Talk:Trusted Computing/Archive 2

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

Archive 1 | Archive 2 | Archive 3

Trustworthy Computing

I'm editing the part that references Trustworthy Computing. Trusted Computing is not the same as Trustworthy Computing. The first is (quoting the article) a "family of specifications from the TCPA, which extend the behavior of a personal computer or server to offer particular cryptographic security services", while the latter is a Microsoft marketing campaign that was a response to the security problems within windows. Microsoft is involved with Trusted Computing, but they don't call it Trustworthy Computing. Though Microsofts "Trustworthy Computing" is mostly marketing hype, from an inside perspective, it's more oriented towards eliminating buffer overflows, ensuring privacy of data, etc. I'm also editing the Trustworthy Computing article that forwards here and changing it into a stub, since they aren't the same thing, unless anyone objects. Timbatron 06:20, 12 October 2005 (UTC)

How to tell if a computer is "trusted"

I'd put this in myself but I don't know the answer. If consumers are to bring market pressure against (or for) this technology, we need a way to know what we are buying. -unsigned

The answer is as simple as self interest. If you can trust it, they can't and vice versa. Like what's a "good" wage for the worker versus what's a "good" return on investment. "Trusted Computing" is ALL about WHO can trust the computer to OBEY them. This is the consumers's first product with a mind. And that mind can be programmed to obey it's creators rather than it's "owner". WAS 4.250 06:38, 12 September 2005 (UTC)

Suppose I go to the store to buy a computer. There is not a green one with a sticker that says "this computer cannot spy on you" and a red one that says "contains trusted computing technology". So the article could be improved by a discussion of how the consumer can tell when they are buying a so-called "trusted computer". Does anyone know the answer? Care to share? Chauncey27 19:09, 20 October 2005 (UTC)

Anonymity section

I love this neglected little article. The recently added anonymity section had these problems:

  • "As new identification technologies such as biometrics and RFID become widespread, it is expected that computer users will be identified with still greater certainty, and that ever increasing amounts of information will be available about them." It is expected by whom?
  • "proponents of TC believe" / "Critics point out" - Great subtle POV. Proponents of TC are obviously acting on blind faith, but critics (with their lab coats and objective facts) "point out" how the proponents are wrong.

Rhobite 00:08, 25 September 2005 (UTC)

Disputed

In light of recent POV edits, I began reverting the page, but came to realise that the entire page is littered with POV and it's going to somebody with more free time than I have to achieve neutrality and factual accuracy. I've marked it as POV in the hope that somebody can take an objective look at this page. StephenHildrey 17:45, 7 December 2005 (UTC)

