[TUHS] TERM for v8

[TUHS] TERM for v8

[Will Senn]

What should I set TERM to on V8 to get the best results on my Mac 
Terminal. If I set it to vt52, vt100, or vt132, only 8 lines appear at 
the bottom of the terminal window (of about 24 lines):

---

root::0:4:m0130,m322:/:
daemon:x:1:1:m0000,m000:/:
sys:sorry:2:1:m0130,m322:/usr/sys:no-login
bin:sorry:3:4:m0130,m322:/bin:
ken:sorry:6:1:m0130,m322:/usr/ken:
dmr:sorry:7:4:mh1092,m069:/usr/dmr:
nuucp::238:1:mh2019,m285,uucp:/usr/spool/uucppublic:/usr/lib/uucp/uucico
uucp::48:1:mh2019,m285,nowitz:/usr/lib/uucp:
"passwd" 20 lines, 770 characters
----

The 8 line window works about like I'd expect - the arrow keys move up 
and down until the screen needs to scroll, then B's and A's show up. I'm 
used to that on BSD. Using the j and k keys work better and when I 
scroll down enough lines, the lines move up to fill the whole terminal 
window and the file can be edited in the full window. Is there a better 
TERM setting that will get 24 lines to show up on file open?

Thanks,

Will

-- 
GPG Fingerprint: 68F4 B3BD 1730 555A 4462  7D45 3EAA 5B6D A982 BAAF

[TUHS] TERM for v8

[William Corcoran]

I had a similar issue with BSD 2.11.
So, I hope this helps.

You can use vt100.   However, make sure your baud rate is 9600.

stty 9600

(SIMH makes the virtual adjustment so you don't need to worry about SIMH settings.)

I am using BSD 2.11 on a Mac with SIMH.  I noticed the console port would have a short window.   But, logging into dz ports were fine.   Setting the console baud rate to match the dz11 settings fixed the issue.

So, I suspect either a setting in vi that changes the window size on slow devices or it's hard coded somewhere.  I will take a look.

Truly,

Bill Corcoran


On Nov 5, 2017, at 12:03 AM, Will Senn <will.senn at gmail.com<mailto:will.senn at gmail.com>> wrote:

What should I set TERM to on V8 to get the best results on my Mac Terminal. If I set it to vt52, vt100, or vt132, only 8 lines appear at the bottom of the terminal window (of about 24 lines):

---

root::0:4:m0130,m322:/:
daemon:x:1:1:m0000,m000:/:
sys:sorry:2:1:m0130,m322:/usr/sys:no-login
bin:sorry:3:4:m0130,m322:/bin:
ken:sorry:6:1:m0130,m322:/usr/ken:
dmr:sorry:7:4:mh1092,m069:/usr/dmr:
nuucp::238:1:mh2019,m285,uucp:/usr/spool/uucppublic:/usr/lib/uucp/uucico
uucp::48:1:mh2019,m285,nowitz:/usr/lib/uucp:
"passwd" 20 lines, 770 characters
----

The 8 line window works about like I'd expect - the arrow keys move up and down until the screen needs to scroll, then B's and A's show up. I'm used to that on BSD. Using the j and k keys work better and when I scroll down enough lines, the lines move up to fill the whole terminal window and the file can be edited in the full window. Is there a better TERM setting that will get 24 lines to show up on file open?

Thanks,

Will

--
GPG Fingerprint: 68F4 B3BD 1730 555A 4462  7D45 3EAA 5B6D A982 BAAF

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20171105/bcc96099/attachment.html>

[TUHS] TERM for v8

[Will Senn]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1409 bytes --]

Bill,

Thanks! You're right about logging into the dz port, worked beautifully. 
On the console port, stty 9600 doesn't seem to take:

# stty
speed: 0 baud
erase = ^H; kill = @; intr = ^?; quit = ^\
start = ^Q; stop = ^S; eof = ^D; brk <undef>
old even odd -raw -nl echo -lcase -tabs -cbreak -tandem nl0 cr0 ff0 bs0
# stty 9600
# stty
speed: 0 baud
erase = ^H; kill = @; intr = ^?; quit = ^\
start = ^Q; stop = ^S; eof = ^D; brk <undef>
old even odd -raw -nl echo -lcase -tabs -cbreak -tandem nl0 cr0 ff0 bs0
#

Strange. Works fine on v7.

