Posts on jacksonchen666.com

Disclaimer: This Blog May Be Bad and Incorrect

Jackson — Sat, 27 Jan 2024 00:17:49 +0000

Today I just decided to read some of my old blog posts and I was like “wtf”

So I decided to make a dedicated blog post to say “wtf” and also “don’t only rely on me”.

If you feel like you can trust me for all of your information, think again. I got the amount of words in my passphrase wrong (in the post about my passwords). I actually got that wrong multiple times until I realized that I was wrong.

There’s a non-zero chance that I have made factual (and grammatical) errors in my other blog posts. So if you are in doubt on whether to trust me, my answer would be “don’t only trust me, verify me”.

QR Code Shenanigans

Jackson — Wed, 24 Jan 2024 14:14:53 +0000

If you couldn’t tell, it is indeed a QR code. An absurdly sized QR code, in fact.

Actually, it doesn’t even have to be that big, I just made it that big because I like being absurd sometimes.

So you may ask: How did you do that?!?!

Well here’s the command I ran:

qrencode -l H -v 40 https://jacksonchen666.com/posts/2024-01-24/14-14-53/ -o content/posts/2024-01-24-qr-code-shenanigans/v40-qr-code.png

It uses qrencode, and specifies -v 40 which tells qrencode to make a QR code which is version 40.

Version 40 of a QR code is the biggest QR code you can get. It can store about 3 kilobytes of stuff, which won’t fit pictures because pictures those tend to be very big.

But the thing is: The data in the QR code is https://jacksonchen666.com/posts/2024-01-24/14-14-53/, not something that takes up anywhere close to 3 kilobytes. So like, why?

Well, because I can and I thought it would be funny and ridiculous, my kind of thing.

2 Years of Sysadmining

Jackson — Wed, 17 Jan 2024 23:34:52 +0000

2 years of sysadmining

On 2021-09-04T12:37Z, I installed Fedora Linux on an Asus laptop. That is pretty much the start of my server adventures.

That date comes from dnf history. It keeps track of packages you install, uninstall, or whatever. It also pretty much never purges the history, so you pretty much always have the first package management thing in your history, which is probably the first time you installed Fedora Linux.

In my initial server setup, I was interested in matrix synapse. Then eventually I found Mastodon, more things (calendars, PeerTube, etc.), and a lot more things happened since then.

And today, I have 2 computers which pretty much monitor each other (both computers run on the same internet and power lines, so it’s not necessarily redundant). I have a crap ton of subdomains with a bunch of crap as well, and questionable infrastructure (status. subdomain has “redundancy” if you use Tor Onions).

Recent downtime

Recently, I had downtime. For some reason, the internet connection just died, and then later came back. I didn’t do anything on my side for that to be fixed.

However, with the internet connection being interrupted then resumed, I uncovered a bug in Out of your Element where it would just tie up all the RAM and swap trying to connect to Discord or something. It was weird.

A new wiki

On the note of running new things, I’ve got a new thing: https://wiki.jacksonchen666.com/.

It’s a wiki where I dump my knowledge. I do feel as though a wiki is much more frictionless than having to deal with the publishing mess that is my website (having to use git for a website with blog posts is not the best thing).

I’ve already gotten a few pages up written.

The goal of the wiki isn’t to replace Wikipedia or anything like that, actually. It’s usually for when I suddenly recall something, and want it to be written down somewhere in a quick notice.

Blog posts though, I will probably continue writing them if I feel that’s appropriate. Just know that I don’t plan to do the 100 days to offload challenge anymore (it was kind of hell when I tried (and succeeded)), so there shouldn’t be much of a blog post spam this year.

Website dead URLs

Yes, I check the things I link. I also do stuff about links when a URL I linked is dead.

I’ve recently done a check after the extended downtime of sourcehut, and probably after a while too.

Verifying Links in Mastodon Using More Mastodon

Jackson — Mon, 08 Jan 2024 10:13:47 +0000

Skip explanation and go to relevant section

Verifying links in your profile fields

In Mastodon (and probably the wider parts of the Fediverse as well), you can put links in your profile fields. Those links can also be verified.

Mastodon gives this thing that it asks you to put into your website:

<a rel="me" href="https://example.com/@example">Mastodon</a>

The important part is the rel="me" part. It says “yes, this is me”.

Of course, it would have to link to your actual Mastodon profile’s URL, the above is just an example. But place that into your website and link your website in your profile fields and you’ll have links verified.

That’s the stuff you probably already know.

Verifying links in your profile fields without having the link visible

The HTML snippet I just gave you is one way to do it. Another way to do it is like this:

<link rel="me" href="https://example.com/@example">

Put that within your <head> element and now you can have links verified while actually having no links visible to the user.

Mastodon does talk about this, but doesn’t give examples. If you needed one, there you go.

Verifying you control another Mastodon/Fedi account

What if you wanted to prove that you control an alt account? Or some other account?

For this example, there will be Account 1 and Account 2:

On Account 1, put the URL for Account 2 in a profile field
On Account 2, put the URL for Account 1 in a profile field
On both accounts, save your profile again (no changes needed, just save it). Then you should have the verification check marks on both accounts.

Important things:

You cannot use your fediverse handle instead of a URL to the profile. It just won’t work.
Both accounts must link to each other.

The “Account 1” and “Account 2” in this example can be any fedi account, provided you control both of them. You can also have any number of fedi accounts linked, provided you have enough profile fields and put links to each profile in all profiles. It can get complicated quite fast.

More Notes on Replacing the Fairphone 4 Screen

Jackson — Tue, 02 Jan 2024 13:00:05 +0000

So this time, I actually did a real screen replacement with a real screen! And I have some more notes based on the previous notes

Last time, I did a “replacement” i.e. Did the replacement procedures but used the same screen. This time, I did actually replace the screen with another screen.

This time, I wanted to fix a couple of things:

Multi-touch unreliability
Dead pixels (tiny and not noticeable during normal use)
Replace my screen protector because I broke it again (I’m on the 6th screen protector so far, I think)

So I ordered a replacement screen and also screen protector. Then it arrived via being fucking thrown over the home gate. Thankfully, the parts did not get damaged after being thrown by the delivery person, as the box had too much padding and the replacement screen had bubble wrap.

