But now people in these countries have the right to ask people to forget about things about them which are true

Incorrect. If the court was saying to remove the page in question, then that would be forgetting things which are true.

However, the court action is directed at the association created by Google between a particular person and a page. By maintaining this association, Google are basically stating "this is one of the most relevant thing about person X", and if what it points to is irrelevant/out of date (even if true) then the result is false.

The right to be forgotten is not about making the world a better place. It is about permitting people to behave badly without consequences.

No, it's about requiring search engines to stop returning irrelevant items about a person when asked for relevant items, and as a result causing harm. Without this law, search engines could report results which are false and do harm with impunity.

If we are going to have some kind of right to be forgotten then it should be judged by independent specialists, pages that should be 'forgotten' should be added to a public blacklist used by ISPs so that it can be checked for abuses

You misunderstand, it's not the page that should be forgotten, but the association created by google between that page and a particular person. Basically, you are effectively asking google "What is the most relevant thing about person X?", and google are returning irrelevant/out of date information. The result due to that association is within Google's control, and that association is what the court is addressing, not the existence of the page itself.

Currently the search engines must remove the link to the article, but the article stays. This is bullshit. If the article contains something to be forgotten it should be removed or redacted. This is the only correct way to do it. Also, there should be an open procedure, with appeals, to decide if the article must be redacted / deleted.

One of the problems with this law is that it's badly named, which creates a lot of misunderstanding. For example, I do not believe search engines must remove the link to the article completely, just not return it for specific search queries. So it's not the article itself that is at fault, but the connection created by the search engine between the article and the search subject

Basically, when you search Google (or any other search engine) for the name of a person, you are implicitly (or explicitly) asking "What are the most relevant things about this person?" If the answers to this query chosen by the search engine are out of date or irrelevant, then the search engine itself is at fault, as it is the one doing the judgment of "most relevant". When these inaccurate results/connections cause damage to a person, and the search engine won't change its practices, then it is only right that the court stepped in.

Would there be as much of a backlash if this law was called "Requirement for search engines to stop returning inaccurate and harmful results about people?" Maybe not, there might still be people saying "But, the article is correct". But this misses the point that the inaccuracy is in the search-engine-generated and ranked connections between articles and people - not in the article itself.

M.

Actually what he bought was a laptop with a refundable copy of windows. If - after the sale - the vendor is claiming that what he bought is something different, i.e. a laptop with a non-refundable copy of windows, then it is the vendor that is in the wrong.
It's not about forcing or not forcing people to buy something they don't want, but requiring vendors to be honest on what they sell, and if they sell an item which has conditions, they are not allowed to unilaterally change those conditions to the detriment of the buyer after the event, especially if the buyer relied on those conditions before making the purchase.

Or how about having a large number of people walk around with phones on, with their path describing the word "N O" or similar?

It may be a licence agreement, but as it the licence is for distribution, not use, it would be a LA, not a EULA.

From my understanding of the OP, you wouldn't have to expect to be paid at all for this work, just to reduce the "value" of the asset containing the personal information, and/or link the obligations of the usage of the date to the ownership thereof.
Currently the personal information is just seen as an asset worth X, with the liabilities of the same asset (the obligations of how it's used) seen as completely separate (and not taken over). If you could get them linked officially, then the people buying the personal information would be required to buy both the data and the obligations.
For a very rough comparison, if you bought a property lease from a company, I wouldn't expect to be to choose to just take over the "right to occupy for the next X years" and not accept the "obligation to pay rent for the next X years" as they are inextricably linked. If the same could be done with the personal information, then the privacy obligation would have to transfer across, as it woudl turn from selling just "a database with personal information" to "a database with personal information with the following liabilities/obligations"

Surely it should be the responsibility of the publisher, as they would be the ones that know what doesn't work a second time? So long as the seller marks it as "used", what else reasonably could they do?

Unfortunately you missed out the [vital] first part of the faq which is:

So, you are only required to make the source code available if your original distribution was in the form of binary+source offer. (3b of GPL) If you made the source available with the binary (as per 3a) you would not have to include that offer, and the offer would not be there to pass on to any third party.

Actually, it is not wrong. If Oracle chose to distribute the source at the same time as the binaries to those they sell to, then they would _not_ be required by the GPL to supply the source to any third party, and they would still be adhering to the GPL.

This is because there are three options in section 3 for source code access. 3a is to distribute source with the same time as the binary. 3b is to distribute the binary with an offer (valid to any third party) and 3c is to distribute the binary including the original offer (if you yourself received this as per 3b, and are distributing non-commercially.)

Thus, if the software was distributed as per 3a (binary+source) to someone who purchased it, then third parties would have no right to request the source per 3b from oracle, and the infringement would by the purchaser for not including the source as per 3a (as they cannot pass on an offer as per 3c, as they didn't get it by 3b themselves)

At a guess:
a) Precedent. They don't want anyone getting the idea that suing them for the "right" amount (i.e. low enough) will result in a payoff
b) Cost. is probably not a large factor, as they're ibm lawyers anyway, I believe, and even if costs are incurred, I'm sure they're prefer to pay this money to their legal team than to the shareholders of the company on the other side of the litigation.
c) Logistics/legality, I'm not sure if it would be as straightforward as you suggest, as unless you have 100% of the shares, there are laws that protect minority shareholders, so it might be difficult to drop the case from the sco side without opening a can of worms from any remaining shareholders. (and there would doubtless be a few that refuse to sell if the above happened, hoping (somewhat logically) that if they were prepared to buy out the whole company to avoid litigation costs, that they would do the same for the last few shares for the same reason.

It was also reported by the bbc http://news.bbc.co.uk/1/hi/england/london/8293784.stm

How exactly would this be in breach of the DPA?

Since they have proven that they can remove the copy from the user's device (by doing so) if they said they could not, that would not be the "correct answer", it would be a lie. And, if the failure to remove the infringing data was a "will not", not a "can not", it would seem to be trivial to prove that any further infringement (by keeping it on the device) was wilful. (if they could remove it but _chose_ not to)