Talk:Cleanfeed (content blocking system)

From Wikipedia, the free encyclopedia
Jump to: navigation, search

Comments[edit]

Jottings from the Clayton (2005) paper for possible inclusion in the article:

History:

  • Illegal to "make" indecent images of children since 1994. make includes saving and browsing. See Sommer (127) and Foundation for Information Policy Research (35)
  • 1996 - Internet Watch Foundation - originally mainly on usenet. UK ISPs sent a report and remove image from news servers (avoids ISP being done for possession?)
  • 2003 -only 2.4% is usenet. IWF accepts reports on sites worldwide - UK site ->UK police, offshore ->national criminal intell. service to interpol. IWF keeps database of URLs and date of illegal material
  • Late 2003 - BT made cleanfeed to stop own customers accessing sites on IWF blocklist
  • June 2004 - cleanfeed goes live. leaked to press (22)

Design: Hybrid system. 1st stage like packet dropping but packets not discarded but routed to 2nd stage content filtering system

  • traffic examined. if from a suspect site (some of which may be blocked) then redirected to 2nd stage filter
  • 1st stage on destination port &ip address
  • 2nd stage web proxy
  • returns 404 if matches item in database
  • 1st stage modified packet routing within customer facing portion of bt network (using bgp)
  • 1st stage uses Ip, 2nd stage uses URLs that are encrypted - compares hashes so that banned list is not accessible

Cleanfeed only uses port 80.

US Child Pornography Prevention Act Ashcroft v Free Speech Coalition (2002)


22 - http://observer.guardian.co.uk/uk_news/story/0,6903,1232422,00.html 35 - http://www.cl.cam.ac.uk/~rnc1/SexualOffencesBill.pdf 127 - http://isig.lse.ac.uk/news/evi.html P. Sommer: Evidence in Internet Paedophilia Cases. In A. MacVean, P. Spindler (Ed.): Policing Paedophiles on the Internet, New Police Bookshop, 2003, ISBN 1-903639-12-2 Invalid ISBN. Secretlondon 03:33, 31 March 2007 (UTC)

This page needs to be split[edit]

From UK / Aus, they're related only in concept, sharing a page seems pretty poor. —Preceding unsigned comment added by 60.242.64.202 (talk) 16:34, 2 January 2010 (UTC)

Could someone provide details on how the UK and Australian systems differ? Are we discussing technology, motivation, scope, administration, law, politics, all of the above...? CanuckMike (talk) 18:34, 17 January 2010 (UTC)
4 years later, I agree, the page is almost entirely UK based. I'm wondering what the naming should be:
Cleanfeed (content blocking system) and Cleanfeed (content blocking system - Australia)
or
Cleanfeed (content blocking system - UK) and Cleanfeed (content blocking system - Australia)
or even the popular[1] term:
Cleanfeed (UK Content Blocking System) and Cleanfeed (Great Firewall of Australia)
Deku-shrub (talk) 19:10, 5 April 2014 (UTC)

Blocking more than just CP?[edit]

I know BT block some boards on 4chan. If this is as a result of the cleanfeed system (the 4chan article says it is), then said system is already going beyond blocking CP, which _is_ banned on 4chan. Just because it shows up occasionally does not mean great swathes of the site should be blocked. Why not extend to blocking all sites with user contributions? No matter if illegal material is banned, it MIGHT show up. Even here!131.111.228.219 (talk) 13:15, 18 November 2007 (UTC)

Do you have a link about the 4 chan stuff? Vodafone's kiddy filter blocks flickr btw. Secretlondon (talk) 22:10, 18 November 2007 (UTC)
And *all* of 4chan!Secretlondon (talk) 22:14, 18 November 2007 (UTC)
I don't believe the 4chan block has anything to do with cleanfeed, it's more likely moot's rather silly idea of a prank. The IWF themselves have confirmed that no part of 4chan.org was ever on their list, what's more, it redirected users to the banned page (banned.php), whereas cleanfeed spoofs a 404 from the server a user is trying to access. I take it you've been on 4chan a while, so you'll know this is exactly the sort of thing moot would do, he'd think it was hilarious. --76.186.118.55 18:29, 2 December 2007 (UTC)
I was wondering this too - the IWF also list material they think is "obscene", as well as incitement to racial hatred. Do we have a source that it is just images of child sexual abuse?
Also we should distinguish between images of child sexual abuse and child porn - note that the latter is a broader category, as it includes fictional material, and images of consenting 16/17 year olds; is it specifically the former which is targetted? Mdwh (talk) 23:34, 16 February 2008 (UTC)
To add to that - my understanding is that no one knows what is on the list? In which case, we can only write what the Government claims is blocked. Mdwh (talk) 12:24, 17 February 2008 (UTC)
The 4chan block was confirmed by a rather nice method, actually - abusing Google search to confirm that 4chan's /b/ board was on the IWF blacklist via a correctly-formatted search query for its URL. I seem to recall people saying that the ISP blocking was quite cleverly done, involving sending the correct custom 404 page/"you have been banned page" for 4chan in order to hide the fact that Cleanfeed was blocking it, though that's harder to confirm. That block was probably CP-related; it's how they handled it afterwards that was dubious. - makomk (talk) 13:12, 21 July 2008 (UTC)