Thanks,
Will


On 11/4/17 11:34 PM, William Corcoran wrote:
>
> I had a similar issue with BSD 2.11.
> So, I hope this helps.
>
> You can use vt100.   However, make sure your baud rate is 9600.
>
> stty 9600
>
> (SIMH makes the virtual adjustment so you don’t need to worry about 
> SIMH settings.)
>
> I am using BSD 2.11 on a Mac with SIMH.  I noticed the console port 
> would have a short window.   But, logging into dz ports were fine.   
> Setting the console baud rate to match the dz11 settings fixed the issue.
>
> So, I suspect either a setting in vi that changes the window size on 
> slow devices or it’s hard coded somewhere.  I will take a look.
>
> Truly,
>
> Bill Corcoran
>


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20171105/40731067/attachment.html>

[TUHS] TERM for v8

[Clem cole]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 2735 bytes --]

This is a feature of vi.   If the baud rate is perceived to be at or less than a specific value (1200 IIRC) the screen is set to 8 lines at a time to keep repainting low. 

This was useful on dialup lines.  

Since you have a virtual serial port the baud rate has no bearing other than upper level applications trying to make inferences about the environment such as this. 

If you set the baud rate to 19200 or 38400 (EXTA and EXTB in the virgin V7 code base) those were the fastest serial speeds in those days. So any SW should infer the ‘best’ behavior.


Sent from my PDP-7 Running UNIX V0 expect things to be almost but not quite. 

> On Nov 4, 2017, at 9:34 PM, William Corcoran <wlc at jctaylor.com> wrote:
> 
> 
> I had a similar issue with BSD 2.11. 
> So, I hope this helps. 
> 
> You can use vt100.   However, make sure your baud rate is 9600.
> 
> stty 9600
> 
> (SIMH makes the virtual adjustment so you don’t need to worry about SIMH settings.)
> 
> I am using BSD 2.11 on a Mac with SIMH.  I noticed the console port would have a short window.   But, logging into dz ports were fine.   Setting the console baud rate to match the dz11 settings fixed the issue.
> 
> So, I suspect either a setting in vi that changes the window size on slow devices or it’s hard coded somewhere.  I will take a look. 
> 
> Truly,
> 
> Bill Corcoran
> 
> 
> On Nov 5, 2017, at 12:03 AM, Will Senn <will.senn at gmail.com> wrote:
> 
>> What should I set TERM to on V8 to get the best results on my Mac Terminal. If I set it to vt52, vt100, or vt132, only 8 lines appear at the bottom of the terminal window (of about 24 lines):
>> 
>> ---
>> 
>> root::0:4:m0130,m322:/:
>> daemon:x:1:1:m0000,m000:/:
>> sys:sorry:2:1:m0130,m322:/usr/sys:no-login
>> bin:sorry:3:4:m0130,m322:/bin:
>> ken:sorry:6:1:m0130,m322:/usr/ken:
>> dmr:sorry:7:4:mh1092,m069:/usr/dmr:
>> nuucp::238:1:mh2019,m285,uucp:/usr/spool/uucppublic:/usr/lib/uucp/uucico
>> uucp::48:1:mh2019,m285,nowitz:/usr/lib/uucp:
>> "passwd" 20 lines, 770 characters
>> ----
>> 
>> The 8 line window works about like I'd expect - the arrow keys move up and down until the screen needs to scroll, then B's and A's show up. I'm used to that on BSD. Using the j and k keys work better and when I scroll down enough lines, the lines move up to fill the whole terminal window and the file can be edited in the full window. Is there a better TERM setting that will get 24 lines to show up on file open?
>> 
>> Thanks,
>> 
>> Will
>> 
>> -- 
>> GPG Fingerprint: 68F4 B3BD 1730 555A 4462  7D45 3EAA 5B6D A982 BAAF
>> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20171104/e0ccf855/attachment-0001.html>

[TUHS] TERM for v8

[Will Senn]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 3883 bytes --]

That makes sense. Any idea where to start looking if the console thinks 
the baudrate is 0 even after setting it (supposedly)? I looked at the 
source code for stty.c (briefly) and it looks like it oughta work, if 
ioctl works... other things work ok nl, cr, etc., just speed doesn't 
have an effect... When I get home later, I'll prolly try putting some 
print statments in stty.c to see what it thinks it's doing. Here's what 
I see on the command line in the console:

# stty
speed: 0 baud
erase = ^H; kill = @; intr = ^?; quit = ^\
start = ^Q; stop = ^S; eof = ^D; brk <undef>
old even odd -raw -nl echo -lcase -tabs -cbreak -tandem nl0 cr0 ff0 bs0
# stty 9600
# stty
speed: 0 baud
erase = ^H; kill = @; intr = ^?; quit = ^\
start = ^Q; stop = ^S; eof = ^D; brk <undef>
old even odd -raw -nl echo -lcase -tabs -cbreak -tandem nl0 cr0 ff0 bs0
#

Will

On 11/5/17 12:44 AM, Clem cole wrote:
> This is a feature of vi.   If the baud rate is perceived to be at or 
> less than a specific value (1200 IIRC) the screen is set to 8 lines at 
> a time to keep repainting low.
>
> This was useful on dialup lines.
>
> Since you have a virtual serial port the baud rate has no bearing 
> other than upper level applications trying to make inferences about 
> the environment such as this.
>
> If you set the baud rate to 19200 or 38400 (EXTA and EXTB in the 
> virgin V7 code base) those were the fastest serial speeds in those 
> days. So any SW should infer the ‘best’ behavior.
>
>
> Sent from my PDP-7 Running UNIX V0 expect things to be almost but not 
> quite.
>
> On Nov 4, 2017, at 9:34 PM, William Corcoran <wlc at jctaylor.com 
> <mailto:wlc at jctaylor.com>> wrote:
>
>>
>> I had a similar issue with BSD 2.11.
>> So, I hope this helps.
>>
>> You can use vt100.   However, make sure your baud rate is 9600.
>>
>> stty 9600
>>
>> (SIMH makes the virtual adjustment so you don’t need to worry about 
>> SIMH settings.)
>>
>> I am using BSD 2.11 on a Mac with SIMH.  I noticed the console port 
>> would have a short window.   But, logging into dz ports were fine.   
>> Setting the console baud rate to match the dz11 settings fixed the issue.
>>
>> So, I suspect either a setting in vi that changes the window size on 
>> slow devices or it’s hard coded somewhere.  I will take a look.
>>
>> Truly,
>>
>> Bill Corcoran
>>
>>
>> On Nov 5, 2017, at 12:03 AM, Will Senn <will.senn at gmail.com 
>> <mailto:will.senn at gmail.com>> wrote:
>>
>>> What should I set TERM to on V8 to get the best results on my Mac 
>>> Terminal. If I set it to vt52, vt100, or vt132, only 8 lines appear 
>>> at the bottom of the terminal window (of about 24 lines):
>>>
>>> ---
>>>
>>> root::0:4:m0130,m322:/:
>>> daemon:x:1:1:m0000,m000:/:
>>> sys:sorry:2:1:m0130,m322:/usr/sys:no-login
>>> bin:sorry:3:4:m0130,m322:/bin:
>>> ken:sorry:6:1:m0130,m322:/usr/ken:
>>> dmr:sorry:7:4:mh1092,m069:/usr/dmr:
>>> nuucp::238:1:mh2019,m285,uucp:/usr/spool/uucppublic:/usr/lib/uucp/uucico
>>> uucp::48:1:mh2019,m285,nowitz:/usr/lib/uucp:
>>> "passwd" 20 lines, 770 characters
>>> ----
>>>
>>> The 8 line window works about like I'd expect - the arrow keys move 
>>> up and down until the screen needs to scroll, then B's and A's show 
>>> up. I'm used to that on BSD. Using the j and k keys work better and 
>>> when I scroll down enough lines, the lines move up to fill the whole 
>>> terminal window and the file can be edited in the full window. Is 
>>> there a better TERM setting that will get 24 lines to show up on 
>>> file open?
>>>
>>> Thanks,
>>>
>>> Will
>>>
>>> -- 
>>> GPG Fingerprint: 68F4 B3BD 1730 555A 4462  7D45 3EAA 5B6D A982 BAAF
>>>

-- 
GPG Fingerprint: 68F4 B3BD 1730 555A 4462  7D45 3EAA 5B6D A982 BAAF

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20171105/4ab41c44/attachment.html>

[TUHS] TERM for v8

[Clem cole]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 4255 bytes --]

I think the problem is before then.  You need to have the baud rate set before the shell is forked.  For serial ports like DZ and DH that is done in by the values of /etc/ttys   

Sent from my PDP-7 Running UNIX V0 expect things to be almost but not quite. 

