Analysis: Apple turns the flamethrower on Android | Hacker News Reader

Analysis: Apple turns the flamethrower on Android

theregister.co.uk

47 pointsyread48 comments

Ahh, Apple. Do you realize that your only value for most of my friends is that your OS is actually a NIX beast?

I wish someone had enough money to come up with decent UI on Linux (pardon me but GNOME looks and works horrible), better typography and write some general, and widely working, framework for supporting Windows drivers.

It should not take more than $20-30m I think. Ubuntu, unfortunately, is not aggressive enough in those, critical, areas.

It seems like that would have been the killer Linux differentiator for some distribution. Design a decent UI that isn't Windows, OS X, etc. UI's are hard, but it seems like there are more ways to do things out there that aren't being explored.

A small and light form factor tablet with a touch interface and a way to hide away a EVDO USB modem would probably sell very well.

I would say skip the big stuff and just do a netbook or netop that is rock solid and requires no user expert knowledge.

Don't see you can't do this completely until you go to a completely touch interface and put up an App Store.

Did you miss the reports about the estimated value of the Linux kernel being about $1.4 billion?

What makes you think that completely overhauling GUIs on Linux to create a new desktop environment and adding binary compatibility with a completely different (closed-source!) OS would cost less than 2% of what it would take to write just the kernel?

Developing a UI framework with comparable technical capabilities to what OS X provides might be doable starting from the best of what's available at the moment given your budget of $20-30m, but building a cohesive desktop environment on top of that would be an extremely huge undertaking.

It will be 2 billions in ten years when linux + new gui has conquered the desktop (or what is left of it) and mobile devices. The difference between 2 billions and $20-30m (or whatever imaginary sum you suggest) is called profit (or rather some share of it).

It does not require major overhaul of the kernel.

I agree that Windows' drivers compatibility is a hard task but I imagine there might be some hack. For instance, you could give up some speed, using virtualization.

If such hack is not available (patents?), it is always possible to come up with a set of rock solid supported devices, might be with some cooperation from hardware vendors. It is exactly how Apple solved this problem, and it seems that regular people are happy to accept such restrictions, as long as everything works great.

Right now, the problem is that if you install Linux on your computer, the official story is that "Linux supports more hardware than any other OS". Last time I checked (6 months ago), though, hibernation is still problematic, and you have to Google to ensure that your printer will actually work.

As for the UI. To be honest, it is not a technical problem. There will be some technical work involved but it does not matter that much. Linux development focuses too much on technicalities. (I write it as software developer)

The biggest issue are fonts. Every time I launch Linux, my eyes hurt. Few hours of work are required to get fonts right and they are still not that good as on OS X or Windows. There is a lot of typography know how available right now, and I bet you'd easily find a top notch typography agency or folks to solve this problem.

The second issue is design and basic usability. You do not have to rewrite everything, initally. The things I mean is fixing many small issues, such as mismatched padding everywhere, and UI inconsistencies, as well as running serious usability studies to find how regular people cope with the basic tasks.

I think it would be the best to pick XFCE for such improvements. Exactly like big corporations picked WebKit instead of Mozilla's engine. XFCE is modular, still relatively small, not that bloated as GNOME, and you could fork it without too much hassle.

Also, it helps a lot that most of the world switches to the web. You longer need to write all possible tools, except good integration with the web. I dream about something like transparent Google Gears integrated into the WM.

Paradoxically, I'd start work with porting TextMate to Linux. VIM or emacs are a bit too hardcore for regular developers, and there is no good editor on Linux for dynamic languages, which is light weight, nice and works perfectly within a minute, or two. Developers are crucial part of the infrastructure and Linux does not attract developers with eye for usability.

I agree that I might underestimate the costs. I don't think, though, that it is going to take more than $50m, and money is not that big problem, actually. The biggest problem is to realize that current Linux experience sucks a lot for people who do not want to spend their valuable time tweaking settings. The improvement is enormous but it is merely a start.

As for multi billion corporations that started and failed, it is not true. Right now, Linux's development is focused on getting servers right. The corporations succeed in this area. RHEL is de facto corporate standard. No one relatively big, except medium-sized Canonical, cares about your mother's experience.

BTW. To be honest, I was very surprised it is only $1.4 billion when I read it. Given the scope of the project, I expected few times more. But this is me.

> "I wish someone had enough money"

Multi-billion dollar corporations have tried and failed, I honestly don't think this is a problem of money.