The notes part

This is in addition to the previous notes

When taking the screen replacement part out of its packaging, there might be a tiny sticky cover on where the front camera is. Take that sticky cover off, you don’t want that there.
A screw in the lower right side of the phone (relative to looking at the back of the opened phone) was resisting to be screwed back in. I had to use a lot more than usual force to get the screw back in. I may have screwed up that single screw in the process.
I had difficulty trying to plug the display back in. I think I just had to push on the connector a little harder for it to connect back together.

After replacement, the replacement has been done. Obviously.

Testing for problems after replacement

I mentioned earlier that I had problems with multi-touch. That is, any finger that isn’t my thumb has very unreliable touchscreen detection.

Well, guess what? Replacing the screen did absolutely nothing. In fact, the same workarounds I found still works on the new screen:

Placing the phone on my MacBook
Plugging headphones in (not wirelessly)

Any one of those workarounds can be used, and it does not have to be all of them.

Multi-touch is important to me because I play rhythm games. In the current state, I would need to play harder levels to progress in skill, and those harder levels require more than just thumbs. But because anything other than thumbs is unreliable (and I need to use more than just thumbs), I can’t properly play.

Dealing with the old part

Now that I have swapped the screen with a new one, what do I do with the old one?

Answer would be to send it off to some electronics recycling center, but I have no idea how to do any of that, and I don’t know where to send it off to.

So I am just going to keep the old display until I deal with it eventually.

Monitor Resolutions and Sizes Review

Jackson — Mon, 01 Jan 2024 23:38:54 +0100

As a very computer person, I have used monitors with differing resolutions and sizes.

Today, I have 2 comparisons to make:

A random 21-inch (or 24-inch) 1920x1080 (HD 1080p) monitor compared to a random 27-inch 3840x2160 (4k) monitor
My MacBook’s internal display compared to my Fairphone 4

21-inch 1080p vs. 27-inch 4k

(I am not exactly sure of the exact inches of my previous 1080p monitor, and it might be 24-inch. So I’ll include a 24-inch 1080p monitor in the comparison.)

My previous monitor was a 21-inch 1920x1080 (1080p) monitor. Because the text wasn’t really that sharp to my liking (I have a MacBook with its sharp internal display, after all), I decided to get a 4k monitor.

The 4k monitor I happened to get is an LS27A70 with the Samsung brand on it. Feel free to look up whatever that is, but the monitor is a 27-inch 3840x2160 monitor.

Putting the numbers into a random online Pixels Per Inch calculator, it returns 104.90 PPI for the 1080p monitor and 163.18 PPI for the 4k monitor.

(Assuming that the 1080p monitor was actually 24-inch, it would be 91.79 PPI)

Well, what do those numbers even mean?

Actually, it doesn’t say much about how sharp text will look on those monitors.

With the 1080p monitor, I can pretty much see the tiny pixels when sitting quite close to it but still being able to see the entire monitor. I don’t like not very sharp text, so I got a 4k monitor.

In my experience with the 27-inch 4k, I have to sit far away enough to see the entire monitor. At that point, the text is pretty sharp and I wouldn’t be able to tell apart the tiny pixels. Looking any closer at the display is pretty abnormal (for me, at least), but does mean I can see the pixels more.

My MacBook vs Fairphone 4

My MacBook has a 13.3 inch 2560x1600 display. The Fairphone 4 has a 1080x2340 display when in portrait mode, and 2340x1080 in landscape mode.

These are getting into “weird resolutions territory”. MacBooks tend to have a 16:10 aspect ratio, and my Fairphone 4 has an aspect ratio of 19.5:9.

Basically, MacBooks have a slightly taller displays than 16:9 monitors (the 1080p and 4k monitors that I talked about earlier), and the Fairphone 4 has a wider display in landscape mode.

Number-wise, my MacBook has a 226.98 PPI display, and the Fairphone 4 has a 409.08 PPI display.

Experience-wise, both have sharp text at reasonable viewing distances. I can start seeing the pixels on my MacBook if I stared at it closely, but on the Fairphone 4? I can’t see pixels even if I tried, except for dead pixels at max brightness in very dark rooms.

Recap

I primarily use a 13-inch MacBook, which has pretty sharp text
A 4k monitor (27-inch) had sharper text compared to the 1080p monitor (21-inch) I had
The Fairphone 4 display (6.3 inch, 1080p but wider in landscape) has a PPI that is way too high for me to see individual pixels (except for dead pixels with max brightness in dark rooms)
The 4k monitor, my MacBook, and the Fairphone 4 all has pretty sharp text, with the baseline being my MacBook

Overall, higher pixels per inch means more pixels to make sharper text, although anything above a MacBook could start being too much (e.g. Fairphone 4).

Also, there’s another factor to account for: Viewing distance. I have not accounted for that, and I think that’s an important factor.

Who Am I?

Jackson — Sun, 31 Dec 2023 23:27:14 +0000

Warning: If you expected a computer related post or an introduction of me, this post is neither. Instead, this is a very recent experience I just had and wanted to just write about for some reason.

So I sit down at my computer, and start doing random stuff (as usual). I check my RSS feeds, and I notice one of my feed being from koshka.love. Of course, out of curiosity (and because I forgot about what it was again), I navigate to it and them got myself into the autism acceptance rabbit hole again.

As I keep reading- wait no, scrambling and solving my rubiks cube while listening to breakcore music from hkmori and simultaneously having Firefox (actually LibreWolf) reader view read out the words in a post, I listen to the words being said. Maybe understanding some of the sentences, or maybe just forgetting it and leaving it somewhere in my brain.

Then I have to go to the bathroom to do bathroom stuff. So I pause the reader view from reading by pressing N, press the pause button for the music, and finally move my mouse cursor to the bottom left to activate the screen saver and lock my computer. Then I head to the bathroom.

And then I am reminded of my existence.

I am a brain trapped in a human body. I am also using different senses to navigate and stuff like that, limbs to interact with things. You know, stuff like a normal human does and has.

Turns out, I am not a faceless entity that is infinitely thin in all dimensions. I am me. Myself. Bigger than nothing.

However, during all that reading and stimming, I was just am. I am. No “me”, I just am. Not in terms of time, I was just… being. And ingesting all the words and sentences from those posts.

