Stupid patent tricks [LWN.net]

May 19, 2004

By Pamela Jones, Editor of Groklaw

If you read about Microsoft's patent number 6,727,830, "Time based hardware button for application launch," issued on April 27, 2004, you're probably thinking that now you've heard everything. A patent on double-clicking and on holding down an application button? How can the Patent Office issue such a patent, since you can probably think of several instances of prior art off the top of your head?

Here is the abstract, explaining the patent:

A method and system are provided for extending the functionality of application buttons on a limited resource computing device. Alternative application functions are launched based on the length of time an application button is pressed. A default function for an application is launched if the button is pressed for a short, i.e., normal, period of time. An alternative function of the application is launched if the button is pressed for a long, (e.g., at least one second), period of time. Still another function can be launched if the application button is pressed multiple times within a short period of time, e.g., double click.

It drones on a while longer, but although they expend a great many words, what they have patented is simply this:

You have a device with a button. If you press the button quickly, you get different behavior than if you hold the button down. Electric typewriters have worked that way for decades, and portable CD players for many years.
Press the button twice in a short period, and you get different behavior. This, of course, is the "double click."

Linux systems are full of software which implements the claimed behavior. Double-clicking is found everywhere. The "hold the button for different behavior" can be found in places like the CD player. The patent specifies a "limited resource" computing device, so they are talking about PDA-type systems. The simple fact is that all computing devices are "limited resource," however.

I asked my right-hand man on Groklaw, Dr Stupid, if he could think of any prior art and he had no trouble in about 10 minutes coming up with these possibilities:

The general concept of short press and long press doing different things is not new at all - many embedded devices use it. What the patent is about is a particular use of this concept for launching programs on a device. That is, clicking once on the icon launches the program with one command line parameter, and a double-click a different command line parameter. Or a hold down is yet another.

The very old FVWM window manager for Linux has a 'maximize' button which works like this:

Click-and-hold: stretches the window vertically.
Single Click: stretches the window but does not cover the taskbar or button bar.
Double Click: stretches the window to cover the entire screen.

So you have normal action, hold-it-down action, and double-click action. It's still shipped with SuSE and most distributions to this day, I believe...

To me, it fails the 'not obvious' test. Another one that I wonder if it might be relevant is here:

'If you wish, you can distinguish single, double, and triple clicks. A double click means clicking a mouse button twice in approximately the same place. The first click generates an ordinary click event. The second click, if it comes soon enough, generates a double-click event instead....'

Microsoft listed 8 prior art documents, each slightly different from theirs. But then you find a long list of what they asserted was unique to this patent. But, without analyzing this one in great depth, certainly we can agree there are patents issued that should not be issued, and the real question is: why does the Patent Office issue them? And why do companies want them?

The answer to the first question is simple: they are understaffed and there is a general policy that you do your best and later the courts can determine if the patent was valid or not.

Why do companies want them? I asked that question of patent attorney Dan Ravicher, head of PubPat, the organization that is dedicated to going after patents that were wrongly issued, and also asked about this specific patent, and here is what he told me:

When I read those claims, I was like, sure, nice try. I doubt Microsoft would ever assert this patent. But, there is still value in building up a portfolio because many valuations are based purely on the objective factor of how many patents or how many claims one has, despite the fact that a wide swath of them are useless. The valuation experts aren't that sophisticated, yet.

A patent, in other words, is an intangible, and you look good to valuation experts if you have a big pile of them.

Does that mean there is no danger? Should something be done? He told me that until Microsoft begins to assert the patent, which so far it seems not to have done, the best thing is just to monitor it. "If Microsoft begins to assert this patent specifically, then we'll review the situation and make a decision about how best to protect the public," he says.

(Log in to post comments)

You want prior art?

Posted May 20, 2004 16:41 UTC (Thu) by spitzak (guest, #4593) [Link]

I have prior art in my own software. Check fltk for the Fl::event_is_click() function. This
function was in there since 1998, and possibly earlier. When the user pushes the mouse
button down it is true, and it remains true for a short amount of time. Checking this on
release can distinguish between a "short" and "long" click.

Grepping through my code for event_is_click() I find the following uses. All date to long
before this patent:

I used it to distinguish double click from hold-release-click.

I also use it to change the behavior of popup menus, a click will just post the menu on
the screen, a hold will instead pick the item the mouse is pointing at.

I use it to make a button that you can click on to increment a variable, and can drag to
change the variable. Holding down the button and not moving does not change the
variable and is different than a click.

I have used this to distinguish between raising windows on click and moving them in the
flwm window manager (you can push, hold, and not move the window, and it does not
raise).

