Re: [mpg123-devel] Please test new network code in/out of mpg123 (now with HTTPS support)!

[Dave Y. <dav...@gm...>]

On 05/19/22 12:02 AM, Thomas Orgis wrote:
> Am Wed, 18 May 2022 18:52:27 -0700
> schrieb Dave Yeo <dav...@gm...>:
>
>> On 05/18/22 08:12 AM, Thomas Orgis wrote:
>
>> Ok, operator error previously. It consistently outputs "Note: stderr
>> terminal width: 0" no matter the terminal width.
>
> So … what do we do with that?! Is our code broken? Isn't OS/2 able to
> provide a terminal width?!

OK, I'd left a stray #ifndef __OS2__ in term_posix.c, deleting it, and 
doing a svn up fixes that, examples,
Note: stderr terminal width: 80
Note: stderr terminal width: 87.


>
> #ifdef __OS2__
> 	int s[2];
> 	_scrsize (s);
> 	if (s[0] >= 0)
> 		return s[0];
> #else
>
> What is wrong here?

Forgetting to clean up after previous testing.

>
>> What it does is cause an extra blank line below the xxx's before the
>> prompt returns. Change the 80 to 79 and that line goes away.
>> Seems that there is a CR or such printed at the end of the line so
>> effectively it is 79 chars without scrolling.
>
> A CR shall not trigger advancing a line. That's the point of it. If
> this is the case, it's a bug in your terminal. So we need to solve this
> second problem. I can do an ifdef(__OS2__) limit = limit - 1; … but
> first, we need to be able to get a proper width value!
>

OK, things have changed with reverting term_posix.c, it always scrolls 
with v, on the other hand the progress meter is now working and the 
display of Title: Artist: etc is now correctly formatted.
The terminal, or officially a VIO window is more like a DOS display then 
a regular window, at that running a DOS program in a window has much the 
same behaviour. I think the cursor just advances to the next character 
which if at the end of the screen means going to the next line.
Sorry about missing the patch in term_posix.c, forgot about SVN ignoring 
changes.
Dave