As I flagged yesterday, yesterday morning the appeal judgment in the IceTV case was handed down by the Full Federal Court. This truly was an all-stars copyright case: interesting issues, an all-star IP bench (including the CJ himself, plus two senior IP heavies – Justices Lindgren and Sackville) (note too – the same bench that sat on the Desktop Marketing case), no less than 4 senior counsel (all of them IP heavies in their own right) plus juniors.

I’d give you background on the case but I don’t really need to – Peter Black has already done that albeit a while ago now, and so have I. In summary, the case raises the question whether Channel Nine could use its copyright in its television program guides to prevent a rival publishing an electronic program guide compiled from a combination of recording what was shown on TV, ‘prediction’ and updating from published guides. Nine lost the original case but have won on appeal: that means they have regained their rights of complete control over programming guides. From a competition perspective, this raises some interesting questions which Joshua Gans has discussed before. Over the fold, my preliminary thoughts on the copyright ruling.

Why does it matter (why such an all-star cast?)
There’s a couple of reasons why Nine (and other channels) are keen to control this information: first, because they license it to lots of people (presumably for money), and second, because control means they can impose licensing conditions on use of the information. And that is important because personal/digital video recorders (PVRs/DVRs) need programming information to work. If you can control use of the information, you can start to control what functions will be allowed on these devices (subject to market power type stuff). And that means you can impose some limits on functions like ad-skipping. Which of course is something that Free to Air channels are very concerned about. I have a vague recollection, which I’ll have to check, that some kind of industry deal has been reached on access to programming guide information, though, so I’m not sure to what extent this control is real.

So, what do we think?
In summary? A judgment that is consistent with the trend of Australian authority, and entirely ['entirely' is a bit strong for a preliminary view!!!] consistent with the reasoning of the (identical) Full Federal Court in Desktop Marketing. But a judgment which also illustrates how far away, in some respects, Australian copyright law is from copyright law in other countries, including the US (with its concepts of thick and thin copyright), and the UK now that the UK has the Database directive. And a judgment that emphasises that Australian copyright law truly does protect information, and not just expression (the judgment is notable for not even talking about that issue). Those would be my preliminary thoughts.

Oh, and this too. The whole judgment skips a more interesting issue on the facts: one that David Lindsay talked about in his talk at the Copyright Symposium last year. Namely, whether sitting down in front of a television, writing down what is being shown, is an indirect means of copying the original program schedule created before those original broadcasts were sent out.

The Full Court has held:

  1. That the program guides are protected by copyright, and in particular, that copyright protects the originality reflected in the programming decisions made by Nine – not just the form of the table;
  2. In assessing whether what was taken infringes copyright (whether a substantial part is taken) you take into account what is ‘original’ about the compilation – in this case, the programming decisions made by Nine employees. Copying of that time/title information – even bits of it – takes the ‘heart’ of the work here. In other words, in a compilation, you can look at how important the information copied is.
  3. That when Ice updated its guide using time and title information from the Nine Weekly schedules (indirectly, through published guides) it copied a substantial part of those program guides, infringing copyright.

In many ways, there’s a lot about this judgment that I think is pretty expected – particularly the first bit. The insistence of the Full Court that you should not, in looking at copyright protection for a compilation, and looking at what copyright protects about the work, try to draw some artificial line between the work that goes into selecting programs to be shown, and putting together the final table, I think is entirely consistent with authority (Olympic Amusements v Milwell springs to mind, as well as the other cases cited by the court). That, I think, while important, is not the most interesting thing about the judgment.

I’m inclined to think that the most interesting aspect of the judgment is how it builds on the Desktop case. For those who are not copyright geeks, Desktop is the 2002 decision of the Full Federal Court which held that a person who copy-typed the White Pages into a database and distributed that was infringing copyright. Importantly, it held that sweat-of-the-brow compilations are protected copyright works in Australia. But in infringement terms, it really was a pretty simple case: it involved wholesale copying.

Where IceTV different is [on the facts and arguments accepted and focused on by the courts - ie on the assumption that the initial template of the programming guide was produced in a non-infringing way] is that in this case, the company copied bits of the compilation – basically, the time and title information of updates to the basic program guide. They did this repeatedly and consistently, but these were, in the end, a series of bits of information necessary to keep the information current and correct. What the judgment of the Full Federal Court essentially says is that if that information is important enough, then taking it can infringe. This approach, I think, potentially gives quite ‘thick’ protection to factual and other elements of compilations.

Now, of course, an obvious response to that is that these weren’t objective ‘facts’ (ie facts out there in the world) – rather, they were ‘created facts’ (facts of creative decisions made by the broadcaster) – and it is the creativity that goes into making those facts that copyright is protecting here. It could therefore be argued that for truly factual compilations – chronologies, telephone books etc – the analysis of infringement in the case of small takings would be different. Maybe. That’s going to be a difficult line to draw in real world cases.

