[MV] Where are we now?

[Chuck Rogers]

Todd (and everyone else):

Those are kind of hard questions to answer without first providing some 
background information that will help clarify things.

The relationship between MacSpeech and Philips is alive and well. 
Technically, the engine used in iListen is the "MacSpeech" engine, but 
it is based on code ported over from the Philips FreeSpeech 2000 
engine. We used this engine as a spring board and have continually made 
it better. At this point, it is pretty far from what Philips originally 
gave us - so to call it a derivative of the FreeSpeech engine would be 
loosely accurate, but also would do the MacSpeech engine an injustice 
as it has evolved so far beyond what FreeSpeech 2000 originally offered 
(at least what it offered in what could be ported).

Now along the way we have encountered several stumbling blocks: Mac OS 
9.1 did not work at all with iListen, and the transition to Mac OS X 
was especially painful - mostly because the support we needed in the OS 
just wasn't there until Mac OS X 10.1.5.

Allow me to digress for just a moment: Those using ViaVoice during 
those times may wonder why ViaVoice beat us to the starting gate with 
both a Mac OS 9 product and a Mac OS X product (even though we 
announced both first). Well, the obvious reason is because they are IBM 
- they had a lot more $ and man hours to throw at the problem. But the 
more subtle answer is because they chose not to use the tools Apple 
provides in the OS.

Something to keep in mind is that at some level EVERY program interacts 
with the OS. The programmers just get to decide how much interaction 
there is. There is a good thing and a bad thing about that. The good 
thing is that developers who bypass the tools in the OS can pretty much 
do what they want. So IBM wrote its own drivers, etc. and were able to 
ship more quickly. The bad thing about that is that when Apple changes 
the OS, it makes it much harder for the programmers since they 
sometimes have to rewrite everything in order to get things to work 
right. This also contributes to bloating, since the application now has 
to have different code for different versions of the OS, etc.

MacSpeech chose to stick with the tools Apple provided, and we 
complained loud and often when the tools were not sufficient for our 
needs. Luckily Apple responded. The result is that iListen completely 
works with Mac OS X 10.3, (including 10.3.6, btw), and it will work 
with 10.4. Take a look at how long it has been since IBM or Scansoft 
issued an update for ViaVoice. Sometime in 2002, I think (and they 
stopped development on speech software for consumers in April of 2003).

OK - so why the history lesson?

I would like to underscore that it is a major task to bring a new 
speech engine to a platform other than that on which it was written. 
Philips continues to serve the high-end consumer market, and we will 
continue to look at options for bringing that engine over to the Mac 
platform. At such a time we feel the improvements in the engine 
outweigh the investment of bringing it over to the Mac, we will pursue 
it with more vigor.

Now, to thwart the obvious assumption some could make that I am saying 
the current version of the Philips engine is not that much better than 
the FreeSpeech 2000 engine, that is not at all what I am saying. What I 
am saying is that while the current version of the Philips speech 
engine is a lot farther ahead of the original FreeSpeech 2000 engine we 
got the rights to in 1999, we have made significant improvements in the 
MacSpeech engine that make the effort of porting the newer version not 
worth the effort at this time.

But the time will come soon when we will need to replace the engine 
that works underneath iListen. Are we working on it? Let's just say we 
are starting to carefully weigh our options. As to when a version of 
iListen will be available with a new speech engine, that is anyone's 
guess - but I would say we are at least a year or more away. We still 
have quite a few things we would like to introduce with the current 
engine first.

One last comment, while we are on the subject. When the time comes that 
we do introduce a product with a new speech engine (which may or may 
not be called "iListen") it is very likely users will have to trade off 
between features of the current version and the benefits of the newer 
engine. While the new engine may provide decreased training times and 
increased accuracy for some people, it is likely to be subject to many 
of the same growing pains version 1 of iListen had to go through. It 
may take us some time to get all the features present in the current 
product into the new one. Hopefully the second time around we will not 
be hampered by such dramatic changes to the Mac OS, however.


Best Regards,

Chuck Rogers, Chief Evangelist
MacSpeech, Inc.