Freedom of Speech[edit]

I went ahead and removed "...which can be seen as an intrusion on free speech rights" and a biased reference. The reason being Australia doesn't have the same 'democratic rights' as America - although many people seem to think so, there's no law preserving freedom of speech at all! 202.180.109.97 (talk) 03:30, 25 November 2008 (UTC)

Except you're wrong. Case law from the High Court of Australia covers the discovered Constitutional freedom of speech in relation to parliamentary elections. Your deletion was correct though, as Cleanfeed doesn't impact on that freedom.Fifelfoo (talk) 00:47, 26 November 2008 (UTC)
Provide a source for your claim that Cleanfeed doesn't impact on freedom of speech. The Internode leak provided evidence that this will effectively slow down internet access across Australia. The UK version of the system has already blocked a number of sites that are considered bastions of free speech, whether by accident or because they allow users too much freedom. The statement that "I would rather be exposed to the inconveniences of too much liberty then to those attending too small a degree of it" seems to me, a pretty good retort to your personal, biased claim that cleanfeed does not in fact violate our free speech. The realities of this are quite simple. The internet, to the old ruling class, both in Australia and in the rest of the developed world, has become a monster they don't know how to control. By using excuses like 'stopping criminally offensive content' they can bring in filtering services, which, both in Australia and the UK have NO public scrutiny for what is blocked. Furthermore, the cleanfeed method merely makes a user think that a site is down rather then blocked so it is easy for your average user to think that the 'politically subversive' material they visit is merely no longer in existence.--122.107.198.87 (talk) 00:54, 6 December 2008 (UTC)
Although I dont entirely disagree, the point being made was that Cleanfeed does not impact on the right to free speech in relation to parliamentary elections, which is - apparently - the only right to free speech covered in either Australian case law or legistlation. Metao (talk) 11:31, 7 December 2008 (UTC)
--
122.107.198.87, could you read the protected freedom in Australia again, as Metao urges, "freedom of speech in relation to parliamentary elections." Australia's Cleanfeed has not yet had a policy document indicating it will impact this limited freedom of speech. In the broader sense of the Enlightenment value of freedom of speech, Australia's cleanfeed is a monster, yet, at the moment, it will only be empowered to block "illegal content," which includes euthanasia instruction, criminal activity instruction (criminality discovered post-arrest usually, particularly in common law crimes), almost all pornography except for M15+ or R18+ pornography behind a user-login system (see Broadcasting Services Act (1992) Schedules 5 & 7 in relation to X18+ material being illegal on line), Cat 1 and Cat 2 publications, communication with "seditious intent", copyright violation, and a few other catagories of illegal material (crypto, arms, etc.). None of which is freedom of speech in relation to parliamentary elections. Now in reality as the Finnish case demonstrates, amateurish police officers will be put in control of lists they know nothing about and "accidentally" block other content. Additionally, as previous attempts to filter the internet in trial demonstrated, filters will block political speech "by accident", as, for example, when a filter trial under the Howard Government "accidentally" blocked the National Party's website... a party in government at the time no less. tl;dr: Australia's cleanfeed is a monster, but it doesn't threaten the extremely limited freedom of speech Australians have, as we only have freedom of speech in relation to parliamentary elections.Fifelfoo (talk) 00:40, 8 December 2008 (UTC)

List of complicit ISPs[edit]

In the wake of the Virgin Killer debacle it seemed that not every ISP implemented Cleanfeed (contrary to the legislation which mandated it). Of the people who complained about it being blocked they all seemed to be customers of a handful of ISPs, most of which were resellers of BT's service (e.g. Sky, Orange, Tesco), a lot of the more 'independent' ISPs didn't seem to implement Cleanfeed at all (like ADSL24 and Zen); for the sake of assisting capitalism's needs for an informed consumer base to make the 'right' purchasing descisions, should this article include a 'name and shame' list of ISPs known to implement the system? W3bbo (talk) 13:40, 10 December 2008 (UTC)

I think linking to such a list would be appropriate, but unless we find one it might constitute OR. We really need to split this article... Metao (talk) 00:40, 11 December 2008 (UTC)

Any estimate as to percentage of daily usenet traffic is filtered by cleanfeed?[edit]

Just wondering how many posts (or how much traffic) on usenet meets the criteria of being filtered by cleanfeed and thus would be considered as usenet spam.