Also, I have a face. That shouldn’t be much of a surprise, but what kind of is a surprise is that I avoid eye contact with myself in mirrors or at all. I don’t like looking at me, and that avoidance is done subconsciously.

Right now, I am staring at a thing called a “monitor”. It is basically the computer screen. The screen itself also blurs any reflection, so any (subconscious) attempts to use the monitor and reflections to see my face will inevitably fail, since it’s so blurry you cannot make details out of it.

That’s great, because then I don’t have to be reminded of my face. Although that applies to the external monitor that I have. In the case of the internal display of the computer, it does not blur reflections or anything really. So I can indeed try and see myself, and I would be able to succeed.

Speaking of my face, why do I not like it?

Well, I have no idea. Maybe I don’t want to be look like that gender or look like that type of people or whatever. Maybe I am reminded of something I cannot recall, except for my subconscious.

And I’m now here writing a post about realizing my existence (after having a brief pause on “escapism”) and why I just don’t really like my face. And the conclusion is: I don’t understand why either.

osu!lazer and Health Drain

Jackson — Sun, 17 Dec 2023 01:47:06 +0000

Important preface

I exclusively use osu!lazer. I have never touched osu!stable in my life (nor with my account), ever. Everything here is in the context of osu!lazer, not osu!stable.

Brief history of my skills in osu!

I joined osu! in 2021-10-27 by creating an account (my osu! account, by the way). By now, I have almost 200 hours or 8 days of playtime for the osu! game mode.

Previously, I had around 60 or 70 percent hit accuracy, and I can do maybe 5 or 6 star maps with a grade of D (less than 70% accuracy in osu!lazer), and not fail.

Or, at least I could.

An osu!lazer update

There was an osu!lazer update on 2023-11-30 that included a change to health drain in the osu! game mode.

Then suddenly after that change and update, I suck at osu!. Much worse than before. Now I’m failing at maps I could previously pass. Like the 5 and 6 star maps that I could pass before the update is now a fail, and my replays are ruined since they are wrong.

Oh no, my self-esteem is completely ruined. /s

…

Wait, osu!catch??

Looking at my osu!catch stats

Yes, I did indeed play osu!catch. I have about 37 hours of playtime and 91% hit accuracy. How the heck am I better at osu!catch?

Well, I don’t know. But in terms of accuracy and grade, I’m pretty good! Well, at least for maps in the 3 or 4 star range.

So how the heck do I have better grades in osu!catch with less playtime? (albeit with less hard maps “compared” to osu!)

Playing easier maps, and getting way better

Well, “easier” and “better”. It’s more like osu! maps which are actually for my skill level, not something that is way beyond my skill level. Those maps are in the 3 to 4 star range.

By playing those “easier” maps and getting “better”, I have actually increased my stats in the game by a lot.

On 2023-12-08, my rank (at least on lazer.ppy.sh) was about 260k. Then 2 days later on 2023-12-10, my rank suddenly jumped to 210k. Today (2023-12-17), it’s 200k.

My maximum combo has reached 799. That would at least mean I would get the medal for 500 combo at least, but since medals aren’t a thing yet in osu!lazer (coming soon(tm)), I didn’t get one.

My hit accuracy has also gotten much higher, going from around the 60% to 70% range to 86.98% right now.

Now that I compare to my osu!catch plays, it’s kind of the same story: I’m actually good at this game (at least at the level I’m currently at).

I also even gained performance points. When I played maps not for my skills, I just never got any at all.

Conclusion

I just had to play the game at the level that worked for me. Well, it was fun trying to play the game with fun maps, but I guess I’m going on numbers instead now.

What it takes to setup a new domain name (for me)

Jackson — Fri, 01 Dec 2023 19:29:37 +0000

While I’m trying to get certificates for reinfo.wiki, I’m going to talk about setting up new domain names for stuff.

Usually, I may put my stuff under a subdomain on jacksonchen666.com, like videos.jacksonchen666.com or status.jacksonchen666.com or files.jacksonchen666.com.

Those kinds of new domains are easy to deal with:

Add more nginx configuration
Use a subdomain as the server_name

And that’s it. The server software part is pretty specific and setup varies, so that’s not included in the list.

But what about the cases of new different domains? Like reinfo.wiki?

Well, that involves a lot more setup. Here’s the case for reinfo.wiki:

Get the domain name
Setup authoritative nameserver (if not setup already)
Add a domain in deSEC.io
Setup DNSSEC
Setup DNS records
Get certificates
Setup the software that’s gonna run the thing (already done)
Setup the Tor Onions
Add nginx configuration
Do HSTS preloading
Update prometheus and blackbox configuration (the uptime checker)
Update links in many places (specific for reinfo wiki as it already existed)
Add CAA records

Yeah, there’s a lot of work involved in a new domain name that isn’t just a subdomain.

So yeah that’s it I guess. There’s not much special about the setup.

The Logistical Mess of Trying to Process a Lot of Videos

Jackson — Sun, 26 Nov 2023 14:51:57 +0000

Warning: Rushed blog post. Yeah.

The last time vs Now

Remember last time? Yeah this time I’m doing it for over 700 gigabytes of videos, with more specific parameters to ffmpeg and with more being added to the pile of videos.

Again, it’s gameplay videos (because turns out I’m back to playing that game again). This time, with more specific parameters.

Goals

This time, I wanted to do the following:

Preserve 4k resolution (for text clarity)
Make the quality of the videos OK-ish but not horrible
Keep 60 fps
Use much less space than the original recording

I settled on using -crf 29 -preset slow to keep file sizes low and quality OK enough.

Not enough time

However, there’s a major problem with -preset slow.

It’s slow (duh). Like way too slow for me to process a single hour video in less than 10 hours.

After trying to process a single video in many weeks, I decided that purely my Macbook with maybe 8 hours of availability to process videos was not enough. I have to go farther with the amount of time available.

An idea

After some thought, I decided that I will now have to use every other computer that is available to me. Here’s the list of hostnames of computers I could use for video processing:

imac
stupid-desktop
stupid-laptop

Actual setup

The actual setup was a bit different.