> On Nov 5, 2017, at 6:01 AM, Will Senn <will.senn at gmail.com> wrote:
> 
> That makes sense. Any idea where to start looking if the console thinks the baudrate is 0 even after setting it (supposedly)? I looked at the source code for stty.c (briefly) and it looks like it oughta work, if ioctl works... other things work ok nl, cr, etc., just speed doesn't have an effect... When I get home later, I'll prolly try putting some print statments in stty.c to see what it thinks it's doing. Here's what I see on the command line in the console:
> 
> # stty
> speed: 0 baud
> erase = ^H; kill = @; intr = ^?; quit = ^\
> start = ^Q; stop = ^S; eof = ^D; brk <undef>
> old even odd -raw -nl echo -lcase -tabs -cbreak -tandem nl0 cr0 ff0 bs0 
> # stty 9600
> # stty
> speed: 0 baud
> erase = ^H; kill = @; intr = ^?; quit = ^\
> start = ^Q; stop = ^S; eof = ^D; brk <undef>
> old even odd -raw -nl echo -lcase -tabs -cbreak -tandem nl0 cr0 ff0 bs0 
> # 
> 
> Will
> 
>> On 11/5/17 12:44 AM, Clem cole wrote:
>> This is a feature of vi.   If the baud rate is perceived to be at or less than a specific value (1200 IIRC) the screen is set to 8 lines at a time to keep repainting low. 
>> 
>> This was useful on dialup lines.  
>> 
>> Since you have a virtual serial port the baud rate has no bearing other than upper level applications trying to make inferences about the environment such as this. 
>> 
>> If you set the baud rate to 19200 or 38400 (EXTA and EXTB in the virgin V7 code base) those were the fastest serial speeds in those days. So any SW should infer the ‘best’ behavior.
>> 
>> 
>> Sent from my PDP-7 Running UNIX V0 expect things to be almost but not quite. 
>> 
>> On Nov 4, 2017, at 9:34 PM, William Corcoran <wlc at jctaylor.com> wrote:
>> 
>>> 
>>> I had a similar issue with BSD 2.11. 
>>> So, I hope this helps. 
>>> 
>>> You can use vt100.   However, make sure your baud rate is 9600.
>>> 
>>> stty 9600
>>> 
>>> (SIMH makes the virtual adjustment so you don’t need to worry about SIMH settings.)
>>> 
>>> I am using BSD 2.11 on a Mac with SIMH.  I noticed the console port would have a short window.   But, logging into dz ports were fine.   Setting the console baud rate to match the dz11 settings fixed the issue.
>>> 
>>> So, I suspect either a setting in vi that changes the window size on slow devices or it’s hard coded somewhere.  I will take a look. 
>>> 
>>> Truly,
>>> 
>>> Bill Corcoran
>>> 
>>> 
>>> On Nov 5, 2017, at 12:03 AM, Will Senn <will.senn at gmail.com> wrote:
>>> 
>>>> What should I set TERM to on V8 to get the best results on my Mac Terminal. If I set it to vt52, vt100, or vt132, only 8 lines appear at the                     bottom of the terminal window (of about 24 lines):
>>>> 
>>>> ---
>>>> 
>>>> root::0:4:m0130,m322:/:
>>>> daemon:x:1:1:m0000,m000:/:
>>>> sys:sorry:2:1:m0130,m322:/usr/sys:no-login
>>>> bin:sorry:3:4:m0130,m322:/bin:
>>>> ken:sorry:6:1:m0130,m322:/usr/ken:
>>>> dmr:sorry:7:4:mh1092,m069:/usr/dmr:
>>>> nuucp::238:1:mh2019,m285,uucp:/usr/spool/uucppublic:/usr/lib/uucp/uucico
>>>> uucp::48:1:mh2019,m285,nowitz:/usr/lib/uucp:
>>>> "passwd" 20 lines, 770 characters
>>>> ----
>>>> 
>>>> The 8 line window works about like I'd expect - the arrow keys move up and down until the screen needs to scroll, then B's and A's show up. I'm used to that on BSD. Using the j and k keys work better and when I scroll down enough lines, the lines move                     up to fill the whole terminal window and the file can be edited in the full window. Is there a better                     TERM setting that will get 24 lines to show up on file open?
>>>> 
>>>> Thanks,
>>>> 
>>>> Will
>>>> 
>>>> -- 
>>>> GPG Fingerprint: 68F4 B3BD 1730 555A 4462  7D45 3EAA 5B6D A982 BAAF
>>>> 
> 
> -- 
> GPG Fingerprint: 68F4 B3BD 1730 555A 4462  7D45 3EAA 5B6D A982 BAAF
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20171105/5fdd81a1/attachment.html>

[TUHS] TERM for v8

[Will Senn]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 4795 bytes --]


I edited ttys:
cat /etc/ttys
1jconsole
12tty00

and rebooted, no effect (the j should mean exta or 19600). I also tried 
18console, but it didn't have an effect either.