I thought some recent edits were not so good too, but have not (yet) looked closely to see what material (if any) is usable in a NPOV way.
However, you need to do more than this to add the POV tag to the page. Generically saying "the entire page is littered with POV" is not useful. It doesn't give editors anything to work with, or any guidance on where to start with changes. You need to state with some specificity what the problems are, and where they are. I'm going to remove the template until such a statement is added here (not that I disagree with the characterization, but its facile to dismissively add a problem template with no criteria for when it might be removed).
I'd also recommend using the npov-section template instead, in most cases. It's more likely that particular sections have problems, and the narrower tag is more helpful for editors who wish to improve it. Lulu of the Lotus-Eaters 18:04, 7 December 2005 (UTC)
You're right - sorry I didn't list any reasons when I tagged the page - I'll try to help all I can in establishing a good TC page. Examples of what I perceived to be POV/misleading/completely incorrect include:
  1. "it allows computer manufacturers and software authors to monitor and control what users may do with their computers"
  2. "the United States Department of Defense's definition of a trusted system is one that can break your security policy"
  3. "Trust in security parlance is always a kind of compromise or weakness—sometimes inevitable, but never desirable as such."
I don't see the problem with any of these bullets. The well defined security parlance seems very worth discussing here. Lulu of the Lotus-Eaters 19:24, 7 December 2005 (UTC)
The problems (I've numbered the bullets above): 1 - innacurate. How will Dell or IBM suddenly be able to monitor or control your PC, just because it has TC hardware inside it? 2 - this quote originated from Bruce Schneier and I cannot find any reference to it on the DoD site, specifically the Rainbow Books [1]. 3 - I suppose in a strict sense that is accurate, though I must say that IMHO "trust" tends to be construed as a positive attribute rather than a negative one; if I "trust" something, it generally means I have a high degree of assurance that it will function correctly. I'm happy to get into semantic debates here, if that results in a more clear article that will successfully inform its audience, rather than pushing a massively biased viewpoint. StephenHildrey 20:03, 7 December 2005 (UTC)
Hmmm... those all seem OK to me still:
  1. Control is largely: "don't run software unless it meets these conditions". Monitor is largely: "phone home to get updated keys and convey machine/software state. Those are pretty much at the core of TC. A general CPU executes whatever instructions are architecturally present, a TC CPU checks whether the machine "approves of" the application as a whole. That's exactly what "control" means.
  2. Schneier is awfully well known and respected in security circles. I don't know for sure that he's quoting DoD correctly, but that is what he claims they define. We could rephrase it without the mention of DoD, but the concept of "can break your security policy" is important.
well known != always right. Schneier has a history of controversial viewpoints on crypto-gram, and the fact that he's well known should not get in the way of accuracy in an encyclopedia - whether that be the establishment of technical facts, or the correct quoting of who actually said what. I think there's a wood/trees problem here that needs addressing in the article page, not the talk page, rather than getting bogged down in the minutiae of security semantics. Alas I feel I'm fighting a losing battle here, so I'll leave that as a challenge for the reader. StephenHildrey 20:36, 7 December 2005 (UTC)
  1. "Trust as weakness" is quite good in cutting through a lot of misunderstandings of computer security. You're right that the word "trust" is often misconstrued in popular conception (the misunderstanding is encouraged by certain vendors like MS). But computer security experts are unanimous and unambivalent in this; trying to double-talk trust into something positive is extremely misleading, and not appropriate for this article.
Lulu of the Lotus-Eaters 20:20, 7 December 2005 (UTC)
  • the Ross Anderson tirade, without balance
Yeah, that's a sort of vanity thing. This article started as a cut-and-paste of the Anderson article, and hasn't wholly recovered from that. Lulu of the Lotus-Eaters 19:24, 7 December 2005 (UTC)
Ross Anderson is entitled to his opinion, but I don't believe WP is the place for it - especially when most of the POV fanboys pushing their agenda here are quoting from an article he wrote over two years ago, the technical details of which are woefully out of date. StephenHildrey 20:03, 7 December 2005 (UTC)
  • the entire Drawbacks section
You only have to look at the headings on the page to see that the article is POV. WP is supposed to be neutral: it is right that a TC article mentions privacy concerns, but it is not the place for poorly-informed, factless rants that dominate what is supposed to be an encyclopedia article. (Side note: I'm typing this using Firefox on a Debian box. I don't want to be "forced to run Windows" any more than the next man - but I do think that we need a factual, objective approach to this article). StephenHildrey 18:29, 7 December 2005 (UTC)

Looking today, I see that "This makes new anti-competitive techniques possible, eventually hurting people who buy trusted computers." has been re-instated. At best this is poor phraseology, but at worst it is inflammatory and typifies everything that is wrong with this article:

  1. it is inherently NPOV - would objective observers say that TC technology hurts customers?
  2. to say it makes anti-competitive techniques possible is fallacious because it implies that they are otherwise not possible (which they clearly are).

StephenHildrey 08:51, 13 December 2005 (UTC)

You are right. I added some attribution to the statements, and they can probably be improved further. I also renamed the section from "Drawbacks" to "Criticism", which seems a lot more NPOV to me (whether you consider the points as "Drawbacks" depends on your interests). Haakon 09:16, 13 December 2005 (UTC)

POV marker and reversion

I came across this article and saw that the intro did not in fact describe the subject matter as the rest of the article portrayed it.

The present intro essentially states that "this is claimed to prevent viruses" but that "many computer security experts disapprove" and various name calling that goes on.