I tried using laptop-server, but I forgot it doesn’t have more RAM and ffmpeg needed 2 or 4 gigabytes of RAM for some reason. So I use it as a middleman of all the video files (using an external SSD), and as a place to hold videos while I’m away.

The computers were setup to run in a while loop, constantly going over all videos that haven’t been processed yet, then processing them.

Now for the other shenanigans.

File synchronization

To get files from my SSD to the computers, I used laptop-server as an middleman with another SSD plugged in. It’s actually my old 2TB Sandisk SSD before I needed to get a larger SSD for more crap.

I would transfer video the file onto the SSD plugged into laptop-server, then make the processing computers pull the video files.

I would then transfer the finished video files back onto laptop-server, then back to my actual SSD where I store my crap.

This extra roundtrip is pretty inefficient, and I have (technically) eliminated the extra roundtrip of transferring files onto the processing computers. But this middleman place is kind of used as a place to hold files before I get the time to actually grab it.

It also allows for using a single rsync command to grab all the processed videos if all of it is on the middleman. And I did use rsync for all of this file transfer thing.

If ran on `LockedMac`

The actual time it took to process 1 hour of video is about 12 hours, which means processing over 72 hours of video would take a long fucking time.

So processing it on my Macbook (LockedMac) would’ve been a major pain and would take way too long.

actually it will still take too long whatever just continue with this ok

Charging for `stupid-laptop`

Of course, due to resource constraints (lack of cables for charging stupid-laptop), I also had to rearrange my charging setup a little bit:

Charging brick for f had to be given to stupid-laptop
Charging cable for f was plugged into stupid-desktop
Another charging cable was provided to stupid-laptop

This was what I can do with what I had. Not much, and it will change when I have to take the charging brick (although not much and not very disruptive).

Speedrunning an install on stupid-desktop

I didn’t have an OS installed on stupid-desktop. So I had to install it.

I followed my own blog post for Alpine Linux with /boot/ and / encrypted, and finished it before I had to leave to do other things. It took about half an hour or an hour.

Doing all the commands was quite manual work, so later I decided to make a script that does all of it. It’s unlikely to be as robust as Alpine’s setup-disk, but hey it works (although I haven’t yet published it).

Networking for `stupid-desktop`

stupid-desktop is a desktop computer. It does not have Wi-Fi, so I had to give it my Ethernet cable so it can get access to the files.

That meant for LockedMac, it would have to use Wi-Fi. Which I had experienced many problems in the past few days just trying to play games.

Storage and space management

I also had to manage the space available and used.

Each hour of unprocessed video is around 20 gigabytes. Yes, 20 gigabytes per hour. That’s about 44444 kilobits per second. Very intense footage, and it makes sense given that it’s also in 4k 60 fps.

I could maybe give each computer 100 gigabytes before there is concern about running out of space for processed videos. So currently, all computers have 100 gigabytes of videos to deal with.

Long-term solutions

This kind of setup was actually for the short term to process everything I had right now and quickly (hopefully).

The longer-term would be to get a computer that can handle a stupid amount of CPU transcoding crap (and other intensive stuff), then use that for my PeerTube and whatever else.

Recovering From Syncthing Deleting My Data

Jackson — Sun, 19 Nov 2023 22:49:58 +0100

Syncthing-fork

So I decided to try Syncthing-fork to see if it was any better at battery life (spoilers: I have no idea).

To my surprise, it has much more tweaking options and stuff, plus more information.

However, I now wanted to migrate from Syncthing to Syncthing-fork because now I wanted to use Syncthing-fork.

The easiest way was to just export the config from Syncthing and import it into Syncthing-fork. Easy and done.

Data loss

I explore around the app, saw a folder with the “revert local changes” button, promptly pressed it and didn’t think about it more.

Much later, I wanted to use my password manager to login to GitHub to find the code that defines the mode function in libqalculate (actually the specifics don’t matter, I just wanted to use my password manager). I open my password manager app, and it says file not found.

So I went ahead and looked in my Sync folder (at the root of user facing internal storage). I found… nothing but .stfolder and .stversions.

Well, fuck. I’ve just lost data, and setting it all back up was going to be a pain.

Complaints sent

I file an issue about the data loss I had, then move on to some other thing.

Data recovery

Later, I look into the .stversions folder. And I realize that my data is literally right there. It’s all there.

So I made a copy of all my data in .stversions by doing an rsync from my phone to my computer in Termux. I think it took over 5 hours.

(In hindsight, I did not need to make a copy of 100GB of data, only the filenames. But hey, who knows what the future can hold.)

Anyways, looking at the filenames, I came up with this python program to change the name to remove the syncthing time marker:

import re

filenames = open("list").read()

result = re.sub(r"(.*)~\d{8}-\d{6}(.*)", r"\1\2", filenames, 0, re.MULTILINE)

if result:
    with open("list2", "w") as f:
        f.write(result)

(Code is under public domain/CC0-1.0 by the way)

The program gets file contents from a file called list, does some regex (which should only remove the last date and time part in the filename added by Syncthing) and writes it to list2.

Renaming the file was done with a shell script with a bunch of mv commands, not the python program.

After making the (hyperspecific to my setup and files) shell script, I put it on my phone and transferred it.

I then also had to restart my phone to fix Syncthing-fork not using the right private keys.

And my data is back, and Syncthing-fork is working.

Bye bye Syncthing (not fork)

After all that, I just deleted Syncthing (not the fork one) from my phone, because I already had another version (it was the fork version).

Post-mortem

So how did Syncthing just delete all my data? And keep it?

Well, here’s how it went:

A receive only folder was set to the entire Sync folder
I revert local changes
It deletes (and saves a copy into .stversions)
I realize the issue (trying to use my password manager)

I have no idea how one of my receive only folders was set to the entire Sync folder, but I guess it was in that state.

The Last Few Days of My Server

Jackson — Sun, 19 Nov 2023 13:34:09 +0000

In the last few days of my server, you may have noticed that it was down sometimes. This is because in the last few days, I did multiple things that kept leading to the downtime of my server. Oops.

The initial issue

On 2023-11-16, I was woken up by some emergency monitoring alerts notifications at 04:00Z. I look at it, and the services I host was not available (website, subdomains stuff, etc.).

