wp-login
pages are probably automatic attempts from a bot to gain access to our hypothetical WordPress installation.
I emptied a page F-theory - From11dSupergravity which seemed to be spammed. It is now empty 146.
Seems to be spam again, see latest revisions, probably from ’Computer Science’ up to ’mylab Classical dynamics books’.
Usually I delete and empty such entries, but in this case I was waiting to see if whoever had created the ‘mylab’ pages had some useful intent behind their attempt. What do others think? As David_C says, the entries start at Computer Science then mylab and so on, and at all from anon1 at 41.178.232.137, who is in Egypt it seems.
I think it’s spam. I never can quite remember what exactly we do. Is there a recipe we could write somewhere?
I will try to remember… although sometimes I get it wrong. :-(
A problem can arise if the page really was spamming something already written, so check the page did not have a History first. If it did exist prior to the Spamming, revert to the latest unspammed version. If not then you delete the contents, check the last number used on an ’empty’ file, (currently 146) and change the name on the offending file to the next (’empty 147’). (I think that is all, but am sure that others will check on what I have said.) Oh! remember the obvious thing that renaming a page will create a redirect so go back and clear that as well.
What Tim has outlined is what I recall Toby recommending to do in these cases.
Ok, thanks. So empty 147-151 are born.
Thanks, David, I usually do this but am travelling and was busy. I have also to prepare the the meeting in Paris at the end of the month!
The page meta topological space looks exceeding flaky to me. Not sure what to do about it. Not sure it’s quite spam, but certainly no good in its current state. The author, ’D. Ely’ has also made Meta matrices, which is likewise very odd.
Both look ’flaky’ to me. Perhaps put a polite warning notice on the shorter one giving a time limit for the author to explain what is intended on these pages. What do you think?
My own opinion, for what it’s worth, is that it’s not worth it spending any time dealing with the author. I would just get rid of them.
Thanks, Todd, I agree. That meta topological space one is particularly useless.
One addition to Tim's directions in #507: You should be able to quickly see how high up the empty pages go by checking http://ncatlab.org/nlab/list/empty. But this only works if you make the content of the new empty pages read
category: empty
instead of being literally empty! I have done this up to empty 152 now.
I have cleared “meta topological space” and have renamed “Meta matrices” to hypermatrix, which is a term that is pretty standard in the context of discussion of hyperdeterminants.
I added a reference to hyperdeterminants, linking to an MO discussion of volume interpretations of hyperdeterminants.
516: I disagree that hypermatrix is another term for tensor in more than 2 dimensions. It is equally wrong as claiming that an ordinary matrix is a tensor. Quadratic matrix can represent a tensor in particular basis, but the matrix does not know its origin. It does not know if it represents a covariant 2-tensor, or contravariant 2-tensor or mixed 1-1-tensor, and so on, nor it knows in which basis this representation is taken, hence it is not giving the complete information on a tensor. It is frosen datum of a tensor provided we know the basis. If we do not have this supplementary datum then it is not.
I am on the other hand not sure if the term is used for non-quadratic matrices of higher number of indices/dimensions.
Sure, basis representations. I have added the word to the entry.
Fundamentally, an -by- matrix with entries from the set is an element of , which is not any sort of tensor. However, if is a field and you wish to add and multiply the matrices in the usual way, then an -by- matrix is a -tensor on , with multiplication given by contraction. Since comes with a preferred basis, you can interpret such a matrix as a -tensor or a -tensor, but this is less direct. This is why people sometimes use the word ‘matrix’ to refer to a -tensor, even though that is (strictly speaking) improper. (Dually, people sometimes use the word ‘vector’ to refer to a list.)
Why are we discussing this in the Spam thread?
I got serious spam from IP 37.25.117.59 with stupid Russian texts overwriting all my text, today, on 09:57 January 18 on HomePage (zoranskoda) and homotopija lekcija2 (zoranskoda). I rolled it back.