I added the following edits to the intro, drawn directly from the article itself:

  • "It is a controversial term signifying broadly that the computer is designed to trust less, and potentially to enforce and require features such as digital rights management in order to be used. A trusted computer is one that a user is forced to trust more, due to lack of choice, rather than because it is "trustworthy". This is a security definition of the term, in contrast to the lay-understanding of the meaning of the word."
  • I changed the reason for concern to reflect the article: "...it allows computer manufacturers and software authors increased control to monitor and dictate what users may do..."
  • I added the explanation of trusted computing policy: ie a system that trusts software and external users less, and therefore is more strict about enforcing or denying different activities, and therefore in theory can itself be trusted more.

In other words:

  • It is a controversial term (factual)
  • What exactly the name signifies (factual)
  • What exactly a "trusted computer" is and why this could be a controversial issue (factual)
  • That the distrust by those who do distrust, stems from the increased control to dictate computer operation, that it theoretically places in manufacturers and third party hands. (factual)

I am reinstating this because I am unable to see any statement there which is 1/ POV, or 2/ not an agreed statement about trusted computing, or 3/ not essential in the intro to understand the shape of the debate (WP:NPOV) or why different sides believe and act as they do.

If anyone differs, please bring here rather than revert or tagging without discussion.

FT2 22:23, 7 December 2005 (UTC)

The main problem with the addition is that it is not very well written. I have to read the added second paragraph a half dozen times to figure out if I even think it's accurate. The main reason, I think, that it reads badly is that it's trying to argue a political position rather than describe facts. There's a big burden to overcome when your first edit to a page on a contentious topic is a complete rewrite of the lead; and this really doesn't seem to meet that burden.
Past that, I think most of the addition—even if it can be rewritten to be intelligible—is that it is not really "above the fold" material. A lot of details that are indeed worth including in the article as a whole don't fit in the intro. The first bit should give readers a basic sense of "what is it, and why do I care", but not yet get into the sides, criticisms, and precise debate. Lulu of the Lotus-Eaters 01:25, 8 December 2005 (UTC)

Comparing versions

I'm just trying to make sense of what FT2 is trying to get at with the new lead. As I say, it's very hard to read.

Consensus version FT2 version Notes
Trusted computing (TC) refers to technology from the Trusted Computing Group (TCG) which is claimed to allow computers and servers to offer improved computer security and protection from computer viruses and the like. Trusted computing (TC) refers to technology from the Trusted Computing Group (TCG) which is claimed to allow computers and servers to offer improved computer security and protection from computer viruses and the like.
It is a controversial term signifying broadly that the computer is designed to trust less, and potentially to enforce and require features such as digital rights management in order to be used. A trusted computer is one that a user is forced to trust more, due to lack of choice, rather than because it is "trustworthy". This is a security definition of the term, in contrast to the lay-understanding of the meaning of the word. (1) The term is not controversial (possibly mildly misleading), the technology is. (2) The rambling criticisms belong (better written) in the main body. (3) the "lay understanding" thing is too argumentative, we don't presume misunderstanding, just describe. I think we can keep the word controversial to pertain to the technology (not term). Lulu of the Lotus-Eaters
However, many computer security experts disapprove of trusted computing, because it allows computer manufacturers and software authors to monitor and control what users may do with their computers. Microsoft calls TC trustworthy computing, and Intel has started calling it safer computing, but Richard Stallman and some others call it treacherous computing. For this and related reasons, many computer security experts disapprove of trusted computing, because it allows computer manufacturers and software authors increased control to monitor and dictate what users may do with their computers. Microsoft calls TC trustworthy computing, and Intel has started calling it safer computing, but Richard Stallman and some others call it treacherous computing.
For policy purposes, trusted computing is an implementation of a trusted system. For policy purposes, trusted computing is a form of implementation of a trusted system, ie a system that trusts software and external users less, and therefore is more strict about enforcing or denying different activities, and therefore in theory can itself be trusted more. Inappropriate to include digression explaining "trusted system" (main body has more, as does linked article). This is the intro and brevity matters. Lulu of the Lotus-Eaters

Fair points. Answers and comments:

  1. Clearly the term is controversial as well as the technology. "Microsoft refers to TC as trustworthy computing ... Richard Stallman and some other critics suggest the backronym treacherous computing", clear evidence of a controversy over the wording independent of and additional to controversies over the technology.
  2. Could you clarify what exactly you mean by the "criticisms" which you feel I've added to the 2nd para? I added the information that TC signifies that the computer is designed to trust less, and potentially to enforce and require features such as digital rights management in order to be used. A trusted computer is one that a user is forced to trust more, due to lack of choice, rather than because it is trustworthy.. It's not clear to me how this is a "criticism". These seem pretty common agreed facts and definitions from both sides. It is also the heart of the reason why a controversy over TC exists at all.
  3. The comment that "the 'lay understanding' thing is too argumentative, we don't presume misunderstanding" seems extremely odd in light of the fact the article goes immediately on itself to say "Trust means something different to security experts than the meaning laypersons often assign".
    Not immediately. The main article has a different role than the lead.
  4. The problem is that the intro to the "trusted computing" article now doesn't actually say what trusted computing is. I find that very odd. It says merely, what it is claimed to achieve and that some people feel it gives over too much control and call it names. For an article intro, thats inadequate. The intro to "trusted computing" has to say what the actual meaning of trusted computing is.

If you find my wording unduely complex, take what's useful from it and write your own, but a short summary of what TC actually means and what it is, belongs in the article intro somehow. For that purpose a 1 sentence explanation of a trusted system, a little more than just "TC is an instance of a TS", is appropriate. That's all I'm trying to achieve. I haven't reverted since I am looking for comments and discussion though.

Clarification question on one aspect of TC: Would it be fair to characterize the critics' stance and backronym as indicating they feel the term "trusted" is a deceptive (or at best misleading) euphemism? FT2 06:53, 8 December 2005 (UTC)

I think readers can figure out that "treacherous" isn't meant as complimentary. Lulu of the Lotus-Eaters 07:39, 8 December 2005 (UTC)

Scylla and Charybdis

And the rest of the above?

Would it be fair to characterize the issues in the debate as follows:

  1. Trusted computing is a term taken from the field of trusted systems
  2. A system described as "trusted" does not necessarily mean it is trustworthy, rather it means it trusts less - both its own users and the outside world.
  3. In effect it means the system is more strict and less accepting of a wide range of activities.

And that because of this:

  • Proponents say that it will make computers safer, less prone to viruses and malware, and more reliable.
  • Opponents say it puts too much power and control in the hands of those who design their systems and software, and potentially forces technologies that can be abused (such as DRM, document sealing) and loss of anonymity upon end users.

And last:

  • Regardless of the debate and the form of the final product (not yet decided), major influences on current computing, including chip manufacturers such as Intel and AMD, and systems software writers such as Microsoft, are planning to include TC in their coming generations of products.

Is this a fair summary of the present position?
FT2 07:57, 8 December 2005 (UTC)

