Remaining DecTalk Issues in Speakup?

Remaining DecTalk Issues in Speakup?

[blinux-list]

Hi All: Several days ago I asked something similar in the Speakup list, but I 
guess maybe this may be a great place to inquire? Last Thursday evening I had 
to restart my Debian machine after 130days. Kernel jumped from 5.10 to 5.15 03. 
First thing I noticed was a 4second delay in the DecTalk was back until I 
ajusted flush-time 
to 100. I think back last September I was able to set this at 10, much easier 
to scroll-and-listen to content. Please consider allowing a 10 setting. Next, I 
bet this involves caps-start-and-stop? If I type or read any capital letter, 
the next item will say a "p" on top of what ever that symbol or even a space. 
So as an example, a capital i followed by a space will say "i p space" This is 
`extremely challenging to read charactor-by-charactor. I looked-and- caps-start 
is set at 160-and-stop is 100. Sure pitch rises on capitals but its these extra 
p's which are in my way. Next, now once again insert2-and-3 for pitch are 
broken. And eventually I will create an audio file to send Samuel to 
demonstrate that just out of the blue, while reading/scrolling, volume, rate, 
and pitch drop suddenly. These have been occurring ever since 2003 when the 
DecTalk drivers joined Speakup. Is Speakup an only console screen-reader with 
DecTalk support? While Fenrir has many more options, I have not heard of it 
support DecTalk?
And lastly on another topic, did something change involving ssh, as now if we 
enable openssh server, any1 can login to my system without any authentication. 
Obviously until we find out how to fix this, its turned off, but maybe since 
last September Debian may have changed something. I would very much appreciate 
suggestions on all of these items. Thanks so much in advance.
Chime

Remaining DecTalk Issues in Speakup?

[blinux-list]

Howdy Chime ,

> While Fenrir has many more options, I have not heard of it support 
> DecTalk? 
fenrir supports dectalk using emacspeak speech back-end. this is 
currently very experimental lacking testers or users.
it requres
- emacspeak somewhere on hard disc
- python-pexpect
- tcl / tclsh
- set speech driver to "emacspeakDriver"
- set? speech serverPath in your settings.conf for emacspeak to your TTS 
server to use:
serverPath=/home/chrys/Projekte/emacspeak/servers/espeak
is what i used for test it with espeak.

again, its still very untested ( i don?t own most of the devices listed 
for emacspeak) and experimental

to maybe not break an existing setup. you can test it by passing the 
settings? wit -o parameter at startup once

sudo fenrir -o 
"speech#driver=emacspeakDriver;speech#serverPath=/path/to/your/emacspeak/server"

or create an copy of the settings.conf, change the required parameters 
and use
sudo fenrir -s /path/to/your/copy/of/fenrir/settings.conf
to use an given setting file.

if wit works, awesome :). if not, let me know. maybe we can fix it.

good luck.


cheers chrys

Am 01.02.22 um 16:28 schrieb Linux for blind general discussion:
> Hi All: Several days ago I asked something similar in the Speakup 
> list, but I guess maybe this may be a great place to inquire? Last 
> Thursday evening I had to restart my Debian machine after 130days. 
> Kernel jumped from 5.10 to 5.15 03. First thing I noticed was a 
> 4second delay in the DecTalk was back until I ajusted flush-time to 
> 100. I think back last September I was able to set this at 10, much 
> easier to scroll-and-listen to content. Please consider allowing a 10 
> setting. Next, I bet this involves caps-start-and-stop? If I type or 
> read any capital letter, the next item will say a "p" on top of what 
> ever that symbol or even a space. So as an example, a capital i 
> followed by a space will say "i p space" This is `extremely 
> challenging to read charactor-by-charactor. I looked-and- caps-start 
> is set at 160-and-stop is 100. Sure pitch rises on capitals but its 
> these extra p's which are in my way. Next, now once again 
> insert2-and-3 for pitch are broken. And eventually I will create an 
> audio file to send Samuel to demonstrate that just out of the blue, 
> while reading/scrolling, volume, rate, and pitch drop suddenly. These 
> have been occurring ever since 2003 when the DecTalk drivers joined 
> Speakup. Is Speakup an only console screen-reader with DecTalk 
> support? While Fenrir has many more options, I have not heard of it 
> support DecTalk?
> And lastly on another topic, did something change involving ssh, as 
> now if we enable openssh server, any1 can login to my system without 
> any authentication. Obviously until we find out how to fix this, its 
> turned off, but maybe since last September Debian may have changed 
> something. I would very much appreciate suggestions on all of these 
> items. Thanks so much in advance.
> Chime
>
> _______________________________________________
> Blinux-list mailing list
> Blinux-list at redhat.com
> https://listman.redhat.com/mailman/listinfo/blinux-list
>

Remaining DecTalk Issues?