Also, can some mention be made here to the software known as noceum, which I think performs a similar function? —Preceding unsigned comment added by 64.231.15.187 (talk) 14:02, 31 May 2009 (UTC)

How to tell if Your ISP is proxying.[edit]

Based on past performance Cleanfeed is almost certainly blocking some URLs on the 'megaupload.com' site despite the fact that Megaupload will promptly delete files they are told about. They have a nice handy URL to tell if your connection is being proxied on their premium account advert page If the IP address for port 80 is different you've got a transparent web proxy if What is my ip shows your real address it's censorship. 90.195.73.153 (talk) 12:11, 7 August 2010 (UTC)

Reasons for this note ...
This technique can be used by anyone to show the technical effect that the structure in the description on the entry causes in real life. The different source IP addresses for connections to different ports and the not quite invisible proxy often cause issues so that normal users are basically forced to avoid the censorship to get a working website. This occurs even when the user is NOT trying to access censored material. The end result is that the censorship becomes useless because everyone ends up avoiding it without actually wanting to bypass it.
This also shows that it's stupidly simple to avoid the filters, use https:// or even just a different port, and so leads into the question of who is supposed to be being protected by this. Especially as it quite possible that people will avoid this without even knowing that they are defeating censorship!
So I think something related to this should be added to the main article but really I don't know how to structure it or really even where to start. 90.195.73.153 (talk) 20:16, 9 August 2010 (UTC)
Adding this to the article would violate WP:NOR. Also, it is not clear which version of Cleanfeed you are talking about (because the page still needs splitting). Metao (talk) 01:55, 10 August 2010 (UTC)
First about splitting, to date there is little or no technical information about the Australian or Canadian variations of this technology. So to date it is reasonable to assume that they will be using exactly the same technology as the UK version. As for the political landscape the only visible difference is that the AU version is to be mandatory for all ISPs rather than the ISP being branded as run by paedophiles if they don't 'voluntarily' do it. Again little difference. Until one of these changes I see no reason to split and good reason to keep in that the UK version is further ahead and according to non-technical sources is exactly the same technology.
As for OR charge, the existing reference "Clayton, Richard (November 2005). 'Anonymity and traceability in cyberspace'."[2] has all the technical attributions for this technique and it's common knowledge that websites like "megaupload" and "Rapidshare" have been public targets of the IWF and similar groups despite them having effective and well publicised methods of dealing with illegal and dubious content.
That just leaves the "show me my IP address" pages that are the standard test for detecting proxies.
So, what's the OR? The choice of proxy detectors? Google will show you others. 90.195.73.153 (talk) 07:08, 10 August 2010 (UTC)
You seem very concerned with Megaupload and Rapidshare (sites I personally find unusable to the point where I filter them out of my google results). "The term "original research" refers to material—such as facts, allegations, ideas, and stories—not already published by reliable sources. It also refers to any analysis or synthesis by Wikipedians of published material, where the analysis or synthesis advances a position not advanced by the sources." This article is about the Cleanfeed filtering systems. Clayton's paper is concerned with traceability - it is not a technical guide to filtering. This article should not include information about getting around the filters, or determining if you are currently being filtered. That kind of information is unencyclopedic and belongs on a site like Wikihow, not Wikipedia. Metao (talk) 07:50, 10 August 2010 (UTC)
Okay in order.
My examples of the Megaupload and Rapidshare sites are chosen for several reasons some of which I've mentioned above. (a) They are two sites I've tested personally using the above technique and others (such as tcp-traceroute) and my ISP has them behind a selective transparent proxy. (b) They are very large and popular websites and so should be a significant percentage of the load on a proxy of the referenced design. (c) They are law abiding companies and have very good responses to illegal and dubious uploads. (d) The have been taken to court and won cases against plaintiffs, sometimes before the court was officially in session. (e) Their websites are of such a design that transparent proxies cause significant problems, in the same way that they cause problems to wikipedia's abuse filtering. (f) It is therefore reasonable to assume that this content filtering is causing real cash losses to these law abiding companies.
These sites and Google search; These sites are designed for semi-private web hosting for sending files to your friends, employees or customers. Almost by definition these files should never appear on Google, generally this means if the file does appear on Google it will either be immediately deleted or it was crap in the first place. So yes, I too ignore any links from these sites that appear on Google.
Usability of these sites; I find they get a lot better if you get off the default port number or use https ... which is kinda the point.
Clayton's paper; I direct you to section 7.3 of this paper titled "Design of the CleanFeed system" which is a more than sufficient description of the system to recreate it. And to section 7.4.2 "Evading CleanFeed" which explicitly lists methods that can be used to avoid this system and how likely they are to continue working. He includes many methods that wouldn't have occurred to me even though I could be considered an expert.
The last point "unencyclopedic": firstly, I hadn't included "how to get around" the filter. True it's blinding obvious once you understand it's a first stage is a filter on port 80 to particular IP addresses but that's not a how-to. But the second part; it is possible you would see listing those two URLs as a 'Manual' on 'How To' detect the proxy. This isn't an unreasonable stance, in fact I see it that way, that's why I used the talk page, however, my (initially implied, sorry) question still stands. How should the important fact that this filter is as useful a chocolate fireguard be included in a nice polite encyclopaedic manner on the main page?
90.195.73.153 (talk) 08:47, 11 August 2010 (UTC)
"My examples" - WP:NOR. I rest my case. You don't get to write examples :) The fact that Clayton discusses ways around the filters does not mean that this article should contain said information. However, your last point is valid. It would not be inappropriate to add content on the effectiveness of web filtering. Such content would have to be carefully written and maintained to present an unbiased, well-referenced explanation - unlike the Content-control software article, which is quite horrible. Metao (talk) 01:59, 12 August 2010 (UTC)