I use Linux as my main desktop and I'm a professional worker. Works just fine, I do hate the Windows and Mac UIs though, so I can understand people have different tastes. I try not to make general statements about other UIs, just different tastes and opinions. Glad you're happy with Mac or whatever it is you're using.

A system in which a software module called an event consumer can indicate an interest in receiving notifications about a specific set of events, and it provides an architecture for efficiently providing notifications to the [event] consumer

Hmmm, Smalltalk implemented this. All objects in Smalltalk implement listening and notification. Apple licensed certain concepts from Xerox, so this may or may not apply. How about Java? Do Java programs that use its notifications violate the patent?

Or what about Javascript libraries? How about Windows programs? It seems to me that event-driven architectures with listeners are everywhere.

This is the natural consequence of software patents. Just as the inner workings of a physical engine can be patented, the inner workings of software can be patented, which means things like patterns.

I wonder if future versions of books like the GoF patterns book will be annotated with warnings that some or all of the patterns are covered by patent and must be licensed?

I wish more people would read the actual patent claims before coming to conclusions about them based solely on their titles. It seems that many news organizations have articles claiming this is a ridiculous lawsuit based on the titles which they claim are overly broad. If however they took the time to read the patent claims, they would see that they are relatively specific. For example: http://www.google.com/patents/about?id=L-IeAAAAEBAJ&dq=5...

Now I'm not saying that software patents are necessarily right, but I am growing weary of people making broad generalizations based on the titles of these patents and having them perpetuated. That is somewhat like a book review discussing the book's title and deciding whether or not it's a good book without having read the book or having any knowledge of its contents.

In other words, leave it to the lawyers. If you went to LN (Lawyer News - doesn't exist, but let's pretend for my analogy) and there was a lawyer claiming that Javascript is stupid because it has the word Java in it and therefore must have all the attributes of Java and he hates Java, we'd all cringe at his lack of knowledge of the subject and wish he'd leave it up to the Hackers who actually know what they're talking about to criticize.

That patent is on NeXTSTEP's NSDistributedNotificationCenter [1], which admittedly might actually be enforceable (I'm not sure of any prior art in this regard).

That said, after reading the claims of a lot of the other patents Apple is citing it is apparent that they are complete rubbish, for example [2] (the observer pattern applied to a desktop environment), [3] (generic proxy objects for IPC/RPC - Smalltalk/CORBA did it), and [4] (WTF???????).

1. http://developer.apple.com/mac/library/documentation/Cocoa/C...

2. http://www.google.com/patents/about?id=zIwLAAAAEBAJ

3. http://www.google.com/patents/about?id=7d4aAAAAEBAJ

4. http://www.google.com/patents/about?id=HrobAAAAEBAJ

When you say "rubish", are you going off the description or the actual patent claim? Descriptions are always written in broad strokes and never really seem enforceable.

Sure, broad generalizations are always a problem, but in this particular case I'm not sure that it matters. The patent description that you linked to doesn't really seem specific enough to make it unique. It seems to be talking about eventing systems in an operating system (as opposed to a language). The idea that there isn't prior art seems dubious.

Also, I think leaving it to the lawyers is part of what got us into this situation in the first place. We have a bunch of lawyers and patent clerks that don't understand the technology so we end up with a lot of silly patents that shouldn't exist.

I can't upvote this enough. Making any conclusion about a patent without concentrating on the claims (at least the independent ones) is ignorant. I admit that I used to do the same, but once you understand the mechanics of a patent (just like any technical system), you are immediately put off when someone is angry over a summary.

Your general point is extremely important and deserves to be underlined in every patent discussion: summaries and abstracts are irrelevant. The patent is the claims.

However, in this case, some of these patents are still quite broad. Almost offensively so.

In other words, leave it to the lawyers.

Perhaps I really shouldn't discuss patents, intellectual property law, or lawsuits here. Even if I read the whole thing, I'm not a lawyer and I don't understand the specialized language being used. I don't know the specifics of the law in different jurisdictions. I really shouldn't have said anything.

What it amounts to is this: The text of an article is insufficient information for discussion, you need expertise to comment on it.

Now that I put it that way, why are any of these articles discussed here on HN? Who is actually qualified to come to conclusions or espouse an opinion about economics, politics, or law? Why do we tolerate discussion about intellectual property law, anti-terrorism laws, or economics by unqualified persons?

It's quite disturbing how little any of us know about things that affect our livelihood.