The page Algebres enveloppantes (Cahiers scientifiques, fasc. 37) (French Edition) (French) Paperback – 1974 (created by an Anon Coward at 197.2.58.222) looks to me like the filename of a torrent-sourced copy of a book. In this case, it’s a book by Dixmier, and the title of that page could have been copied verbatim from this Amazon page.
I made that last one a real page and linked to it from universal enveloping algebra.
Mysterious spam from 2405:204:8485:ab49:a8f3:c856:6f17:9f86 at floor. It just says ’3.6’.
What is that type of identifier as a search did not produce any results?
It’s an IPv6 address.
I put some content at floor. (Such content may be discussed in a separate discussion.)
Unclear whether it is technically spam, but have deleted a page ’real line and circle with one point remove is homeomorphic’ created today (its only content was a copy of the title), as well as the announcement which it generated on the nForum.
Deleted page created with ’(X,d) is a pseudometric Hausdorff space…’ in the title, with content a proof of this. Have noted down IP address and author (joeljoeljoel). I believe that the same page may have existed before (which I deleted). I also deleted the thread with the nForum announcement of the page creation, amongst other reasons to avoid making it possible to create the page simply by clicking on the links in the announcement.
This is a strange sort of spam.
Some automated scripts may attempt to create spam that, thanks to Instiki's quirks or our filters, ends up not really working. (For instance, if there was supposed to be spam link in a user homepage field and the script makes the post even though we don't have such a field, although this is just my speculation.)
As to whether we should be deleting such spam, much less things are not even clearly spam as in the previous few comments, that's being discussed elsewhere.
We seem to be receiving spam attacks. Look at the latest additions for Kalb-Ramond field (by Edward Witten!), Hilbert module, circle action, Roger Penrose.
Sometimes a page appears not to be changed, e.g., the last version of From String structures to Spin structures on loop spaces by Britney. I guess they could change a web address and it wouldn’t show.
We do need to keep an eye on latest changes. Looking back what’s this at contact instanton? Seems that someone just filled in a link there.
I don’t have time right now to roll these back. Can we find out if these changes are coming from the same place?
I have blocked in nginx the IP address 69.156.53.123 which the two entries I checked were edited from. No time to look further into it or doing any rolling back. Let me know if there are other IP address which should be blocked (the IP address can be seen by looking at the History of the page).
I rolled back Kalb-Ramond field.
I rolled back Hilbert module There was a change made to the Context part of it with what looked like a suspicious link. The change was: by Sara Seger on November 26, 2018 at 16:22:57 from 69.156.53.123 as well. From the same IP address was circle action, as David C noticed and I rolled that back as well.
Thanks, everyone! Yes, let’s keep an eye on LatestChanges.
I have deleted the spam edits from the database now. Felt this was important to keep things clean, especially the one with ’Edward Witten’ as author. Have also deleted the page ’contact instanton’ and the corresponding nForum thread.
Spam on HomePage. I will roll back. It is from 115.96.106.12.
Thanks. Have deleted the revision and your rollback now. Not blocked IP address for the moment, but have noted it. Will block given a second occurrence.
Spam at Bluevhale. The address (95.15.183.5.) is in Turkey. I have left it to others to delete this properly.
Blocked IP address 72.182.53.175 which had rolled back the pages Freyd cover and adjoint functor, and had made several edits to the Sandbox. Both rollbacks and all other edits have been deleted.
Also blocked the IP address mentioned at 95.15.183.5 and deleted the created page Bluevhale (no other edits had been made with this IP).
Cleaned up the nForum afterwards.
I do begin to think that we must do something about anonymous edits, maybe making them ’pending’ or something.
see simplicial complex. Revision 39 was spammed and the page overwritten by 137.189.88.229.
Thanks Tim! I have deleted that revision now, so Revision 39 is just your rollback, i.e. is identical to Revision 38. I have also blocked the IP address.
Deleted new spam at quantum information, and removed traces of spam from earlier this year. One of the ’sophisticated’ attacks where blocking IP address will not work. Added content to spam filter, but the text was changed this time compared to the previous attempts.