Will


On 11/5/17 12:04 PM, Clem cole wrote:
> I think the problem is before then.  You need to have the baud rate 
> set before the shell is forked.  For serial ports like DZ and DH that 
> is done in by the values of /etc/ttys
>
> Sent from my PDP-7 Running UNIX V0 expect things to be almost but not 
> quite.
>
> On Nov 5, 2017, at 6:01 AM, Will Senn <will.senn at gmail.com 
> <mailto:will.senn at gmail.com>> wrote:
>
>> That makes sense. Any idea where to start looking if the console 
>> thinks the baudrate is 0 even after setting it (supposedly)? I looked 
>> at the source code for stty.c (briefly) and it looks like it oughta 
>> work, if ioctl works... other things work ok nl, cr, etc., just speed 
>> doesn't have an effect... When I get home later, I'll prolly try 
>> putting some print statments in stty.c to see what it thinks it's 
>> doing. Here's what I see on the command line in the console:
>>
>> # stty
>> speed: 0 baud
>> erase = ^H; kill = @; intr = ^?; quit = ^\
>> start = ^Q; stop = ^S; eof = ^D; brk <undef>
>> old even odd -raw -nl echo -lcase -tabs -cbreak -tandem nl0 cr0 ff0 bs0
>> # stty 9600
>> # stty
>> speed: 0 baud
>> erase = ^H; kill = @; intr = ^?; quit = ^\
>> start = ^Q; stop = ^S; eof = ^D; brk <undef>
>> old even odd -raw -nl echo -lcase -tabs -cbreak -tandem nl0 cr0 ff0 bs0
>> #
>>
>> Will
>>
>> On 11/5/17 12:44 AM, Clem cole wrote:
>>> This is a feature of vi.   If the baud rate is perceived to be at or 
>>> less than a specific value (1200 IIRC) the screen is set to 8 lines 
>>> at a time to keep repainting low.
>>>
>>> This was useful on dialup lines.
>>>
>>> Since you have a virtual serial port the baud rate has no bearing 
>>> other than upper level applications trying to make inferences about 
>>> the environment such as this.
>>>
>>> If you set the baud rate to 19200 or 38400 (EXTA and EXTB in the 
>>> virgin V7 code base) those were the fastest serial speeds in those 
>>> days. So any SW should infer the ‘best’ behavior.
>>>
>>>
>>> Sent from my PDP-7 Running UNIX V0 expect things to be almost but 
>>> not quite.
>>>
>>> On Nov 4, 2017, at 9:34 PM, William Corcoran <wlc at jctaylor.com 
>>> <mailto:wlc at jctaylor.com>> wrote:
>>>
>>>>
>>>> I had a similar issue with BSD 2.11.
>>>> So, I hope this helps.
>>>>
>>>> You can use vt100.   However, make sure your baud rate is 9600.
>>>>
>>>> stty 9600
>>>>
>>>> (SIMH makes the virtual adjustment so you don’t need to worry about 
>>>> SIMH settings.)
>>>>
>>>> I am using BSD 2.11 on a Mac with SIMH.  I noticed the console port 
>>>> would have a short window.   But, logging into dz ports were fine. 
>>>>   Setting the console baud rate to match the dz11 settings fixed 
>>>> the issue.
>>>>
>>>> So, I suspect either a setting in vi that changes the window size 
>>>> on slow devices or it’s hard coded somewhere.  I will take a look.
>>>>
>>>> Truly,
>>>>
>>>> Bill Corcoran
>>>>
>>>>
>>>> On Nov 5, 2017, at 12:03 AM, Will Senn <will.senn at gmail.com 
>>>> <mailto:will.senn at gmail.com>> wrote:
>>>>
>>>>> What should I set TERM to on V8 to get the best results on my Mac 
>>>>> Terminal. If I set it to vt52, vt100, or vt132, only 8 lines 
>>>>> appear at the bottom of the terminal window (of about 24 lines):
>>>>>
>>>>> ---
>>>>>
>>>>> root::0:4:m0130,m322:/:
>>>>> daemon:x:1:1:m0000,m000:/:
>>>>> sys:sorry:2:1:m0130,m322:/usr/sys:no-login
>>>>> bin:sorry:3:4:m0130,m322:/bin:
>>>>> ken:sorry:6:1:m0130,m322:/usr/ken:
>>>>> dmr:sorry:7:4:mh1092,m069:/usr/dmr:
>>>>> nuucp::238:1:mh2019,m285,uucp:/usr/spool/uucppublic:/usr/lib/uucp/uucico
>>>>> uucp::48:1:mh2019,m285,nowitz:/usr/lib/uucp:
>>>>> "passwd" 20 lines, 770 characters
>>>>> ----
>>>>>
>>>>> The 8 line window works about like I'd expect - the arrow keys 
>>>>> move up and down until the screen needs to scroll, then B's and 
>>>>> A's show up. I'm used to that on BSD. Using the j and k keys work 
>>>>> better and when I scroll down enough lines, the lines move up to 
>>>>> fill the whole terminal window and the file can be edited in the 
>>>>> full window. Is there a better TERM setting that will get 24 lines 
>>>>> to show up on file open?
>>>>>
>>>>> Thanks,
>>>>>
>>>>> Will
>>>>>
>>>>> -- 
>>>>> GPG Fingerprint: 68F4 B3BD 1730 555A 4462  7D45 3EAA 5B6D A982 BAAF
>>>>>
>>
>> -- 
>> GPG Fingerprint: 68F4 B3BD 1730 555A 4462  7D45 3EAA 5B6D A982 BAAF