I have just read the claims of the patent you linked. The first 3 claims [1] describe events (basic MVC stuff) as has been used in the wild since at least 1979!!!

The patent has a filing date of May 13, 1994. I had written Smalltalk code in use at Fortune 100 clients using these very same techniques for several years prior to this filing date. At the time, I was clear I did not invent these techniques.

Its my understanding there are penalties for an inventor listed in a patent fraudulently claiming they created the invention. Perhaps its time we started enforcing these penalties. I find it impossible to believe that any engineer at Next or Apple did not understand they were not the original inventor of what is claimed in this patent.

[1] first 3 claims from http://www.google.com/patents/about?id=L-IeAAAAEBAJ&dq=5...

1. In a computer including at least one event producer for detecting that an event has occurred in the computer and generating an event and at least one event consumer which needs to be informed when events occur in the computer, a system for distributing events comprising:

storing means for storing a specific set of events of which said at least one event consumer is to be informed; event manager control means for receiving the event from the event producer, comparing the received event to the stored set of events, and distributing an appropriate event to an appropriate event consumer; and distributor means for receiving the event from the control means and directing said control means to distribute an appropriate event to an appropriate event consumer. 2. The system according to claim 1, wherein said distributor means comprises a distributor module for each kind of event possible in the computer.

3. The system according to claim 1, wherein a plurality of event consumers are included in the computer and the plurality of consumers comprise:

broadcast consumers having no relationship with other consumers, the broadcast consumers operating independently of other consumers and of the order in which consumers are informed of the event; and sequential consumers having relationships with other consumers, the sequential consumers requiring that no other consumer be told about an event while they themselves are processing the event and having an ability to influence when they receive the event relative to the other consumers.

The first three claims are establishing the basis for claims 4 through 9. Likewise claims 11 through 17 depend on claim 10 and claims 19 through 24 depend on claim 18. Though they are written linearly, patents have a hierarchical structure.

What you appear to be doing is mistaking each claim for an independent claim of invention, which is a common mistake, but a mistake nonetheless. When the claim says "the system according to claim x", it is growing more specific as to what is actually being claimed as the invention. With that in mind it is not surprising that the independent claims (1, 10, 18) would not be novel. You have to read as far as the tail end of each thread to understand what is actually being claimed as the invention.

I can't speak to the content of this patent when read correctly, but I suggest re-examining it with this in mind.

Thanks for the pointers. I have written a few patents. Well, I wrote the invention part and paid supposedly good patent attorneys to clean it up and write the claims and educate me on the process.

I "thought" my lawyers told me the first few claims (1-3?) have to stand on their own and the rest can build on the invention. Meaning, if the first few are not unique inventions, the whole thing fell apart. More info would be useful.

I wouldn't doubt your lawyers! They were probably right in a context that does not generalize exactly.

I am not a lawyer and speak from no authority but curiosity, so I could be completely wrong or misleadingly close to being right, but I think the main point is confirmed by this, from the "Claims" section of the first result for "how to read a patent" (http://www.bpmlegal.com/howtopat5.html): "Each dependent claim must be narrower than the claim upon which it depends." The full example is too long to quote, but I encourage you to read it.

The obvious problem with this is that it is an implicit convention in an otherwise very explicit kind of document. It's no wonder at all that it is not widely understood.

The article starts with some naive and clearly exaggerated statements about what might happen before finally getting to the real point, which is:

> The case will take years to resolve, but that's no bad thing for Apple: manufacturers are going to be hesitant to create Android handsets...

If Apple thought they could get a clean victory they would go after Google.

Perhaps they're going for the settlement to use as ammunition for going against Google?

I quite doubt Apple's going to attack Google directly. Their weird App Store rejections and now the HTC lawsuit (read, not Google) indicate that Apple prefers a passive-aggressive stance.

If Apple sues Google, Google will stop paying Apple for Safari search ad revenue and make iPhone users hate Apple by blocking Maps and YouTube and breaking iPhone Mail's Gmail button. I'm sure Apple wouldn't want that.

Google taking ReMail off of the App Store is a very nice stroke of theirs against Apple. This gives Android phones the big advantage of Google's gmail search.

This is misconception that is being repeated over and over. They ARE going after Google. It's very clear.

It's also silly to think that HTC is in this alone. If HTC loses, Google loses. They'll throw whatever money/support they need to to back up HTC.