The problem that critics, who come along as editors every couple months, have is that they make a sort of category error. Trusted computing is firstly a marketing name for a bunch of loosely related technologies/concepts put out by a particular organization. A lot of the criticisms are a bit like criticizing the sawzall brand of reciprocating saw because it doesn't really "saw all" things... and then inventing backronyms to make fun of the name and the like.
There's a real primary sense in which what TC is is exactly what the TCG wants it to be, no more and no less. They invented, and probably trademarked, the term. I'm actually not terribly happy with the whole "treacherous" spiel in the lead at all (though it's certainly fine lower down), but I finally gave in after a bunch of free software advocates kept reinserting it. The current form is the least noxious one we've had, so I'll settle for that. And yes, I'm a far more radical critic of TC than are any of the editors who put in unencyclopedic rants; but I know WP is an encyclopedia.
"Above the fold" it's our job to tell readers what a thing is, not to go into every argument about it that we can think of. Below the TOC we have plenty of room for more detailed discussion. Push my analogy a bit more... if the "sawzall" trademarked product had notable criticisms, those would likewise go below the TOC. Righ now the article is a redirect to the generic term, but suppose it had its own article, and that some people felt the model had issues (safety, reliability, price, whatever): the sawzall article should start with "Sawzall is a trademark name for a tool produced by Milwakee company for blah, blah". That's neutral and factual. Below the TOC, we might get sections for "safety issues", "price manipulation", "deceptive name", etc that critics raise (purely hypothetically, AFAIK this is a fine model of saw with no such notable problems). It's fine, FT2, that you dislike TC, but we're writing an encyclopedia, not a critical tract. Lulu of the Lotus-Eaters 08:15, 8 December 2005 (UTC)
Thats a very strange reading of the information added. I doubt you can point to anywhere I have advocated a side or that I "dislike" or "like" TC? As pointed out, every thing edited in has been about "what is TC" -- definitions and characterizations of the notable views of proponents and opponents. Your comment above is that the article should "tell readers what a thing is". My main problem with the present intro is that it does not tell readers what TC is at all. It merely presents views of two sides on it, without ever defining the "it" in the first place...
The other problem I'm beginning to have is that you have three times now ignored questions I've asked. After you posted the above table (which was a good idea) I posted some 4 comments on it. You ignored them. I asked a clarifying question. The answer is "yes/no". My question isn't "can readers figure it out" but "is it a fair charecterization of a position". Ignored. I have now posted a summary of both sides of the subject matter as best I can see it, favoring neither and advocating none, and your reply is essentially 1/ your personal view on the topic, 2/ a straw man it's fine, FT2, that you dislike TC... this is not a critical tract, and 3/ to ignore my question again.
Whether or not TC is some individual company's view or not, the facts of the matter are, that what I have bulletted above is a summary of the TC debate as it is currently framed. If not, then you need to say what a more accurate framing of the debate is, read my comments rather than read into them, and reply, rather than ignore. It would help us progress this question if you could.
FT2 12:50, 8 December 2005 (UTC)
Perhaps you should write a separate article on the Trusted computing debate then? What TC is is a trademarked marketing term created by a particular organization.
I think you're probably right about a separate Trusted computing debate article. As far as I'm aware though, "trusted computing" is not a trademarked term - at least according to the USPTO. StephenHildrey 19:32, 8 December 2005 (UTC)
And yes, it's true that I haven't repeated for a third and fourth time every problem about every "question" you've asked. There's a reason the consensus version became consensus. Adding lots of argumentative, poorly written, run-on sentences to the lead doesn't help get us a NPOV article. And not everything that is in the main article needs to be in the lead, after all. Lulu of the Lotus-Eaters 19:11, 8 December 2005 (UTC)

I think this is getting misleading. The information added is self-evidently factual, or at the least direct summaries of what the article states are factual. You have accused me (notably via a straw man) of unsubstantiated bias, and have failed when asked to produce any bias evidence. You have not declined to answer the "3rd or 4th" time, but also the 1st and 2nd. You have labelled the additions "argumentative" when actually it's pretty clear only one person is "arguing". By another straw man you comment "not everything" has to be in the intro whereas my concern is that the actual definition itself which is the point of the intro, is missing. The additions are factual, and the present intro does not describe the actual subject matter which is unreasonable. Please explain your reasons for these below, for which I have asked several times in good faith. Thank you. FT2 21:02, 8 December 2005 (UTC)

Let's try to get some other editors to opine. There's no point in going around in circles. But the consensus lead that has been there for a number of months reflects lots of compromises among previously active editors, let's not throw all that away without some more input. Lulu of the Lotus-Eaters 21:44, 8 December 2005 (UTC)
That would be sensible. Let's give it a few days and see what others feel. But if none do, then we still need to think about a definition of TC for the intro in any event, and how to characterize the two sides' view. Even if TC is just "exactly what the TCG wants it to be" then we can state what TC is, what they "want it to be" and the basis and charecterization of the two sides' views, rather than just what one side says it will achieve. Both from website and press reports, as well as from the ensuing article, it seems that this can be fairly well characterized. FT2 22:06, 8 December 2005 (UTC)

Richard Stallment says

Richard Stallman says: "The technical idea underlying treacherous computing is that the computer includes a digital encryption and signature device, and the keys are kept secret from you. Proprietary programs will use this device to control which other programs you can run, which documents or data you can access, and what programs you can pass them to. These programs will continually download new authorization rules through the Internet, and impose those rules automatically on your work. If you don't allow your computer to obtain the new rules periodically from the Internet, some capabilities will automatically cease to function. Programs that use treacherous computing will continually download new authorization rules through the Internet, and impose those rules automatically on your work. If Microsoft, or the US government, does not like what you said in a document you wrote, they could post new instructions telling all computers to refuse to let anyone read that document. Each computer would obey when it downloads the new instructions. Your writing would be subject to 1984-style retroactive erasure. You might be unable to read it yourself. Treacherous computing puts the existence of free operating systems and free applications at risk, because you may not be able to run them at all. Some versions of treacherous computing would require the operating system to be specifically authorized by a particular company. Free operating systems could not be installed. Some versions of treacherous computing would require every program to be specifically authorized by the operating system developer. You could not run free applications on such a system. If you did figure out how, and told someone, that could be a crime." WAS 4.250 22:11, 20 December 2005 (UTC)