Question about Cleanfeed Canada[edit]

Halton police in Canada claim to have the ability to detect computers attempting to access CP content. Using these methods they have then gone on to issue warrents to the ISPs to hand over the subscribers information.

My question is does this relate to Cleanfeed in anyway? Since this article states that Cleanfeed in Canada is not used by police for surveillance or prosecution, but to the extent of what they claim a lot of access to private networks would be required. That is assuming they're not refering to simple bait "honeypots" and capturing IPs.

http://www.thestar.com/news/article/951154--police-can-detect-computers-accessing-child-porn?bn=1

99.240.36.63 (talk) 22:03, 14 March 2011 (UTC)

Inclusion of a 'Criticisms of Cleanfeed' section and some minor tweaks[edit]

There are some grammatical errors on the page that I wish to correct and I'm also going to add a section on 'Criticisms of Cleanfeed' as it is clear that there are some issues being debated involving censorship, ways to circumvent Cleanfeed, and lack of transparency regarding its blocked content. I chose to note these under criticisms, rather than controversies or its potential to affect free speech/ expression, etc. in order to preserve neutrality. OneMadRabbit (talk) 04:05, 9 November 2011 (UTC)

I'm also going to expand this criticisms section. OneMadRabbit (talk) 01:36, 5 December 2011 (UTC)

Possible Split Cleanfeed/IWF URL/BT case?[edit]

The page is a real jumble of information and it seems to cover three separate topics, which are connected but aren't really defined as being separate entities in the article, perhaps it might be worth splitting into maybe two (or at a push; three) separate wiki entries? Suggest:

Cleanfeed[edit]

The system, it's implementation, the history, the technology, criticism of. So pretty much the page as it is now, with the IWF stuff removed and linked to on a separate page (see below), a note made that the cleanfeed system uses the IWF URL list for child sexual abuse images only - all other information and the other block(s) are the result of ISPs implementing their own blocks through legal threats, court judgments, etc.

perhaps put a subsection here about the MPA v British Telecommunications Plc case and it's impact on BT's Cleanfeed? or maybe a new section for this too?

IWF URL[edit]

aims (which ARE different from cleanfeed), the implementation (in cleanfeed), history, criticisms (these are different criticisms to cleanfeed (both since the court case and government intrusion). Cleanfeed is clearly being used for purposes which were never intended by the IWF, who's only remit and concern is child sexual abuse images. I think there are still some valid criticisms here BTW! Accountability, etc.)

I think the two were originally so dependent on each other, and effectively the same, that they deserved to share the same page here (on wikipedia), but I think the recent Newzbin case and political interference has split the systems/issues to such an extent that they no longer share the same ambitions/aims.

Obviously if agreed each page would reference the other, and they would be interlinked and have some crossover.

Any thoughts?

Bullblade (talk) 16:56, 12 November 2013 (UTC)

I did this, Child abuse image content list Deku-shrub (talk) 23:26, 26 April 2014 (UTC)
@Deku-shrub: no, you didn't. You've split out all the technological details away from the actual article on the system. Ironholds (talk) 17:33, 10 November 2014 (UTC)
@Bullblade: Right, so I've written much of Web blocking in the United Kingdom so believe me when I say I know my web blocking! I've even compiled a list of known technologies all the ISPs use - as you can see Cleanfeed is only, and I mean only used by BT. All the other ISPs use their own blocking systems which ingest the Child abuse image content list from the IWF. As a result, talking about Cleanfeed in a UK context only makes sense historically, unless you want to talk about something special with BT's system, which is no longer at all special. Deku-shrub (talk) 21:04, 10 November 2014 (UTC)