Hacker News Reader

Comment by wereHamster | Hacker News Reader

Please go read the patent, it describes an "apparatus and method" that is much more advanced than traditional tiling window managers. The patent doesn't even compare to simple terminal based tiling of sub-shells. It also describes a GUI to control the positions, and describes in detail how it's supposed to work. There is a section "Description of the Related Art" in the patent which makes references to existing (at that time, year 2008) approaches and why they are not good enough. Example: "[...] Windows provides [...] commands which are accessible from the Windows taskbar. However, these commands are not as intuitive as they could be and they do not provide a desired level of control when positioning windows".

View on HN

>There is a section "Description of the Related Art" in the patent which makes references to existing (at that time, year 2008) approaches and why they are not good enough

In that section, it mostly compares itself to Microsoft Windows' stacking (not tiling) window manager. There's no comparison to ANY of the already existing tiling window managers.

It was filed on June 2nd, 2008.

There were a ton of tiling oriented windows managers in place for many, many, years prior. Many of them supported automatic layout, and many supported a variety of grid layouts as depicted in the patent.

Here's a video from 1987, 21 years prior to the filing of this patent, showing the Siemens RTL tiling window manager: https://www.youtube.com/watch?v=MFd0FDv3fQc You can see the selected layout in the upper left hand corner of the screen. The video itself shows the user choosing where new windows will land, and the automated tiling/fit.

As mentioned earlier, a ton of other tiling window managers existed well before the filing of this patent. Many of them with automatic tiling, and several grid layout options. I don't see anything in the patent that sounds novel to me.

The patent doesn't appear to describe tiling window management, but the composition of a window layout using a number of menu options that integrate neatly with the existing environment's window controls, and that visually demonstrate their effect.

Indeed the patent describes a user interface to accomplish this that looks very similar to screenshots of Moom.

My experience with the USPTO is that statements like "I don't see anything in the patent that sounds novel to me" doesn't amuse them and they take this sort of thing seriously. Whether you agree with software patents or not, I think it's important to not be too dismissive of that fact.

[1]: https://www.google.com/patents/US20090300541#forward-citatio...

>The patent doesn't appear to describe tiling window management

It does though. The screenshots show grid options, and describe how the end user can place windows in the grid. It specifically complains about how MS Windows has overlapping windows. Grid + no overlap == tiling.

>Indeed the patent describes a user interface to accomplish this that looks very similar to screenshots of Moom.

Sure. The argument about "novel" is that the patent holder wasn't the first to do something "similar to Moom". The idea that 2008 was the first time someone had the notion of grid layouts and allowing an end user to place a window in them is just silly.

> The idea that 2008 was the first time someone had the notion of grid layouts and allowing an end user to place a window in them is just silly.

That's not what's being patented. The gui menu to select placement of windows is what's being patented.

Did that exist before 2008?

Yes, it did.

I know I used window managers with almost the same grid square select in the 1990's. Not able to conjure up a screenshot yet. I did find one that's very close, here: http://www.badros.com/greg/papers/scwm-extensible-wm.pdf

That is clearly a UI for dynamic window placement and sizing, just not with the exact grid dynamic, though you can define new constraints, so it's possible it can do it as well. It's not the one I remember, but it's hard to Google up old UI's in HN time.

Given some time, I'm sure I could find something exactly like it.

It may be silly, but remember that the US uses first-to-file, not first-to-invent.

That only comes up when two people file for patents that overlap. You cannot patent known inventions.

First to file, though, is just about the patent being granted. Once you start using your patent to bully people, prior art, even non-patented prior art, can be used against you.

Edit: "Once the patent has issued, you can challenge it at the USPTO by filing a request for reexamination of the patent"[1]

[1] https://www.uspto.gov/patents-maintaining-patent/patent-liti...

Edit2: It's also possible there are overlapping, earlier patents. This one looks promising...compare the drawings: https://www.google.com/patents/US5712995

> Edit2: It's also possible there are overlapping, earlier patents. This one looks promising...compare the drawings:

There's also:

https://www.google.com/patents/US5796402

That's not what 35 USC. § 273 says. It says that commercial prior use may be legitimate defence for that prior use, which isn't the same thing at all.