Because it was 04:00Z at the time, I just decided to put the investigation off later.

Later at about almost 07:00Z, I dig around and check what’s available:

Onion services
I2P eep sites
Yggdrasil network stuff

Based on what was available, I concluded that clearnet services (so going to https://jacksonchen666.com and etc.) were broken.

The issue I found was that ddclient was just… not working for some unknown reason.

ddclient is used to change the IP address for my domains when my public IP for the server changes. However, it just decided to remove the IP address from my domains, resulting in the issue and downtime.

I fixed it by getting my server to do the IP update again. Probably by running ddclient.

(Status post issue)

More issues arise

So, after that was resolved, we had… the same issue again!

I have no idea why the issue keeps reoccurring. I just tried to do the same thing again, running ddclient and… it immediately exiting. I just hoped it worked.

I continue trying to figure out the issue at around 08:00Z. I still have no idea.

I login to my deSEC account (on 2023-11-16 08:00Z) and look around. Then I get hit with a 24 hour rate limit (on 2023-11-16 08:20Z).

Yes, deSEC has rate limits that apply for 24 hours. In deSEC’s documentation for rate limits, there is a rate limit name called user and it gives 2000 requests per day (or 24 hours).

I have somehow exceeded that rate limit. I also did nothing to exceed that rate limit.

It seemed weird, but my server was luckily online with the web, so I just decided to wait out the rate limit.

My deSEC account, the next day

After waiting for 24 hours (now 2023-11-17 09:30Z), I decided to try to login again. And I am still under a 24 hour rate limit.

At this point, I decided to email deSEC support for this issue, asking why I still have a 24 hour rate limit.

I then receive an email back saying:

Devices with the IP address [laptop-server IPv6 address 1] and [laptop-server IPv6 address 2] are doing thousands of dynDNS requests, sometimes several times per second.

(That’s a partial quote, and I have replaced the IPv6 addresses with the actual device. It was originally an actual IPv6 address)

I checked the server and… oh…

The server has multiple instances of ddclient, all of them quickly retrying (within less than 60 seconds), all trying to do dynDNS requests, then causing my account to be rate limited because of all those requests adding up.

Turns out, ddclient defaults to being a daemon, not being foreground. I assumed that when I ran ddclient, it would just stick output to my terminal. Nope, it just went on its own.

So here’s how I got into the rate limit again:

I ran ddclient without the foreground flag
Multiple instances of ddclient were just spamming the deSEC API
That cause it to hit the worst-case rate limits
Repeat

So I just went ahead and stopped the ddclient service, and killed all ddclient processes.

The problem should be solved, so I went ahead and asked deSEC support:

Can the rate limit on my account be removed earlier?

They responded:

That’s not simple on our side, so we would only do it in extraordinary circumstances.

So I decided to come back after the rate limit again.

Even more issues

I earlier stopped all ddclient processes. However, on 2023-11-18 11:40Z, my server IP address actually changed and there was nothing to fix that issue automatically (not even ddclient because I stopped it).

So of course, when I noticed the issue (at around 2023-11-18 14:10Z), I went ahead and just started ddclient again. And it’s fixed.

(Status post issue)

Hopefully no more issues

Now the server should be back to normal, and everything should be back to normal. Things should be running just fine as before the initial issue.

Upon reviewing the past few days and the mistakes I made, I’m not sure if I can say that I am without issues (there’s another blog post about another mistake coming soon(tm)).

Recap

Server removed IP address
Alerts at 2023-11-16 04:00Z
I fixed the issue
I was rate limited on deSEC (2023-11-16 08:20Z)
Next day, I try to login to deSEC, still rate limited (2023-11-17 09:30Z)
I found the issue for the second rate limit, multiple instances of ddclient
I wait for the rate limit
The server IP changes, and ddclient isn’t running (2023-11-18 11:40Z)
I fix that issue (2023-11-18 14:10Z)

Hey, I'm trying to juggle a million things. Be patient, OK?

Jackson — Tue, 14 Nov 2023 16:46:27 +0000

You may have noticed that it has been 6 days since my last blog post. The thing is, I have like a million giant things to juggle:

Drew DeVault captions (I’m updating all current captions)
Reinformation wiki (it isn’t online yet, but there’s still work to be done)
My servers (maintaining them)
My life (maintaining it)
Switching to SSH for git commit signing?

That’s at least some of the things I’m currently dealing with, maybe not today and maybe yesterday.

So that’s why I haven’t been doing blog posts. Because I’ve been doing everything else now, so there’s not much I can do about blog posts. Plus I don’t have ideas.

Tridactyl with iTerm2

Jackson — Wed, 08 Nov 2023 22:21:11 +0000

NOTE: This post contains multiple licenses in use. The different license use is in the script section.

Heads up: This is a quick infodump near midnight. Expect severe incoherence.

Tridactyl is basically vim keys in Firefox. iTerm2 is a terminal emulator.

Tridactyl supports a special feature: you can use your own editor in the terminal, something like vim.

I wanted to do something like that, but for some reason, I just couldn’t figure it out. Until today.

Today, I have a script and lots more information to bring to you to the table. More specifically, how to setup Tridactyl with iTerm2.

Complications with iTerm2

iTerm2, from what I can tell, is a macOS application. It does not behave like on Linux, where you can run the program to start a terminal, then the program exits when the terminal exits.

This presented an interesting problem to solve: How do you make iTerm2 behave like a program which would open the terminal, and exits when the terminal is closed?

Well, the answer is I made a Python script using the Python API for iTerm2. That’s a thing, and you have to enable it too.

Script for that thing

Setting up Tridactyl

Steps:

Install
Setup native messaging
Setup editor command
???
Profit

Installing Tridactyl

Easy. Go to the Firefox add-ons store and add to Firefox.

Setup native messaging

Setting up Tridactyl native messaging was easy:

Type :installnative in the browser
Paste the curl | sh command into your terminal
Let it do its thing
Type :native in the browser to check if it works

Of course, running with a Firefox based browser that is actually not Firefox (it’s LibreWolf) and uses a different path for native messaging and everything.

So I had to replace step 3 with “Let it do its thing, then copy the files over to the approprivate directory”.

