← Back to team overview

speechcontrolteam team mailing list archive

Developement Team Updates and Remarks



        NOTE: This is the first (or maybe second) of a series of
        near-regularly published emails from the head of development for
        SpeechControl. This is to not only ensure everyone knows what's
        happening development-wise, but it's a means of decision making
        with the rest of the team, as prescribed by Manuela (hajour).
        NOTE: Since ALL of SpeechControl is getting this e-mail, I've
        taken the liberty in briefly explaining any technical terms that
        may arise.


        Bazaar branch: lp:libopenmary-c++
                ISSUE: Currently, development for this library is at a
                standstill. We're encountering a serious run-time error
                with a segmentation fault. Segmentation faults occur
                usually when bad pointer arithmetics has been done and a
                request to an invalid, null or other inaccessible
                location in memory. The method has been removed from the
                code, but you can get the code in the 'segfault.zip'
                archive attached.
                ** This is filed as our first bug **
                COMING-FEATURE: We need to incorporate the ability to
                either communicate to PulseAudio via D-Bus, or link to
                PulseAudio after we resolve the issue stated above. This
                is because openMary doesn't automatically plays
                synthesized text; it returns it as an audio stream.
                Please post suggestions for fixing this issue on the bug
        Bazaar branch (inactive): lp:speechcontrol-daemon
                ABOUT: This project's to be the crown jewel of
                SpeechControl. It'll interact with Speech Dispatcher,
                the Speech API engine for Linux, bring in libopenmary-c
                ++ , and when production starts, it'll bring in
                libjulius, a modified version of Julius we'll either
                borrow from Simon or make on our own. This won't be an
                application or library, but rather a daemon, an
                inhibited application that runs like a server of sorts.
                strikeu's working intently at getting the daemon up and
                running, but his overall progress cannot continue until
                libopenmary-c++ is at least on its knees.


        We are making progress! That, I can ensure you. But our team of
        roughly 15 developers has only 4 active individuals, myself
        excluded. That means 5 people have put together what you see
        here. On the behalf of development; we implore you to come to
        #ubuntu-speechcontrol-devs (speak to Kiempe [Undifined] before
        entering; the room's locked @ #ubuntu-speechcontrol) and get
        with the action! Rome wasn't built in a day, but many lights
        make light work! 
                                                              Thank you.

Sent from Ubuntu

Attachment: segfault-20111001-rev20.tar.gz
Description: application/compressed-tar

Attachment: signature.asc
Description: This is a digitally signed message part