Jouke

OSCON day 3

Besides attending some talks, I spent most of this day preparing and giving my own talk.

I got a very flattering introduction by Damian Conway, who explained that after his Presentation Aikido talk, several people had come up and asked him to introduce them, just like I did. He told the audience he had turned them all down, except me. If that’s not flattering…

Anyway, although there were only a few people in the room, the talk went very well I think, and the people that were there, were really interested in the subject. It showed, because they kept asking questions during the talk, especially during the demo. I had planned the demo to last for not more than 15 minutes, but because all kinds of questions kept being asked (and I was too enthusiastic about the subject not to answer them), the demo took more than 30 minutes. In a 45 minute talk, that means you’re running hopelessly out of time. I had not kept track of time, and after the demo I had only 3 minutes left, and slides for 15 minutes. I rushed through the last part, which wasn’t a good plan on hindsight.

The good thing is that most people kept asking questions and giving suggestions after I left the room. I actually even got the offer of someone who wanted to help out improving the online help for pVoice. This is a good thing since it will greatly improve the useability of the software.

After I finished talking with the audience, I ran downstairs to the ‘Works in Progress’ session to do the last part of the presentation again. I had only 10 minutes in this session, and used it as a ‘Cry for Help’. Let me use this journal entry as another Cry for Help. I need people to help with development. I don’t expect anyone to spend as much time on pVoice as I do, but I hope people will help porting pVoice to other platforms, especially Linux and Mac OS X. It really shouldn’t be too hard. It’s all written in Perl and uses wxPerl as the GUI library. wxPerl compiles on many platforms (including OS X), but I know I’ve used several Win32 specific calls, and there really should be workarounds for those calls on other platforms. Eventually I should be able to port pVoice to linux myself, but I don’t own a Mac, so I can’t do that myself.
The Speech synthesis is done in only one subroutine, so it shouldn’t be too hard to make that work on other platforms too. I’m still planning to write a generic TTS (Text To Speech) module that interfaces with the different TTS modules for different platforms. If I find the time. If someone else does it, I’ll use that and it will be a huge step towards making pVoice multi-platform.

The slides of my talk (for those who didn’t have the chance to be there) are on http://jouke.pvoice.org/oscon-2004/index.html, so take a look at that to find out what pVoice is all about these days.

OSCON day 2

The second day of tutorials for me consisted mainly of hallway converstations. For a while I was at Rocco’s POE tutorial, but I couldn’t really concentrate, so I decided to be better off leaving and thinking about my own presentation the next day.

I had short conversations with people like Tim O’Reilly, Damian, Ovid, “cLive;-)”, kudra, (Josh)ua, petdance and many more. This is the great part of conferences: the social networking thing.

The evening was fun with Larry’s State of the Onion, Paul Graham’s talk on Hacking and Damian’s ‘Life, the Universe and Everything’. I had seen Damian’s talk in Germany two years ago, but he has made quite a few changes in it, which make it sound as if it were almost a new talk. It was hilarious (or how you are supposed to spell that anyway).

I tried to get my talk for the third day rescheduled, only to find out that ‘The Daily Source’ had already gone into print, so there couldn’t be any more changes. Instead I found Damian willing to introduce me for my talk, and I’ve arranged a short talk on pVoice during the ‘Works in Progress’ session, because I heard many people say “if your talk wouldn’t be scheduled with the Lightning talks, I’d surely be there”. Now there’s a chance for them to go and listen to what I have to say

It was about 2am when I went to bed, and 8 when I got up, so I’m quite tired. Still, the talk has been finished on time, and I have enough adrenaline in my vains to stay awake…let’s see how it goes…

OSCON day 1

We’re approaching the end of the first day of tutorials. My first day consisted of listening to Damian and to Damian.

The first talk was ‘Best Practice Perl’, which was -as usual with Damians talks- a good one. I didn’t agree with everything he mentioned, and as he said, that wasn’t nessecary, but he certainly brought up a good number of useful points I hadn’t really thought of.

Now I’m at the ‘Presentation Aikido’ talk. Really, really having a good time here. I definately need to work on my slides again tonight.

Of course I’ve seen a number of people I saw again for the first time, and others I hadn’t seen for far too long. These include (but are not limited to…I’m good at forgetting things) cwest, kudra, schwern, dha, merlyn, sky, petdance and some others…

I have no clue what I’m going to do tonight, other than review my slides, because I got many good ideas from Damian’s last talk. Maybe I’ll just keep hanging around in the lobby…

OSCON day 0

It’s nice to be in Portland before everyone else does. Actually checking in in the hotel on Saturday and spending most of Saturday and Sunday in the hotel lobby allows you to see all the regular visitors of the hotel go, and the geeks coming in.

While I was practising my worst habit (smoking), Saturday evening, suddenly Nick Clark stood before me. He was the first familiar face to show up here. Sunday, Jesse Vincent and a few others showed up. We quickly formed the first group of people wearing geekshirts, working op laptops and discussing Perl there.

Around 4 Ingy, Autrijus and Whiteg arrived and some 10 of us went to a cheap but tasty Mexican restaurant. Now, the forementioned three and I are sitting in our hotelroom which has been transformed to some kind of dorm. Wireless access here, so we’re happy

I’m all set for day 1!

pVoice 2.2 beta 2

I keep calling it beta releases, although they’re actually short updated versions….but anyway, there’s a new version out.

Just two new features: you can now load different datafiles and the one-switch operation is now fully supported.