I ran something like this:

cp ~/Library/Application\ Support/Mozilla/NativeMessagingHosts/tridactyl.json ~/Library/Application\ Support/LibreWolf/NativeMessagingHosts/tridactyl.json

Setup editor

I had to set it to something like this:

:set editorcmd /opt/homebrew/bin/bash -c "'/Users/jackson/Library/Application Support/iTerm2/iterm2env/versions/3.10.4/bin/python3' '/Users/jackson/Library/Application Support/iTerm2/Scripts/neovim_in_new_window.py' %f %l %c"

Replace /Users/jackson/Library/Application Support/iTerm2/iterm2env/versions/3.10.4/bin/python3 with whatever is correct by first checking /Users/jackson/Library/Application Support/iTerm2/iterm2env/versions/ for folders, then finding the python3 executable.

Because of some weirdness in macOS, you must use absolute paths for everything.

Debugging `:editor`

If running :editor gives you something like TypeError: e2 is null, it probably means you setup the editorcmd incorrectly.

Run the following in your browser’s Tridactyl:

:set logging.native debug

Then check the console for the current webpage you’re on, then run :editor again.

When finished debugging, run the following to revert debugging config:

:unset logging.native

About the script

I had to read a lot of the iTerm2 Python API documentation for this…

So, you cannot simply launch the iTerm2 binary and expect terminal windows to show up.

So I had to make a script that kinda emulates the terminal emulators on Linux, where you can run the program, a terminal shows up, then the program exits on terminal closing. That’s what the script does basically.

It also takes character numbers and moves the cursor to the right place.

The Script

NOTE: This code is licensed under the MIT license. The rest has its own license on the licensing page.

Heads up: This code is hardcoded to my configuration. Replace all absolute paths as necessary, and replace “LibreWolf” with “Firefox” if you use Firefox instead.

I put this file at ~/Library/Application Support/iTerm2/Scripts/neovim_in_new_window.py. It also needs the Python API to be enabled in iTerm2

#!/usr/bin/env python3.10
# vim: tw=0

# tridactyl :editor but with iterm2 (which doesn't have the right behavior so
# this is basically a wrapper script for that)
# https://github.com/tridactyl/tridactyl/issues/4697
# https://github.com/tridactyl/tridactyl/issues/684

# MIT License
# 
# Copyright (c) 2023 JacksonChen666
# 
# Permission is hereby granted, free of charge, to any person obtaining a copy
# of this software and associated documentation files (the "Software"), to deal
# in the Software without restriction, including without limitation the rights
# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
# copies of the Software, and to permit persons to whom the Software is
# furnished to do so, subject to the following conditions:
# 
# The above copyright notice and this permission notice shall be included in all
# copies or substantial portions of the Software.
# 
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
# SOFTWARE.

import iterm2
import sys
import subprocess
# This script was created with the "basic" environment which does not support adding dependencies
# with pip.

async def main(connection):
    # Your code goes here. Here's a bit of example code that adds a tab to the current window:
    app = await iterm2.async_get_app(connection)
    #filename, line, character = "", 0, 0
    filename, line, character = sys.argv[1:4]

    #if len(sys.argv) >= 2:
    #    filename = sys.argv[1]

    #if len(sys.argv) >= 3:
    #    line = sys.argv[2]

    #if len(sys.argv) >= 4:
    #    character = sys.argv[3]

    # defaults already assigned earlier
    command = f"/opt/homebrew/bin/bash -l -i -c \"/opt/homebrew/bin/nvim {filename} '+normal!{line}Gzv{character}|'\""
    print("#", command)

    new_window = await iterm2.Window.async_create(connection, command=command)
    print("#", new_window)

    await app.async_activate(raise_all_windows=False, ignoring_other_apps=True)
    await new_window.async_activate()

    #print(new_window.current_tab)
    async with iterm2.FocusMonitor(connection) as monitor:
        while True:
            # TODO: make it trigger if it's the last window and it closes
            await monitor.async_get_next_update()
            if new_window not in app.windows:
                break
    print("# end")
    # focus web browser
    subprocess.run(['osascript', '-e' 'tell application "LibreWolf" to activate'])

iterm2.run_until_complete(main)

Does it work?

Yes!

OBS Multi Track Audio Setup

Jackson — Mon, 06 Nov 2023 20:58:20 +0000

NOTE: There is a guide on the OBS project website which is basically this blog post. You can use that if you want.

Introduction

In OBS Studio, you can have up to 6 audio tracks, all of which can be recorded into the recorded video file.

With separate audio tracks, you can have things like your computer sounds, microphone sounds, and music, all on a separate audio track, allowing you to individually mess with each audio track post-production.

Now, how do you do audio tracks?

Advanced Audio Properties

Look at your Audio Mixer dock.

It should have some audio outputs, having your computer and microphone audio (or maybe something else). To make them separate into audio tracks, click on the 3 dots menu and click “Advanced Audio Properties”.

Now you should see a new window for the Advanced Audio Properties:

Next thing is to manage the audio tracks each audio thing outputs to.

Selecting audio tracks for audio outputs

Tip: Video players (and video services) default to using the first audio track¹, so use the first audio track for all audio you want to have by default (computer, mic, and music audio combined), and use the rest for whatever you want (unfiltered mic, only computer audio, etc.).

In my case, I want to include a filtered microphone output, and my computer audio in my video for viewers. I do not want to include an unfiltered microphone output to the viewers by default.

That means I would have to include my computer and filtered microphone into Audio Track 1. The rest of the 5 tracks I can just use for whatever, like a dedicated audio track for computer sounds, filtered microphone, and unfiltered microphone (for better post-production quality, if there even is any kind of post-production).

So for me, it looks something like this:

My computer sounds and microphone output to Audio Track 1, along with 3 dedicated audio tracks for each output. The unfiltered microphone (named “Mic/Aux Unfiltered” in the picture) does not output to Audio Track 1, which is quite important, because I do not want it included in the default audio track.

Your actual setup may differ from mine. That’s fine, as long as it works for you.

Video output settings

After doing the checkboxes, you should also do the same for checkboxes in your video output settings. Select the audio tracks you want to save, which should be the audio tracks you used and selected in the Advanced Audio Properties.

In Simple mode for Output mode, you can tick boxes.