Removed some spam from the Latest Changes thread for bitcoin.
Anonymous has added and had removed 3 times a spam link in An Essay on the Foundations of Geometry
maybe papersowl.com
should be added to the spam filter.
Addressed now,
Just for the record: the account https://nforum.ncatlab.org/account/1976/ just spammed half a dozen nForum threads.
Blocked the account and cleaned up the spam now.
I’ve now added “Tango” and “tango” to the spam filter - a cursory Wikipedia search didn’t indicate any mathematics-related meanings, but please let me know if this causes any issues.
We should also blank or remove the nLab pages.
@Mike Shulman, re. #3: Yeah, I was wondering about that. :-) Which is preferable?
For standard uncontroversially clear-cut malicious spam we should delete the pages for good, with no trace in the history left.
Others here (Toby) have used to urge that as soon as there is a shred of possibility that something is meant as an actual contribution, and be it as obnoxious as it may, then we should keep it in the History. And I see that point, given what’s going on in the world.
But standard malicious commercial spam is uncontroversially recognized and should be deleted for good.
@Urs, re. #5: Okay, so on the basis that the two ’Tango’-related pages are clearly inappropriate for the nLab, I’ve now deleted the relevant pages from the database. Thanks!
Thanks Alexis! Yes, I always delete pages/revisions in these circumstances to keep the history clean, and also delete the corresponding nForum threads usually. Just in case you need it, this is the procedure I follow for deleting nLab pages.
1) Look up the id in the pages table. E.g. something like SELECT * FROM pages WHERE name=’xxx’
2) Do DELETE FROM revisions WHERE page_id=xxx
3) Do DELETE FROM wiki_references WHERE page_id=xxx
4) Do DELETE FROM pages WHERE id=xxx
For deleting an nForum thread, I delete both the discussion and any comments with that discussion id.
For instance I think I saw recently an example where Rod rolled back some spam. I would usually delete that revision (and the rollback) in the database to keep the history clean.
Just in case you were not already aware, one needs to restart the nLab server for spam filter changes to take effect by the way (/bin/unicornctl restart). There is no downtime with these restarts. Apologies if you already figured this out, but I thought I’d mention it!
@Richard,
re. #7:
Thanks - I’d done the deletions from ’pages’ and ’revisions’, but I hadn’t thought to check ’wiki_references’, and I wasn’t sure whether to delete the relevant nForum threads as well. In this particular instance, should we keep both threads, since non-spam discussion has been had in them?
re. #8:
Oh, I was indeed not aware of that - thanks for bringing it to my attention!
Removed page with title ’antiderivative’ with spam content from IP 2600:8802:5902:e600:b1ba:8912:6dfd:f3ed, along with corresponding created nForum thread.
Comments #549 - #556 were originally in a different thread created by the edit announcer upon creation of a spam page with title ’Tango support center’. Moving here to allow deletion of the original thread (which I have now done).
Hope this resolves your question in #556 satisfactorily, Alexis!
If I remember I usually make a note of spam deletions in this thread here.
Hi Alexis,
I forgot one more thing that I do as part of the deletion procedure: to remove the page from the cache. E.g. rm www/nlab-prod/cache/views/nlab/show/xxx
. Some other things are also sometimes cached, e.g. history and revision pages, but those are viewed less often, and I sometimes don’t bother with those; just have a look around in the cache directories :-).
Removed nForum spam about tech support, added relevant word to spam filter.
I have deleted a page with title ’Every discrete topological space Is locally connected space’ with similar one-line content. I believe this was ’accidental spam’. Urs had made a small edit to the page to try to make it more reasonable, but Todd agreed with deletion, so I am going ahead and doing so.
Deleted a page ’Set theory proof’ which was the same kind of thing as in #561.
Deleted a page with title ’URL’ and no content.
Deleted a page ’Fréchet spaces is not banach space’ and its associated nForum thread. Similar cases to the above.