[arch-general] archlinux.org/pacman/pacman.conf.5.html from Aug-2019

[David C. Rankin]

On 6/3/21 1:00 PM, David Rosenstrauch via arch-general wrote:
> 
> 
> On 6/3/21 1:09 PM, David C. Rankin via arch-general wrote:
>> I did all 3 updates from different tabs in the same konsole (KDE), so the
>> terminal is VT100 compatible. It really looked like there was a bad ANSI
>> escape somewhere. It was night/day different than the last time I updated and
>> I've probably update each of these boxes of 1000 times during the past 10
>> years or so.
> 
> My guess would be that the old (KDE Trinity) version of konsole that you're
> using doesn't support some modern feature of terminals.
> 
> DR

No, that's not it either, I have used ALL ANSI escapes (and tput, etc..) in
the terminals without issue.

The key here seems to relate to the "Total" line triggering a new line part
way through the package download. The faster the host, the less noticeable the
jitter. This is from the newer box that I though was NOT affected:

 linux-docs-5.12.9.arch1-1-x86_64
 Total (0/4)                                0.0   B  0.00   B/s 00:00
[--------------------------------------]   0%
 linux-lts-5.10.42-1-x86_64x86_64          22.0 MiB  5.75 MiB/s 00:04
[######################################] 100%
 Total (1/4)                               20.9 MiB  6.08 MiB/s 00:17
[######--------------------------------]  16%
 linux-lts-headers-5.10.42-1-x86_64        74.3 MiB  5.85 MiB/s 00:13
[######################################] 100%
 Total (2/4)                               95.5 MiB  6.08 MiB/s 00:05
[############################----------]  74%
 nodejs-16.3.0-1-x86_64.42-1-x86_64        22.7 MiB  5.57 MiB/s 00:04
[######################################] 100%
 Total (3/4)                              117.8 MiB  6.06 MiB/s 00:01
[##################################----]  92%
 Total (4/4)                              127.8 MiB  5.73 MiB/s 00:22
[######################################] 100%

It was affected, it just wasn't as noticeable when pacman was running because
the screen refresh was fast.

Also note, I do all updates over SSH (and Wifi) to headless boxes, so there
many be some network delay contributing as well.

But the point being, for 12 years I have never seen anything like this issue,
and now, it is quite notable.


-- 
David C. Rankin, J.D.,P.E.