IRC: #techbytes @ FreeNode: Friday, January 15, 2021

Techrights-sec

The IPFS bandwidh reports in IRC are still "âââ" and not ASCII graphics.

http://techrights.org/irc-archives/irc-log-techrights-140121.html#tJan%2014%2023:20:02

Techrights-sec

-TechBytesBot/#techbytes-techrights.org | IRC: #techrights @ FreeNode: Thursday, January 14, 2021

https://theasciicode.com.ar/extended-ascii-code/graphic-character-low-density-dotted-ascii-code-176.html

schestowitz

I got those to work OK in most IRC clients that I have tested with. However, the python program I use for HTML logs does not cope. It's something in the encoding. One option I have is to write some program to convert, but an easier workaround is leaving out all those lines, or blanking out bits of them. Those are standard unicode characters, so another option is to experiment with another type of indicator for volume. What can be used

Jan 15 07:27

schestowitz

other than horizontal bars? Hex?

Jan 15 07:27

Techrights-sec

Jan 15 07:28

-TechBytesBot/#techbytes-theasciicode.com.ar | ASCII code Graphic character, low density dotted, American Standard Code for Information Interchange, The complete ASCII table, characters,letters, vowels with accents, consonants, signs, symbols, numbers graphic, character, low, density, dotted,ascii,176, ascii art, ascii table, code ascii, ascii character, ascii text, ascii chart, ascii characters, ascii codes, characters, codes, tables, symbols, list, alt, keys, keyb

Jan 15 07:28

schestowitz

I considered those, but the gradients are 3 or 4 in scale, unlike 0-8

Jan 15 07:29

Techrights-sec

https://theasciicode.com.ar/extended-ascii-code/graphic-character-low-density-dotted-ascii-code-176.html

Jan 15 07:30

Techrights-sec

That has 219 and 220 and32 -- in decimal

Jan 15 07:30

Techrights-sec

Otherwise python's unicode support needs to be figured to get 0 to 8 scale

Jan 15 07:30

schestowitz

I am going to experiment this morning, we start workout shortly, then I need to record InteLeaks (my new routine)

Techrights-sec

It might be tricker to stack them in two rows. That would give 0 to 8 but

Techrights-sec

only if they line up.

schestowitz

That would add more clutter to IRC, doubling it

schestowitz

I also wrote a 100-line python program to make the LEDs pulsate 24/7 to show up/down traffic

Jan 15 07:32

Techrights-sec

I haven't used python3 for text processing at all. I was going to

Jan 15 07:33

Techrights-sec

move to rakulang from perl5 instead.

Jan 15 07:33

schestowitz

In the page you have cited, the chars I use and the ones you have highlighted (and I had originally considered) are both under "Extended ASCII", so good chance is, python would choke on those too

Jan 15 07:34

Techrights-sec

U+2581 through U+2588 would work, but those are what you are trying?

Jan 15 07:35

schestowitz

What I really die for right now is a GOOD, proper image->unicode covertor, for images in bulletins. I know some exist, but nothing on Debian and not free software. The ones I found are dumb and shitty algos

Techrights-sec

I presume you've seen the main documentation?

https://docs.python.org/3/howto/unicode.html

Techrights-sec

-TechBytesBot/#techbytes-docs.python.org | Unicode HOWTO — Python 3.9.1 documentation

schestowitz

I did not check python docs. One option is to sed those 8 symbols

Techrights-sec

Sed would be a bad choice, it is a matter of setting the encoding / decoding

Techrights-sec

correctly for input, I would guess.

Techrights-sec

https://docs.python.org/3/howto/unicode.html#reading-and-writing-unicode-data

schestowitz

That might introduce new side issues, so I am reluctant, let me check the code...

Jan 15 07:38

schestowitz

#!/usr/bin/env python

Jan 15 07:39

schestowitz

# -*- coding: utf-8 -*-

Jan 15 07:39

schestowitz

funny thing is, one original author was jeff waugh (yeah, that one)

Jan 15 07:40

Techrights-sec

Jan 15 07:42

-TechBytesBot/#techbytes-docs.python.org | Unicode HOWTO — Python 3.9.1 documentation

Jan 15 07:42

schestowitz

(to self) TODO: 1) bulletin image->unicode (did lots of searching, installed programs, used web-based stuff... all crap), 2) encoding issue for bot, 3) connection issues.

Jan 15 07:45

schestowitz

I think I'm OK with the tools I use for videos now. OBS wasn't explored, but I'm fine with what already works anyway and it's actively maintained, too...

Jan 15 07:46

schestowitz

I've just grepped logs, over 50,000 hits on videos in techrights over the past 5 days

Jan 15 07:47

schestowitz

seems the issue boils down to xchat, not python

Jan 15 07:52

schestowitz

Jan 14 02:20:02 <techrights-ipfs> â–• IPFS downstream, 60 minutes: â–ƒâ–‚â–‚â–ƒâ–„â–…â–…â–ƒâ–‚â–…â–ƒâ–„â–‚â–‚â–‚â–ƒâ–‚â–ƒâ–…â–‚â–„â–^aâ–ƒâ–ƒâ–‚â–ƒâ–„â–‚â–ƒâ–‚â–„â–‚â–ƒâ–„$

Jan 15 07:53

schestowitz

Jan 14 02:36:41 * mmu_man has quit (Ping timeout: 256 seconds)

Jan 15 07:53

schestowitz

This is strange as originally those things showed up fine in xchat, but not in konversation, and now it is the opposite

Jan 15 07:53

schestowitz

yes stranger: on another machine of mine xchat handles fine with it, maybe it's an issue with my local configs in xchat

Jan 15 07:55

schestowitz

test: ▅▆▅▆▅▅▆▆▆▆▇▅▇▆▅▆▇▇▆▅▆▅▆▆▅▆▇▆▆▇▇▆▆▅

Jan 15 08:04

schestowitz

▕

Jan 15 08:04

schestowitz

[07:20] <techrights-ipfs> ▕ IPFS downstream, 60 minutes: ▅▆▅▆▅▅▆▆▆▆▇▅▇▆▅▆▇▇▆▅▆▅▆▆▅▆▇▆▆▇▇▆▆▅▆▇▆▆▅▅▅▅▇▆▇▆▆▇▆▆▇▅▆▆▇█▇▇▇▇▇▆▆▇▆ ▕ IPFS upstream, 60 minutes: ▃▄▂▂▃▃▄▅▃▃▂▃▂▂▄▃▂▃▄▅▂▅▃▃▄▂▃▃▄▃▄▃▃▃▃▂▂▃▂▂▄▂▂▂▃▃▄▂▃▂

Jan 15 08:04

Techrights-sec

I am not familiar enough with the parts to guess. I would otherwise check

Techrights-sec

the data at each stage of work to see what is actually present.

schestowitz

I have just checked on 3 laptops. xchat-gnome deals OK with it, not 2019 build for debian buster

schestowitz

I at least narrowed that down to the IRC client itself, the input files that go into python are already garbage

schestowitz

I will experiment for a bit