In Advanced mode for Output mode, you can tick boxes in the Recording section, and name the audio tracks in the Audio section.

Selecting audio track for streaming

(Applicable to advanced output mode only)

For streaming, you can only choose 1 audio track to stream.

If you followed the earlier tip, you can just use audio track 1. If you didn’t follow the earlier tip, determine or create an audio track where all the audio you want is combined into 1 track. Then select that track.

Recap

OBS Audio Tracks can be helpful to separate different audio outputs from one another. What you do is tick boxes in Advanced Audio Properties and Audio Tracks in your output.

Well, from at least what I can tell. The default track seems to be the first track, but that could be different for different files. But using the first audio track has worked for me. ↩︎

The Shift Key

Jackson — Fri, 27 Oct 2023 21:41:43 +0000

The shift key. You can use it to make characters UPPERCASE. or you could not use it and be lowercase.

Physical shift key like behaviors

Here’s something you might not know: You can hold the shift key on your smartphone. Yes, hold it, and type in uppercase until you release it. It acts like a physical shift key when held down.

However by default, the shift key on smartphone is sticky for usability reasons. Sticky as in you press it and it stays on for the next character. If you double tap the shift key though, it will enable basically the equivalent of caps lock.

(You can have sticky keys on your computers too. If you were on Windows and pressed the shift key 5 times in a row, you will get annoyed by a prompt from Windows asking if you want to enable sticky keys.)

Sticky keys out of the way, here’s a quick recap:

You can use your shift key on your smartphone keyboard like a physical shift key
Shift keys are sticky by default on smartphone keyboards
Double tap the shift key to turn on caps lock

You’ve probably known the last 2 by now if you’ve used a smartphone keyboard. You probably didn’t know about the first point though.

Chinese smartphone keyboards

Now on Android phones made by Chinese phone manufactures. The software is… not great. Like most of it is not great, and also quite foreign in behaviors.

Two notable different behaviors in keyboards made by Chinese manufactures is:

You cannot hold the shift key down, it only reacts when you release on the shift key.
Caps lock by default, not single character shift by default.

These behaviors may vary between Chinese phone manufactures, but I’ve absolutely encountered that behaviors on what I call “Chinese phones”.

These behaviors may come from extreme time constraints, or just the developers were somehow not exposed to the behaviors on iPhones or other Android phones, or they had an “innovative” idea.

Fixing an Email Problem

Jackson — Tue, 24 Oct 2023 22:28:24 +0000

Something is missing

On 2023-10-20, I noticed that my server wasn’t sending me emails.

Usually, my server sends me sysadmin related emails like server updates and stuff.

But I realized for the last few days, my server has not sent me a single email ever since. Not an alert, not an update, not anything really.

I made that realization after my server has stopped emailing me for 10 days. Which to be honest is quite a bit, but they weren’t extremely important emails.

I tried to figure out what was going on, but for some reason I could not figure it out. For what I knew, the server was apparently trying to send some emails and it didn’t work. The emails were stuck in the queue, being retried and retried and retried.

More specifically, it seemed like it was sending to the wrong email server. It was sending to the receiving server, not the sending server.

I did not understand how this happened. I also did not know what was the source of the problem. Basically, I was completely clueless, unable to understand problem, and left without emails to receive from my server.

Desperately searching for a solution

I tried and tried, but found nothing.

I decided on last straw to ask my email provider (Migadu) why my emails were not sending.

I provided some logs, and went to play some osu!. Then when I checked my email, there it was! A response.

They came back asking for my IP addresses, after looking at the postfix logs and thinking that I might be IP banned.

I provided the IP addresses, and they came back with:

You’ve disabled sending to noreply address, so that [email] is getting rejected.

The other machine is sending to other addresses.

Oh.

So basically, imac was sending emails successfully. Those emails were some test emails that I had made to test that imac’s email was working.

On laptop-server, it was trying to send a “bounced email” notice to the sender, who was… noreply!

I have setup noreply to not accept any emails. If you try, it won’t work. The only thing the noreply email is allowed to do is send emails from any address on my domain.

So, now I disable zeyple on laptop-server and send a test email. And it works.

So I respond with:

Oh… I just realized that, it’s sending to the wrong addresses.

Sending the email to my receivable address gets me an email, so I guess it’s another issue with my setup somehow.

Thanks for your help though!

I have realized that I did not read mailq properly. Turns out the MAILER-DAEMON part was the sender and I thought the receiver part was actually the sender.

How this issue even happened

I have setup my postfix server to encrypt emails that are being sent. This was done with PGP. I used zeyple, which requires the gpg Python module.

Unfortunately, in doing some cleanup with the autoremove command in dnf, dnf has removed the gpg Python module. It was basically a cascading failure:

The zeyple program fails to run.
Postfix tries to send an email, but running the zeyple program means exit code non-zero.
Postfix tries to send an email to the sender, saying the email could not be sent.
The email addresses postfix was trying to send to would not accept emails, either because the email address didn’t exist or doesn’t accept emails.
Postfix now constantly retries to send the bounced email notification to nobody, and gets rejected on every time.

So if something went wrong with my server that required an email, I probably would not have heard it.

Why imac worked

Because it did not encrypt emails to me using zeyple. The problem was in the email encryption on laptop-server, due to a missing Python module. imac had none of that. So it had no issues.

Conclusion

So what did I learn?

Well:

I need to learn to read carefully.

I have ignored one important aspect of mailq: it includes the sender and the receiver, and I have ignored the sender because it did not look like an email address. It was “MAILER-DAEMON”.
This issue could’ve been entirely avoided if I had made a package for it with the proper dependencies. The root cause of this issue is because a python module got removed, so an important program could not be run anymore.
I should probably centralize my logging, but I’m not actually sure how to do that with zeyple.

And thanks to Dejan at Migadu for helping me get on the right track.

Starting at the Wrong Place

Jackson — Wed, 18 Oct 2023 19:50:24 +0000

I think there’s something that’s common between everyone that I’ve seen quite a bit in technology: Nobody reads. At least, people read, but they find the starting point too quickly, and land on the wrong starting point.

I’ve had times where I read something, then completely misunderstand it, proceed to continue on the basis of that misunderstanding, and then realize later that I’ve misunderstood the thing.