Intro

Coming back a few days after the above, the intro still has the same major failing -- it doesn't characterize "trusted computing", nor does it characterize the debate about it. I've taken care to respect above opinion in a revised draft that I hope will be acceptable. A quick explanation of the revised intro:

Overall structure:

  1. What TC is:
    • That it's a creation of the TCG (respects the talk page view that it is not a general term but a corporate group creation)
    • That its a term taken from the field of trusted systems (respects the talk page view that it is just one instance of a trusted system, and not a "field" itself)
    • That trusted in this sense and usage doesnt necessarily mean the same as trustworthy (stated in main article)
    • That it means the computer follows its programming with lower chance of incorrect disruption, or of undertaking forbidden activities defined by its programmers. (basic definition of a trusted system or computer, precis of definition in article)
  2. The nature of the controversy over TC:
    • That there is a controversy surrounding the latter aspect of TC
    • Proponents say it will make them 1/ more reliable, 2/ less prone to malware, and 3/ will enhance computer security.
    • Opponents say that 1/ the trust in the underlying companies is not deserved, 2/ can be abused and 3/ may force technologies and lack of anonymity on users.
    • It is also one possible means suggested to meet the next generation needs of corporate and other markets for a means of document and copyright protection. (Point both sides technically agree upon, nobody has denied that some corporates and users want such features and that TC could potentially deliver such features)
  3. Security specialists and other notable 3rd party views:
    • Security experts say (whatever they said in the original intro).
    • Some such as Stallman call it (backronym from original intro).
  4. Current state of play:
    • Major corporations are planning to use it in their upcoming products.

It seems that is a stance that accounts for all the notable views we have had, summarizes the debate, doesnt side with corporates, experts or users, and describes the subject. It's taken a lot of thought to try and convey every notable aspect that users have raised here in a balanced structured manner. If any of it is still perceived as inaccurate or unbalanced please bring it here to discuss first.

FT2 14:43, 13 December 2005 (UTC)

Good job. I'm happy with the writing quality and neutrality of today's version of the lead. I actually would still rather move the backronym lower in the article, but I've found prior editors try hard to get it into the lead. In my mind, the substantial criticisms are important for the lead, but wordplay on the name is a minor footnote (worth including later, certainly). Lulu of the Lotus-Eaters 19:09, 13 December 2005 (UTC)
Wow. In that case all I can say is, apologies for any stress caused by previous versions that missed that mark. To me it was important to observe that the article intro didnt define the article subject or the controversy surrounding it, or characterize the debate, and thats what I've tried to see fixed. One could justify the backronym I imagine, on the basis its cited and from a notable credible source, and that it represents that source's characterization of the subject. But as you say, its a minor thingy.
If anything's missing for me, it's the "anti-TC" view that TC based computers would potentially have a crippling effect on non-proprietary software and services, and the IT market, I think that's the opposing view to the "pro-TC" point about the corporate and govt demand for data and copyright protection, and could be justified as an addition, as it is notable, a significant concern, not hinted at, and balancing an existing point. Something like: There are significant concerns that TC would have (or may even covertly be intended to have) a crippling anti-competitive effect on the free software markets, private software development, and the IT market in general.
I think that's factual, important and not stated. But I'll leave it here as a thought for now.
FT2 20:24, 13 December 2005 (UTC)
I think that sentence is reasonable for inclusion in the lead. The article later does get to that concern, but it is probably central enough to merit "above the fold". Lulu of the Lotus-Eaters 20:43, 13 December 2005 (UTC)
I agree: that sentence seems reasonable because it is stating that concerns exist, rather than expressing the concerns. Nice work FT2 - the article is stronger for your efforts. StephenHildrey 20:46, 13 December 2005 (UTC)

Wow. And, added. FT2 14:05, 14 December 2005 (UTC)