Page 5 of 5
Re: Became a Linux build discussion - originally "How to encourage more developers to be involved"
Posted: Wed Sep 17, 2025 6:56 am
by MrFlibble
mcarans wrote: ↑Wed Sep 17, 2025 1:49 am
MrFlibble wrote: ↑Fri Sep 12, 2025 1:29 am
mcarans wrote: ↑Thu Sep 11, 2025 10:16 pm
This is great news! Now that I have Oolite looking like it might work on modern MSYS2 in Windows, I want to look at the espeak or espeak-ng library. What have you settled on for Linux? If you're able to use the system espeak-ng, I'd like to try to do the same in Windows rather than building the old hacked espeak.
espeak-ng FTW!
Have you seen any issues with speech with blocking using espeak-ng on Linux? For example, try to change the view when a message is spoken and see if it hangs until the message is finished.
Just tested a launch, with speech from an OXP that talks a lot when you leave station. Flipped view back and forward several times without issue during speech.
Re: Became a Linux build discussion - originally "How to encourage more developers to be involved"
Posted: Wed Sep 17, 2025 8:10 am
by another_commander
MrFlibble wrote: ↑Wed Sep 17, 2025 6:56 am
Just tested a launch, with speech from an OXP that talks a lot when you leave station. Flipped view back and forward several times without issue during speech.
Can you please try the sequence described
in this post and let us know if you get the same behaviour on Linux?
Re: Became a Linux build discussion - originally "How to encourage more developers to be involved"
Posted: Wed Sep 17, 2025 6:35 pm
by MrFlibble
another_commander wrote: ↑Wed Sep 17, 2025 8:10 am
MrFlibble wrote: ↑Wed Sep 17, 2025 6:56 am
Just tested a launch, with speech from an OXP that talks a lot when you leave station. Flipped view back and forward several times without issue during speech.
Can you please try the sequence described
in this post and let us know if you get the same behaviour on Linux?
Would love to... I don't get the "Identification system locked onto navigation buoy" speech. I wonder if the oxp that blathers at me in the station bailiwick is causing that to be overruled?... Disabling oolite.oxp.Thargoid.TrafficControl.oxz...
Ah!... So, spoken messages 'on', rather than comms only. I now remember why I'd had the latter set
That's grim... OFF = oh eff eff, as bad as the "aitch oh oopy" bars. Would it perhaps be prudent to lowercase text fed to espeak-ng until such time as a parallel set of variables are conjured for screen and voice (international options too of course!).
However, running that test, I get the "Identification system locked onto navigation buoy" speech cut off mid sentence, and "aft view" immediately spoken and actioned, so I think my fork has inadvertently cured a bug which didn't affect me
I wonder if there's a way to get overlapping messages 'mixed' (optionally)... When I switch to rear view, is the station buoy any less targeted?
Re: Became a Linux build discussion - originally "How to encourage more developers to be involved"
Posted: Fri Oct 03, 2025 7:04 pm
by mcarans
MrFlibble wrote: ↑Wed Sep 17, 2025 6:35 pm
That's grim... OFF = oh eff eff, as bad as the "aitch oh oopy" bars. Would it perhaps be prudent to lowercase text fed to espeak-ng until such time as a parallel set of variables are conjured for screen and voice (international options too of course!).
However, running that test, I get the "Identification system locked onto navigation buoy" speech cut off mid sentence, and "aft view" immediately spoken and actioned, so I think my fork has inadvertently cured a bug which didn't affect me
Turns out the issue was a bug specific to Windows xaudio2 that I've fixed in forks of espeak-ng and pcaudiolib.
Agree we'll need to tailor speech options for espeak-ng.
How goes your Linux update fork?
Re: Became a Linux build discussion - originally "How to encourage more developers to be involved"
Posted: Fri Oct 03, 2025 9:14 pm
by MrFlibble
mcarans wrote: ↑Fri Oct 03, 2025 7:04 pm
MrFlibble wrote: ↑Wed Sep 17, 2025 6:35 pm
That's grim... OFF = oh eff eff, as bad as the "aitch oh oopy" bars. Would it perhaps be prudent to lowercase text fed to espeak-ng until such time as a parallel set of variables are conjured for screen and voice (international options too of course!).
However, running that test, I get the "Identification system locked onto navigation buoy" speech cut off mid sentence, and "aft view" immediately spoken and actioned, so I think my fork has inadvertently cured a bug which didn't affect me
Turns out the issue was a bug specific to Windows xaudio2 that I've fixed in forks of espeak-ng and pcaudiolib.
Agree we'll need to tailor speech options for espeak-ng.
How goes your Linux update fork?
Have stalled for a while due to other matters. Will be able to pick up again from about the middle of next week.