Except it could hurt apple if HTC happens to also have a lot of nice little patents which the iPhone is infringing (not so far fetch). I have a feeling that this will never end up in court and a cross patent+cash deal will happen. Disclaimer: IANAL so whatever I just said could be total crap.

> Except it could hurt apple if HTC happens to also have a lot of nice little patents which the iPhone is infringing (not so far fetch).

Or Google, should they decide to go on the offensive.

I think the interesting point is that if HTC settles using either cross licensing some patent of their own, or by paying to license whichever of Apple's patents are upheld, the outcome will be to limit Android to those who can defend themselves similarly.

This might actually be good for innovation. Android - strong as it is, is not a leap forward from the iPhone, but it's now the center of gravity for most people who need a mobile os to go on their touchscreen devices. If there was more incentive to develop a truly radical mobile ux on top of Linux, it might be better for all of us.

If any of this comes to pass, then the open-source aspect of android will become much more important: development will simply shift to countries that don't have software patents

Well, some of the patents are hardware, and you still need to sell your device in countries that will recognize these patents to make it worth making.

I expect a settlement out of this, there just seems to be something else going on that is behind closed doors. It just doesn't quite make sense how it is being reported.

Well, some of the patents are hardware, and you still need to sell your device in countries that will recognize these patents to make it worth making.

Sure, but as long as someone is selling the hardware (perhaps even licensing the appropriate patents) the hacker/end-user would still be able to compile and load software to make it all work.

I expect a settlement too, if only because I can't imagine that HTC has enough money to invalidate all of the patents/show that they don't infringe on them.

After the Kodak / Sun patent suit, I wouldn't count on HTC's ability to invalidate any of the patents. The power patent might be unavoidable.

I'd say the Nexus One and Droid are leaps forward from the iPhone at least in terms of hardware capability. Remember that this is a suit against a hardware manufacturer, not the software developer.

If Apple was confident of their devices' ability to outcompete HTC's in the market, they wouldn't have bothered to file this suit. They're suing because they are afraid of these things.

If you don't need a clean victory to win, why would you pick a fight with the best paid lawyers in town and give away your hometown vs. foreigner advantage?

I think it's pretty clear they are going after Google. By going after one manufacturer you put the brakes on the rest. Either it was a piece of HTC technology that violated patent or it was Google tech.

BTW. I hate the idea of software patents and wish it ends badly for Apple.

"the use of a proxy object to communicate between threads"

WTF. How the hell do they get away with establishing non-obviousness for these things? (Wouldn't that cover OTP, Go channels, Java Channels, etc.?)

I have no beef with Apple per se (thus would have you wish that for all companies) but clearly software patent system is fundamentally flawed.

I would rather that law makers deal with the IP issues in a mature way then wish ill-will to someone using the current system. We really need to get some pressure where it matters (government). The courts do what they are told for the most part.

This article has the clearest description of a "patent arsenal", and it's purpose, that I have seen.

What could the Free Software Foundation be thinking about this, I wonder. Particularly patenting the hand gestures sounds really absurd.

Doesn't Apple know that one cannot proceed by clinging on the patents and legal protections alone.

Doesn't Apple know that one cannot proceed by clinging on the patents and legal protections alone.

They're doing no such thing. They're trying to slow others so that they can stay ahead.

For the longest time I refused to buy new smartphones because I don't agree with Apple's iPhone lockdown (If I buy a product, it's mine. Nowhere does it say I'm leasing an iPhone, Apple should have no say about what I do with it.) and because multi-touch was a make-or-break feature.

By keeping manufacturers afraid of implementing multi-touch Apple keeps consumers like me from going to them because they seem behind on the times.

If Apple thinks I'm going to buy an iPhone in the future because they're using legal methods to hinder the competition, they are dead wrong.

They know that some people would be bothered by this but they don't have any great reason to care as the competetive edge this might give them on the 99% of their market which does not care would far outweigh the damage done by bothering the 1%.

99% of their market doesn't even know about this trial, or even what software patents are. Only the one 1% that is bothered about this knows about and understands the issue. In the end it will be a big loss of money for Apple and its shareholders and HTC too. A lose-lose situation, well done Apple...

From the geek mindset, apple is the new microsoft. Apple has nice products, but seemingly agressive in their approach. Don't they realize that competition is good? Of course not...

Hmm,

Does this action actually make Nokia's suit look like a pre-strike in a war that Apple had implicitly declared?

It does indeed.

Reply on news.ycombinator.com