Re: Ambiguous text entry?

[seanadams]

Phil,

We can't use it if we believe it's patented.



On Thursday, November 20, 2003, at 10:30 AM, Phil Barrett wrote:

> dean blakketter wrote:
>
>> That's neat. When you do the entry, does it show just the numbers or
>> some matching string?
>
> Right now (or rather until 5.0, which means I'll have to do it again,
> which is no bad thing) just the numbers, but I have several ideas for
> refining it.
>
> As long as it isn't going to be pulled because of the patent issue.
>
> Phil
>
>
>>
>> On Nov 19, 2003, at 10:32 AM, Phil Barrett wrote:
>>
>>> (Forgive me if this has already been discussed and dismissed)
>>>
>>> The current Search mechanism for entering text is clumsy compared to
>>> single-key methods commonly used on mobile phones.
>>>
>>> I have implemented ambiguous text entry, where the user just enters
>>> 74688<right> to find 'shout' (or any other letter sequence with that
>>> mapping to number keys), and I plan to extend this using a
>>> dictionary of the actual words used in artists, albums and songs.
>>>
>>> But I am worried that this is technology covered by the T9 patents,
>>> which are heavily protected by AOL/Tegic. (To see the eleven patents
>>> in detail, go to www.uspto.gov and search for Tegic)
>>>
>>> What do we think? Is it too risky to include this technique in open
>>> source software?
>>>
>>> Disclaimer: IANAL, but I have had dealings with software patents.
>>>
>>> Phil
>>>
>>>

[Phil Barrett]

> We can't use it if we believe it's patented.

Indeed. That's what I thought.

Pity.


> On Thursday, November 20, 2003, at 10:30 AM, Phil Barrett wrote:
>
>> dean blakketter wrote:
>>
>>> That's neat. When you do the entry, does it show just the numbers
>>> or some matching string?
>>
>> Right now (or rather until 5.0, which means I'll have to do it again,
>> which is no bad thing) just the numbers, but I have several ideas for
>> refining it.
>>
>> As long as it isn't going to be pulled because of the patent issue.
>>
>> Phil
>>
>>
>>>
>>> On Nov 19, 2003, at 10:32 AM, Phil Barrett wrote:
>>>
>>>> (Forgive me if this has already been discussed and dismissed)
>>>>
>>>> The current Search mechanism for entering text is clumsy compared
>>>> to single-key methods commonly used on mobile phones.
>>>>
>>>> I have implemented ambiguous text entry, where the user just enters
>>>> 74688<right> to find 'shout' (or any other letter sequence with
>>>> that mapping to number keys), and I plan to extend this using a
>>>> dictionary of the actual words used in artists, albums and songs.
>>>>
>>>> But I am worried that this is technology covered by the T9 patents,
>>>> which are heavily protected by AOL/Tegic. (To see the eleven
>>>> patents in detail, go to www.uspto.gov and search for Tegic)
>>>>
>>>> What do we think? Is it too risky to include this technique in open
>>>> source software?
>>>>
>>>> Disclaimer: IANAL, but I have had dealings with software patents.
>>>>
>>>> Phil
>>>>
>>>>

[Kevin Deane-Freeman]

Quoting Phil Barrett <philb (AT) philb (DOT) co.uk>:

> > We can't use it if we believe it's patented.
>
> Indeed. That's what I thought.
>
> Pity.
another ancient system that is badly in need of a kick in the backside...
manipulating patents is actually being touted as a "significant" part of my
company's business, now. No shortage of careers for lawyers...

-kdf

> > On Thursday, November 20, 2003, at 10:30 AM, Phil Barrett wrote:
> >
> >> dean blakketter wrote:
> >>
> >>> That's neat. When you do the entry, does it show just the numbers
> >>> or some matching string?
> >>
> >> Right now (or rather until 5.0, which means I'll have to do it again,
> >> which is no bad thing) just the numbers, but I have several ideas for
> >> refining it.
> >>
> >> As long as it isn't going to be pulled because of the patent issue.
> >>
> >> Phil
> >>
> >>
> >>>
> >>> On Nov 19, 2003, at 10:32 AM, Phil Barrett wrote:
> >>>
> >>>> (Forgive me if this has already been discussed and dismissed)
> >>>>
> >>>> The current Search mechanism for entering text is clumsy compared
> >>>> to single-key methods commonly used on mobile phones.
> >>>>
> >>>> I have implemented ambiguous text entry, where the user just enters
> >>>> 74688<right> to find 'shout' (or any other letter sequence with
> >>>> that mapping to number keys), and I plan to extend this using a
> >>>> dictionary of the actual words used in artists, albums and songs.
> >>>>
> >>>> But I am worried that this is technology covered by the T9 patents,
> >>>> which are heavily protected by AOL/Tegic. (To see the eleven
> >>>> patents in detail, go to www.uspto.gov and search for Tegic)
> >>>>
> >>>> What do we think? Is it too risky to include this technique in open
> >>>> source software?
> >>>>
> >>>> Disclaimer: IANAL, but I have had dealings with software patents.
> >>>>
> >>>> Phil
> >>>>
> >>>>