-- 
GPG Fingerprint: 68F4 B3BD 1730 555A 4462  7D45 3EAA 5B6D A982 BAAF

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20171105/2a42e8fd/attachment.html>

[TUHS] TERM for v8

[Bakul Shah]

Based on the following snippet my guess is that the kernel
doesn't report the right baud rate for the console.

> >> # stty 9600
> >> # stty
> >> speed: 0 baud

[TUHS] TERM for v8

[Will Senn]

On 11/5/17 1:00 PM, Bakul Shah wrote:
> Based on the following snippet my guess is that the kernel
> doesn't report the right baud rate for the console.
>
>>>> # stty 9600
>>>> # stty
>>>> speed: 0 baud

Yeabut, it works fine on tty00 (the DZ):

# stty exta
# stty
speed: 19200 baud

...


-- 
GPG Fingerprint: 68F4 B3BD 1730 555A 4462  7D45 3EAA 5B6D A982 BAAF

[TUHS] TERM for v8

[Bakul Shah]

On Sun, 05 Nov 2017 13:28:18 -0600 Will Senn <will.senn at gmail.com> wrote:
Will Senn writes:
> On 11/5/17 1:00 PM, Bakul Shah wrote:
> > Based on the following snippet my guess is that the kernel
> > doesn't report the right baud rate for the console.
> >
> >>>> # stty 9600
> >>>> # stty
> >>>> speed: 0 baud
> 
> Yeabut, it works fine on tty00 (the DZ):
> 
> # stty exta
> # stty
> speed: 19200 baud
> 
> ...

I have no clue about v8 but at least in v7 these were
different drivers.  Looking at v7/usr/sys/dev/kl.c I don't see
klioctl() like dzioctl() in the dz driver. So my guess is
console ioctls return EINVAL.

[TUHS] TERM for v8

[Paul Winalski]

On 11/5/17, Bakul Shah <bakul at bitblocks.com> wrote:
>
> I have no clue about v8 but at least in v7 these were
> different drivers.  Looking at v7/usr/sys/dev/kl.c I don't see
> klioctl() like dzioctl() in the dz driver. So my guess is
> console ioctls return EINVAL.
>
That would seem to indicate that stty isn't checking for ioctl()
failure.  Or at least it isn't displaying an error message if that
happens.

-Paul W.

[TUHS] TERM for v8

[Bakul Shah]

On Sun, 05 Nov 2017 15:42:19 -0500 Paul Winalski <paul.winalski at gmail.com> wrote:
Paul Winalski writes:
> On 11/5/17, Bakul Shah <bakul at bitblocks.com> wrote:
> >
> > I have no clue about v8 but at least in v7 these were
> > different drivers.  Looking at v7/usr/sys/dev/kl.c I don't see
> > klioctl() like dzioctl() in the dz driver. So my guess is
> > console ioctls return EINVAL.
> >
> That would seem to indicate that stty isn't checking for ioctl()
> failure.  Or at least it isn't displaying an error message if that
> happens.

Yes, v8 stty ignores all ioctl results! v8 has cons.c for the
console device and it basically ignores ioctls.