Seems like the patent is already by a Microsoft engineer: http://pdfpiw.uspto.gov/.piw?Docid=05371847&homeurl=http%3A%...

And look at the images too, they help you visualize what the patent is about.

And the patent pretty much describes the mouse based interface which Moom implements, and I can't think of any prior art of that.

You should be able to find prior art and invalidate many clauses of the patent, like the general idea of grid based layouts, especially driven by keyboard shortcuts and mouse gestures, but the general concept of a graphical drop-down menu for arranging windows will probably stay.

I used to use GridMove [0], which looks quite similar. This was certainly prior to 2008, though I can't find a version history right now.

0: http://jgpaiva.dcmembers.com/gridmove.html

Edit: first alpha version May 30 2006 (http://www.donationcoder.com/forum/index.php?topic=3824.msg2...).

> Please go read the patent, it describes an "apparatus and method" that is much more advanced than traditional tiling window managers.

I read/skimmed the patent, I cannot find anything described which is obviously novel. Care to provide a quote of these particular features?

A judge is not going to skim the patent. They will patiently read each and every independent claim, and evaluate whether Moom implements any of them.

In evaluating claims 1 and 2 (¶9) I imagine that a judge will get to this screenshot[1] and compare it to this diagram[2] and ask the authors of Moom for their justification. This justification can indeed be that it isn't novel, but I do not think the judge is likely to accept it because the patent examiner did indeed grant the patent.

In evaluating claim 3 (¶17) it may be that Moom is not infringing, but it is difficult to tell from screenshots and a video.

[1]: http://web.archive.org/web/20110309004337/https://manytricks...

[2]: https://patentimages.storage.googleapis.com/US20130219331A1/...

> because the patent examiner did indeed grant the patent.

I am not a patent lawyer but I don't think that how it works. The patent examiner is not qualified to do and does not do an exhaustive check for novelness. They mainly do a procedural check.

I've been through the process both in the US and in Europe.

"Exhaustive" has a specific legal definition.

The examiner indeed identifies existing work[1] and asks for an explanation.

They usually focus on existing grants, but they also look at other applications (that are published). They do not have to pay as much attention to un-filed prior art because the US uses a "first-to-file"[2] system.

[1]: https://www.google.com/patents/US20090300541#backward-citati...

[2]: https://www.uspto.gov/sites/default/files/aia_implementation...

From your comments in this thread, your understanding of first to file is incorrect.

First to file doesn't invalidate un-filed prior art or give it lower presedence.

First to file only really matters when there are overlapping patent applications. Basically a situation where one person files first but the other applicant claims they actually invented first but didn't publish the invention.

Had the person who claims they invented first published their invention, they'd have a year to file.

If that year was up, their prior art could still prevent the other applicant from being granted a patent the same as it would be under first to invent.

> First to file doesn't invalidate un-filed prior art or give it lower presedence.

I'm not saying it invalidates anything.

I'm saying this kind of arm-chair lawyering isn't going to impress a judge.

> First to file only really matters when there are overlapping patent applications.

35 USC 102 (b).

35 USC 102 (b) covers the exception that allows an inventor up to a year to file after publishing their invention. It prevents an inventor's own disclosure from being used as prior art to deny the inventor's patent request--provided the patent was filed within a year of the disclosure.

What do you think it says?

>I'm not saying it invalidates anything.

You said this:

>They do not have to pay as much attention to un-filed prior art because the US uses a "first-to-file"[2] system.

That implies that un-filed prior art was made less important by first to file, which is incorrect.

>I'm saying this kind of arm-chair lawyering isn't going to impress a judge.

Good thing I'm not trying to impress a judge. I'm just trying to stop the spread of misinformation on first to file.

No, unfiled prior art counts as always, as long as it's published somewhere. It's only secret prior art that stopped counting.

Also, the FITF system only became effective for patents filed on or after March 16, 2013, whereas this patent was filed back in 2008.

Yes, and they have an incentive to pass as many patents as possible (because more patents is more money).

Are you saying that patent examiners are paid for each application that is filed?

I don't even believe that the author knew about the WinKey/arrow key shortcuts. They let you size a window to half a single viewport without the rigmarole described in the application.

Reply on news.ycombinator.com