Another point of interest is that in reading the judgment, there’s a really interesting series of paragraphs (109-113) where the court is talking about what IceTV took. Reading those paragraphs, I got the sense that the court was a bit uncomfortable with the finding that only individual time/program title bits were being copied, and that much of that information was not ‘copied’. On the one hand, the court notes that ‘Ice did not take all its time and title information from the Weekly Schedules … the so-called ‘predicting it over’ process adopted by Ice enabled it to construct the starting templates’. But then the court says that ‘those templates … lacked the crucial time and title information that would make the templates useful’. Now, I would have expected the court to say that the templates lacked some of the crucial information, or parts of that information. It might be a mistake to read too much into it, but I can’t help but wonder whether the court wasn’t entirely convinced that the whole template wasn’t in some sense copied. Maybe that takes it too far. Interesting to speculate though.

[update Friday arvo]:
But what about the indirect copying?
Here’s another weird thing about the judgment. I had the pleasure of catching up with my old friend David Lindsay today, and we were talking about the paper he gave at the copyright symposium last year, and the point he made in that – one that most of us talked about when the original judgment was handed down – namely, what about the indirect copying?

There’s a long line of authority that says you can copy copyright material indirectly: you don’t need to see the original for you to be copying a copyright work. For example, in the Plix case, someone was held to have copied drawings of kiwifruit containers, when they produced similar kiwifruit containers based on regulations that specified, in detail, the dimensions/requirements of the containers. This was copying even though they never saw the original drawings.

In IceTV, the original template for the program guide was said to have been produced by someone watching television for a long time and writing down everything that was shown. This method was used, in effect, to record the program guide. After that, they used predicting software and little updates to produce new program guides. The Full Court focuses on whether the updates are infringement. But that’s moot, of course, if the original template was produced through copyright infringement.

So the issue is: are you copying the program guide when you write down what was broadcast? Or has the causal link between the original and your version been ‘broken’ by the fact that you are writing down what was shown, not copying from the compilation? Now, that’s a hard question. I don’t think it’s obvious at all. To address it would require delving into some quite tricky issues.

So it’s very notable that the Full Court says, in effect, nothing about that argument. And I can’t help but wonder again, out loud and speculatively, whether something of a background concern about that is reflected in some of that language that suggests the court thought that more than just the updates were copied. Hmmm.
[end update]

What are the implications?
For the broadcasting area, obviously, the implication is that pretty much no matter what you do (and IceTV really did do quite a lot to try to avoid copyright infringement here, on the facts found by the trial judge) – you simply cannot produce a television program guide without the cooperation of the broadcasters. Maybe from a policy perspective it’s a good thing that broadcasters get to license (and get to get paid) for this stuff – in the sense that licensing it clearly is a part of their business model.

On the other hand, it has some pretty strong implications for related markets. As I mentioned at the start, the obvious one is the market for DVRs/PVRs. If we do end up with less fully-featured devices in the Australian market (which seems to be the situation now, at least) – well, you know what to blame. Copyright law, of course (and the broadcasting regulation system which has given, and continues to give, such power to the free to air channels).

More broadly, I’m sure there are some other businesses who are pretty pleased by the judgment. In the UK, the arguments that suggested that broadcasters did not have rights of control over programming information were based on ECJ decisions that the betting agencies and sports authorities did not have extensive rights over their info (for example, of racing fixtures) used in sports betting. While the decision in Desktop no doubt gave these guys a lot of comfort (and the physical/timing differences/difficulties in getting prior access to info probably protected the


Oh, and just one other weird things in the judgment

There’s just one other thing I wanted to mention which I thought was really weird about the judgment. There’s this little throwaway line in paragraph 117, which says this:

We assume [for the sake of argument] that some similarity in arrangement or form must be shown to establish that Ice reproduced a substantial part of Nine’s copyright work, although the assumption is doubtful in the light of the definition of ‘material form’ in s 10(1) of the Copyright Act.’

In other words, it seems that the court is suggesting that if you’ve copied copyright material, your end result does not need to look the same as the original – and that is because the Act includes a definition of material form which allows for electronic storage of a work (copyright infringement requires a ‘reproduction in material form’). Frankly, that’s a bit odd. The definition of material form wasn’t introduced for that purpose – if I remember rightly, it was put in there to ensure that all those electronic copies stored in, for example, computer hard drives counted as ‘copies’ of works. Obviously, when they’re in storage they look nothing like the original copyright work (when you store a book in electronic form, the copy in memory doesn’t look much like the book. Dur!). The requirement of ‘objective similarity’ between infringing work and original comes from the concept of what counts as a reproduction – it’s a common law requirement, that I don’t think the legislature intended to do away with. That said, Desktop Marketing also made pretty clear that in compilation cases, the form may not be particularly important when all the information is reproduced and reproducible – that implication arises from the particular nature of compilations, perhaps. But let’s not throw the baby out with the bathwater here. Weird. Just weird.