I use this to distinguish between clicking on selected text just moving the insertion point
and starting drag & drop (OS/X also does this). (this usage may date only to 2002).

I use it in the fluid GUI designer to distinguish between dragging a selected set of widgets
and just selecting the one clicked on. Holding down without moving does not change the
selection, while a quick click does.

Nuke (Commercial software from Digital Domain) uses middle-mouse click to restore the
zoom in windows to the default, while holding down the middle mouse either causes
nothing to happen or causes the zoom to go to 1:1. This behavior was copied to other
software using the Qt toolkit, indicating that this behavior can be copied without using
fltk. This behaivor has been in there since 1994 and has been observed and used by
several hundred artists at Digital Domain.

Nuke allows the user to drag an animated parameter from one text field to another to
cause linking. This is used to distinguish an attempt at a drag from a click. This code
was written *by a different author* indicating that I am not the only source of this
behavior! In a similar place a button can either pop up a menu or start a drag and drop
depending on how long the user holds the mouse button down.

In the main "DAG" view of Nuke clicking on a "node" selects only it, but holding the
mouse button down on a selected node drags it and all others without changing the
selection. Again, holding the mouse button down and not moving it will not change the
selection and is thus prior art.

The Nuke image viewer uses this to distinguish between moving a permanent
color-indicator to the mouse, or just temporarily displaying the color under the mouse.

Some Nuke viewer versions (disabled in the current one) used this to distinguish between
deselecting all selected objects and changing the point at which Nuke is updating the
viewer to be the line the cursor is pointing at. (this does not work in current ones due to
internal inability to support it, the GUI will be restored as soon as possible).

The "is_click" is also visible in the Nuke SDK for making plugins, even though you are
not allowed to call fltk from that. I did not bother to search the plugins for usage...

I believe many systems already have prior art by using methods identical to mine for
distinguishing a double click from a push-hold-release-push sequence. Holding the mouse
button down for a longer time on the first click converts that click from "possible double
click precursor" to "not a double-click precursor".

Patent is not on double click in general

Posted May 22, 2004 19:41 UTC (Sat) by giraffedata (subscriber, #1954) [Link]

The patent doesn't claim double presses in general. The most famous prior art for the double press is Microsoft Windows!

The patent is upfront about the fact that double presses already exist. The claim is for a specific use of a double press.

The patent claims use of a double press to "launch" a different "function of an application." It also speaks of a "hardware button," which I get the impression is meant to allude to a button on a watch more than a mouse button. It's all kind of gibberish, but I agree with Dr. Stupid that what they're trying to describe without getting too specific is a button on a device whose entire user interface is just a few buttons. Pressing a button does something that would be called running a program in a full-size computer, and what arguments that program gets depends on how many times you press.

Looked at that narrowly, I can't think of any prior art. But it does have the obviousness problem that most "software patents" have.

Patent is not on double click in general

Posted May 25, 2004 15:16 UTC (Tue) by hazelsct (guest, #3659) [Link]

PalmOS has done this for many, many years. Clicking the contacts button launches the contacts application, holding it down launches that application to beam the owner's "business card" to another user.

There is no doubt a ton of prior art here. The idiots in the patent office who granted this need to be fired, let alone software patents thrown in the dustbin of "intellectual property" history.

Patent is not on double click in general

Posted May 27, 2004 10:58 UTC (Thu) by mmacok (guest, #20088) [Link]

My cell phone is Nokia 6310i.

When I shortly press the power button when the device
is off, nothing happens. When I press it for a longer
time, the device is switched on.

When the device is on and I press the power button
for a short time, the menu with different "profiles"
appear. When I press it for a longer time, the device
is switched off.

Stupid patent tricks

Posted May 20, 2004 18:44 UTC (Thu) by hathawsh (guest, #11289) [Link]

There are a lot of patents that should be monitored. Is there a database of bad patents somewhere on the web? It would be useful to simply record which patents are bad, and not worry about paying lawyers until patent holders attempt to enforce bad patents. This would benefit two groups of people:

1) People who are concerned about abuse of the patent system could register their complaints about specific patents in this database. Since no attorneys would be involved at this point, the only cost would be for Web hosting.

2) People who get sued for infringing bad patents could have an instant resource for defending themselves.

So, does such a database exist? If not, there are a few things a new site could do to attract attention, like provide an RSS feed of new patents.

(BTW, related to this specific patent, car stereos have done this for a long time. Push a button to switch to a new station, or push and hold the same button to associate that button with the station you're currently listening to.)