[blinux-list]

Thanks Chris: Not only would I love to try this, but ideally it would be easier 
if I could connect with you or another Fenrir expert by phone or on Zoom. I 
know nothing about Emacs, but just running an EmacsSpeak server is what I 
wanted to try in Speakup, but either it didn't work or I didn't know what I was 
doing? Now, just as important, in Fenrir, could I run a Speakup style keyboard 
configuration? It would certainly make a possible transition easier. I mean, 
the fact of Fenrir having a pronunciation dictionary as well as an easier way 
of ajusting volume are rather apealing. Thanks so much in advance
Chime

SSH server authentication (was Re: Remaining DecTalk Issues in Speakup?)

[blinux-list]

On Tue, 1 Feb 2022, Linux for blind general discussion wrote:

> And lastly on another topic, did something change involving ssh, as now if we 
> enable openssh server, any1 can login to my system without any 
> authentication. Obviously until we find out how to fix this, its turned off, 
> but maybe since last September Debian may have changed something.

Hi,

YOu didn't say which version of openssh-server or even Debian that you are 
running, but unless you're running unstable, I doubt that anything was 
changed.  Certainly nothing has changed in Bullseye since September.

But a few things to be aware of:

1.  The configuration for the SSH server is /etc/ssh/sshd_config

2.  The Debian default used at least to allow root login. I always turn 
this off because I think it's asking for trouble.

If you have no root password (also a bad idea) and root login is allowed, 
anyone could log in as root without a password.

3.  You may want to consider requiring remote access using SSH keys.  This 
means that you can turn off password authentication altogether and only 
people with valid keys can log in.

4.  You might also want to consider using a non-default port.  Yes some 
people will eventually find you, but in my experience, you get much less 
noise by doing this.

HTH,
Geoff.

SSH server authentication (was Re: Remaining DecTalk Issues in Speakup?)

[blinux-list]

Hi Jeff: Yes its Unstable. Usually any1 here needs a dot pem file to get in. 
For some reason, even though openssh is installed, it doesn't tab-complete, so 
I cannot give you a version number. Not only that but man -k doesn't find 
anything. Also, even when the server was on, no1 could login with root, but 
thanks for your suggestions.
Chime

Remaining DecTalk Issues?

[blinux-list]

Howdy,

Sure, we can voice about that.
Where are you from? What timezone you live there (just for time scheduling)

About the keyboard layout. Well this is completely configurable. I don?t use an screenreader, so i do not really know speakups shortcuts. The current layout is made by storm_dragon. So yes it?s doable, but not the case by default. If someone creates a layout similar to speakup i will of corse push it upstream master.

The dectalk stuff looks also very simple. If the emackspeak server won?t fit, i can hammer an python backend.

Cheers chrys

Von meinem iPhone gesendet

> Am 01.02.2022 um 17:12 schrieb Linux for blind general discussion <blinux-list at redhat.com>:
> 
> ?Thanks Chris: Not only would I love to try this, but ideally it would be easier if I could connect with you or another Fenrir expert by phone or on Zoom. I know nothing about Emacs, but just running an EmacsSpeak server is what I wanted to try in Speakup, but either it didn't work or I didn't know what I was doing? Now, just as important, in Fenrir, could I run a Speakup style keyboard configuration? It would certainly make a possible transition easier. I mean, the fact of Fenrir having a pronunciation dictionary as well as an easier way of ajusting volume are rather apealing. Thanks so much in advance
> Chime
> 
> _______________________________________________
> Blinux-list mailing list
> Blinux-list at redhat.com
> https://listman.redhat.com/mailman/listinfo/blinux-list
> 

Remaining DecTalk Issues?

[blinux-list]

Hi Chris: I am in Southern California, Pacific Standard time, where right now 
it is 937AM. I trust Fenrir also supports the new embedded voices in Voxin? I 
think Gilles-and-others have suggested Fenrir. Thanks in advance
Chime

SSH server authentication (was Re: Remaining DecTalk Issues in Speakup?)

[blinux-list]

On Tue, 1 Feb 2022, Linux for blind general discussion wrote:

> Hi Jeff: Yes its Unstable. Usually any1 here needs a dot pem file to get in.

hmm OK well you may want to look at 
/usr/share/doc/openssh-server/changelog.Debian.gz for any package changes. 
I'm not running Syd anywhere so can't be of much help.

> For some reason, even though openssh is installed, it doesn't tab-complete, 
> so I cannot give you a version number.

Firstly, openssh-server is the package name, the binary is called sshd. 
And it's in /usr/sbin which won't be in a regular user's path.

If you want to see the package version for any package, you can run:

dpkg -s openssh-server

HTH,
Geoff.

Remaining DecTalk Issues?

[blinux-list]

Howdy Chime,

good to know.  I will PM you about a call. I do not find time.

Yes, fenrir supports speech-dispatcher. Voxin is available for that.

Cheers chrys

> Am 01.02.2022 um 18:40 schrieb Linux for blind general discussion <blinux-list at redhat.com>:
> 
> ?Hi Chris: I am in Southern California, Pacific Standard time, where right now it is 937AM. I trust Fenrir also supports the new embedded voices in Voxin? I think Gilles-and-others have suggested Fenrir. Thanks in advance
> Chime
> 
> _______________________________________________
> Blinux-list mailing list
> Blinux-list at redhat.com
> https://listman.redhat.com/mailman/listinfo/blinux-list
>