And I’ve thought of something: Can misreadings be attributed to incorrect/poor emphasis in the design of things?

Finding a kernel bug

Let’s say you run into some problem with the Linux kernel. You go and report it, and then you get told to “bisect the bug” and linked to the documentation about bisecting kernel bugs.

Now, you the reader, do as I instruct before you go do it:

Open that page
Casually find the first step in the process (no looking harder or something)
Come back to this page

After you done that, I have a question:

What were your eyes drawn to first?

If you were drawn to the git bisect start command, you started on the wrong step. And I won’t fault you for doing that.

There are multiple issues with that page:

The list of steps is an unordered list. This makes noticing “you’re on the wrong step” harder, and has definitely lead to confusion.
The first bullet point does not stand out, but the git bisect start command does stand out. That’s probably why your eyes was drawn to that first.

Finding the first line of my bio

Here’s my bio, roughly to how it is shown in Tusky:

jackson ☑️
@jacksonchen666
word, annoy, joke, question mark

serial bookmarker (18k), excessive screenshoter, data hoarder, computer user, self hosted server manager, minecraft accounts holder, mediocre circle clicker, fruit catcher, Phigros player…

What was your eye drawn to first?

(The rest of the bio is trimmed off, since that’s not the main topic.)

If you missed “word, annoy, joke, question mark”, well I wouldn’t blame it on you either. I would’ve missed that, and in fact, I did!

There’s big spacing after the display name, username, and the first line of my bio.

You might have quickly skipped to the second line of my bio (“serial bookmarker[…]”) as your eyes was drawn towards that when looking for the bio, but in fact, that was not where the bio started. It was grouped with the display name and username.

The problem here is separation. There was no immediately clear separation with the names stuff from the bio. No lines, nothing. Just suddenly, bio.

SourceHut

Here’s your goals:

Find the “Permalink” link
Find the “Summary” link
Find the owner of the git repository ("~jacksonchen666")
Find line 4 of the file
Do all of the above with just your eyes¹

If you’re ready, you can head to the .gitignore file for my website, do all of those, and try timing yourself (in your head is fine too).

After you done that, here’s my question: How fast/easy was that?

Did you have to look at the entire page for a specific thing, or did you hit the right thing very quickly?

If you’re not used to looking at sourcehut pages, you may had to take some more time to search the page. If you use sourcehut quite a bit though, you would know where you want to look for what you want, and you would do that with minimal error and quickly.

When emphasis on certain elements are considered, it can be easier to find things, or there wouldn’t be any issues with finding things.

Conclusion

I’ve explored some stuff where people can get confused on where the “starting point” is at. I also shown sourcehut and how it tries to not be too confusing, and easy to find the things you want.

If you only read the start of paragraphs of this blog post, then you might’ve missed a lot. That kind of misreading is not included, only “oh wait, I looked at the wrong thing” kind of thing.

Sorry if you’re vision impaired. I’m not sure of an alternative. ↩︎

100 Days to Offload Finally Complete

Jackson — Wed, 18 Oct 2023 17:56:33 +0000

This is the final update on the 100 days to offload challenge for 2023.

I’m on the list. I have completed the challenge.

That’s it. I have another blog post that I’m in the process of right now, expect that very soon and possibly even today.

Day 104

Jackson — Tue, 17 Oct 2023 17:33:31 +0000

This is part of the 100 days to offload challenge, also including other blog posts in 2023.

100 days to offload update

It is day 104 of the 100 days to offload challenge. I have overachieved by 4 days now.

I made some empty blog posts, but the extra 4 should’ve made up for it anyways.

That is all you should know on the current status of the 100 days to offload challenge.

Previous 100 days blog post

Website changes considerations

After getting rid of pretty much all tags in my posts, I’ve started to tag them a little more.

Previously, the tags would be a mess. I would have a single tag for a single blog post. No grouping, just tags. Like another way to view blog posts, although horribly.

This time though, I’m starting from scratch. It’s not yet available for browsing purposes, but hopefully soon.

Posts on jacksonchen666.com

Disclaimer: This Blog May Be Bad and Incorrect

QR Code Shenanigans

2 Years of Sysadmining

2 years of sysadmining

Recent downtime

A new wiki

Website dead URLs

Verifying Links in Mastodon Using More Mastodon

Verifying links in your profile fields

Verifying links in your profile fields without having the link visible

Verifying you control another Mastodon/Fedi account

More Notes on Replacing the Fairphone 4 Screen

The notes part

Testing for problems after replacement

Dealing with the old part

Monitor Resolutions and Sizes Review

21-inch 1080p vs. 27-inch 4k

My MacBook vs Fairphone 4

Recap

Who Am I?

osu!lazer and Health Drain

Important preface

Brief history of my skills in osu!

An osu!lazer update

Looking at my osu!catch stats

Playing easier maps, and getting way better

Conclusion

What it takes to setup a new domain name (for me)

The Logistical Mess of Trying to Process a Lot of Videos

The last time vs Now

Goals

Not enough time

An idea

Actual setup

File synchronization

If ran on LockedMac

Charging for stupid-laptop

Speedrunning an install on stupid-desktop

Networking for stupid-desktop

Storage and space management

Long-term solutions

Recovering From Syncthing Deleting My Data

Syncthing-fork

Data loss

Complaints sent

Data recovery

Bye bye Syncthing (not fork)

Post-mortem

The Last Few Days of My Server

The initial issue

More issues arise

My deSEC account, the next day

Even more issues

Hopefully no more issues

Recap

Hey, I'm trying to juggle a million things. Be patient, OK?

Tridactyl with iTerm2

Complications with iTerm2

Setting up Tridactyl

Installing Tridactyl

Setup native messaging

Setup editor

Debugging :editor

About the script

The Script

Does it work?

OBS Multi Track Audio Setup

Introduction

Advanced Audio Properties

Selecting audio tracks for audio outputs

Video output settings

Selecting audio track for streaming

Recap

The Shift Key

Physical shift key like behaviors

Chinese smartphone keyboards

Fixing an Email Problem

Something is missing

Desperately searching for a solution

If ran on `LockedMac`

Charging for `stupid-laptop`

Networking for `stupid-desktop`

Debugging `:editor`