On Sat, Apr 23, 2022 at 04:15:38PM +0200, Adrian Orjales wrote: > Hello all hello. > I am trying to build my own OS based on Arch Linux, but I am having some > problems with the screen reader. > I have an arch linux fresh installation, with systemd, espeakup and > espeak-ng installed with pacman. > After installing the espeakup package, I noticed that the system was very > slow on boot, and some times espeakup did not wake up after boot. > After some research, I've discovered that the cause of the slow boot was > that the systemd espeakup service unit uses systemd-udev-settle.service and > waits for it to end to start, to ensure all hardware is discovered. > But as I've read, this operation is very discouraged because the > udev-settle discover process is asynchronous and it slows down the boot > process. to be honest, I'm not sure that the slowdown in booting is due to this. > Furthermore, when this service ends, espeakup has the same problem, > sometimes it talks, sometimes it is not able to talk. > The problem of no speech is solved if I run systemctl reload espeakup && > systemctl start espeakup manually from command line. did you set the default sound card? > Now, I have tried removing the wants and requires of > systemd-udev-settle.service and left only the sound.target. The slow boot > is solved, but espeakup does not start properly, and I don't know how can I > solve this. I'm afraid it's not possible to get rid of udev-setle.target without changes in espeakup. > One of the info that I've found says that sometimes, systemd sound.target > unlocks the awaiting services when the first sound card is detected, but I > have a Nvidia gpu that is detected as a soundcard, but alsa's default card > is my onboard intel hda. > I know that this list is about speakup, not espeakup, but I'm sure that you > like me know that we are not much people talking about these topics and is > difficult to find help. > Thanks in advance -- Sincerely, Alexander