I thought it was wise to release this, since I spoke about it last week and people were really interested. Now they can download what they’ve seen. Another reason is that I can now really concentrate on the hard stuff that will go into version 2.2: plugin support. I’ve been thinking and talking about this for a few months now, and I have the idea that I’ve finally completely thought it out….although I’ve said that before…

If it will take a long time to make it work, then at least there is a version released with all features before that…

arrived at the Marriott

After spending a week at the La Quinta Convention Center, I’ve now arrived at the Marriott Downtown hotel. The room kind of scared me. We’re actually supposed to sleep there with 4 people, but the room (and the individual beds) is smaller than the one I had at the La Quinta.

Wireless coverage is reasonable. I don’t seem to be able to connect to anything at my hotelroom, but the lobby has numerous different networks I can connect to. However, I don’t seem to be able to connect to an SMTP server (which is not that big of a problem since I can ssh to my server and send mail from there), nor can I connect to an irc server.
It’s at the very least much better than the La Quinta, which lacked all sorts of internet access. I don’t have to go to the Pioneer Square to check my mail…

It’s again extremely hot today. About 100F. Yesterday I didn’t really have a problem with the heat (it was 102 yesterday), but today it seems like the first day has worn me out. I’m staying in today to modify some slides an work on some code…

more later…

Another presentation…

Yesterday I did the presentation at the Center of Spoken Language Understanding. I think it went pretty well. Especially so, because after the talk ms Lois Black (neuropsychologist I think) and Jan van Santen (a professor who told me that Kevin Lenzo actually graduated under him) and I had a rather long conversation about Krista and pVoice.

If I wasn’t already impressed with the friendlyness of Portland people, these two were absolutely the best so far. They were really interested in pVoice, and we spoke about the possibility to add certain features (like using sounds for input), about licensing issues (or actually the lack thereof) and the chances of getting more grants to further develop pVoice. It might implicate that certain added features won’t be Open Source. This however doesn’t have to be a problem, because I think Open Source is not a goal by itself. It’s a means of getting other people involved to improve the software and to let people add their own features. I’m very happy to have chosen the Artistic License as the license under which pVoice is distributed….

Of course the time was too short to really get into details, but they’re going to see if there is a way to get more funding from certain organizations. This of course is very exciting!!

keeping my fingers crossed

Talk at the OHSU

Yesterday was fun.

In the morning I finished up my slides and added some new features to pVoice. It’s now possible to use pVoice with only one switch. Of course, this had already mainly been covered in AAC::Pvoice, but pVoice itself didn’t support it yet. I had to make some minor changes to AAC::Pvoice to be able to switch between one switch, two switch and touchscreen operation in one pVoice session. It turned out that when you change the setting from one switch to something else, the timer that invoked the automatic scanning kept running and couldn’t be stopped, because AAC::Pvoice::Input thought that the current setting wasn’t one-switch. It resulted in a segfault no less…and no errormessage. Well, it’s been fixed now.

The presentation went well. It was located at the Child Development Rehabilitation Center (CDRC). There were only about 10 attendees, which was mildly disappointing. But I always think that one interested attendee is better than 100 sleeping ones. In this case, all 10 were very interested.
I think the simplicity of pVoice, and the support for multiple input devices along with the multi-linguality of it, and the fact that pVoice is free, impressed the audience. Hopefully they’re going to try out pVoice in their professional situation.

One of the attendees turned out to be someone I vaguely knew. It’s a Dutch scientist working on Speech technology, called Esther Klabbers. We once exchanged some emails about a Dutch implementation of Festival.
Anyway, she invited me to have a drink afterwards. We had a lot of fun, a little too many beers, and after we picked up her boyfriend, we had some yummy Japanese food.

She also invited me to come over to her faculty to do the presentation again tomorrow (Friday) for her colleagues. I’m really looking forward to it.

This afternoon I’ll be trying to act like a real tourist and have a real good look around in the Portland area. I’ll take my camera and try to shoot a bit more pictures than I did so far.

the latest changes to pVoice and AAC::Pvoice have been committed to cvs, so if you want to take a look, check it out. More info on the cvs repository can be found at http://opensource.pvoice.org.

Portland day 3

Yeah, I missed a day in my journal…I know…

Not a lot to tell so far. Yesterday I went for dinner and drinks with Ovid, ua and Todd (sorry, I didn’t get your last name). We had fun, good Lebanese food, cocktails and local beer…mmmmm…

Today I have to finish the presentation I’m doing tomorrow at the Child Development and Rehabilitation Center at the OHSU here in Portland. It’s about pVoice of course . I announced that it would also be about other Open Source Assistive software, but actually, there’s not a lot to tell. Apart from “The Hawking Communicator” and “Skipper” (which intends to be a linux distro for people with all kinds of physical disabilities), I’ve hardly found any active open source projects in this area. Of course, there are a number of initiatives like emacspeak and linux-things for blind people (braille interfaces and stuff), but the focus in my talk will be on people with Celebral Palsy, and I can’t say I’ve found a lot of projects useable for these people.

Anyway, I’ve got enough to tell about pVoice, so I’ve got enough preparation to do…

Portland day 1

Well, it seems like I’ve arrived in Portland, OR last night. I don’t think that I’ve ever had such a bad journey as I had yesterday. Apart from being a long journey, a lot of things didn’t exactly go the way I planned.

Ooops..that’s my battery saying it’s almost empty, so I’ll have to finish this later. I’m now in the Starbucks coffeehouse at Pioneer Square…wonderful initiative…Personal Telco in Portland++