[Patrick Hurley]

>> We can't use it if we believe it's patented.
>
>
> Indeed. That's what I thought.
>
> Pity.

Have we considered just asking them? Seems like there would be little
harm in checking to see if they think this violates their patents (of
course they do, even though it seems tenuous in my reading), if so would
they be willing to provide a license for it use here (in any area
outside their general business focus) for an open source project?

[Phil Barrett]

On 20 Nov 2003, at 19:47, Patrick Hurley wrote:
>>> We can't use it if we believe it's patented.
>> Indeed. That's what I thought.
>> Pity.
>
> Have we considered just asking them? Seems like there would be little
> harm in checking to see if they think this violates their patents (of
> course they do, even though it seems tenuous in my reading), if so
> would they be willing to provide a license for it use here (in any
> area outside their general business focus) for an open source project?

Nothing ventured, nothing gained, I suppose.

Who do you reckon is more likely to get a positive response? An
official approach from Slim Devices, or a lone developer contributing
to an open source project? If there's a license involved I guess it
would have to be to the company.

Phil

[Patrick Hurley]

Phil Barrett wrote:
> On 20 Nov 2003, at 19:47, Patrick Hurley wrote:
>
>>>> We can't use it if we believe it's patented.
>>
>> Have we considered just asking them? Seems like there would be little
>> harm in checking to see if they think this violates their patents (of
>> course they do, even though it seems tenuous in my reading), if so
>> would they be willing to provide a license for it use here (in any
>> area outside their general business focus) for an open source project?
>
>
> Nothing ventured, nothing gained, I suppose.
>
> Who do you reckon is more likely to get a positive response? An official
> approach from Slim Devices, or a lone developer contributing to an open
> source project? If there's a license involved I guess it would have to
> be to the company.
>

I guess the lone developer if he/she got any response might be more
likely to get the positive response. Odds are we will need to use
something archaic like USPS and stamps for them to even respond, but you
can apparently start with info (AT) tegic (DOT) com of course at some point the
license would have to be given to Slim Devices as they are the ones who
need to be indemnified (sp? usage?).

[Phil Barrett]

A follow-up to last week's discussion about possibly patented
techniques for text entry...

I contacted Tegic, and received a reply yesterday from Rick Romatowski,
Director of Product Marketing, Tegic/AOL:

> Thanks very much for writing. Regrettably, we're not in a position to
> provide T9 technology or related patents for open source licensing. If
> you believe that your prototype would "arguable fall under the scope
> of one or more" Tegic patents, I'd urge caution. By policy, AOL
> protects, enforces and defends its intellectual property rights.

So it looks like that idea is dead in the water. Ho hum.

Phil

[Robert Moser II]

Makes me wonder how cd3o does it. Of course their server is closed
source so maybe they licensed the patent.

Prime example of why software patents suck. We can't take a simple idea
(match /^[abc][def][abc][tuv]/i to get Beatles) and implement it
entirely from scratch because someone has staked a claim on it.

Phil Barrett wrote:

> A follow-up to last week's discussion about possibly patented techniques
> for text entry...
>
> I contacted Tegic, and received a reply yesterday from Rick Romatowski,
> Director of Product Marketing, Tegic/AOL:
>
>> Thanks very much for writing. Regrettably, we're not in a position to
>> provide T9 technology or related patents for open source licensing. If
>> you believe that your prototype would "arguable fall under the scope
>> of one or more" Tegic patents, I'd urge caution. By policy, AOL
>> protects, enforces and defends its intellectual property rights.
>
>
> So it looks like that idea is dead in the water. Ho hum.
>
> Phil