Do all the things like ++ or -- rants, post your own rants, comment on others' rants and build your customized dev avatarSign Up
From the creators of devRant, Pipeless lets you power real-time personalized recommendations and activity feeds using a simple APILearn More
Search - "logs"
Manager: THE SERVER IS DOWN THE SERVER IS DOWN!!!!
Dev: Ok I’ll look into it
*5 mins later
Dev: Wow these are really strange logs, it’s like config values are being changed all over the place while I’m looking at it
Manager: Well I figured while you were looking into it I’d go i to the server settings and change everything I could find in order to try and get the server back up again. Two sets of hands are better than one, Is it up yet???
Manager: I THOUGT YOU SAID YOU’D LOOK INTO THIS. I NEED ANSWERS NOW. WHAT IS TAKING SO LONG?!?!?
Manager: Hey fullstackchris, the maps widget on our app stopped working recently...
Dev: (Skeptical, little did he know) Sigh... probably didn't raise quota or something stupid... Logs on to google cloud console to check it out...
Google Dashboard: Your bill.... $5,197 (!!!!!!) Payment method declined (you think?!)
Dev: 😱 WTF!?!?!! (Calls managers) Uh, we have HUGE problem, charges for $5000+ in our google account, did you guys remove the quota limits or not see any limit reached warnings!?
Managers: Uh, we didn't even know that an API could cost money, besides, we never check that email account!
Dev: 🤦♂️ yeah obviously you get charged, especially when there have literally been millions of requests. Anyway, the bigger question is where or how our key got leaked. Somewhat started hammering one of the google APIs with one of our keys (Proceeds to hunt for usages of said API key in the codebase)
Dev: (sweating 😰) did I expose an API key somewhere? Man, I hope it's not my fault...
Terminal: grep results in, CMS codebase!
Dev: ah, what do we have here, app.config, seems fine.... wait, why did they expose it to a PUBLIC endpoint?!
Long story short:
The previous consulting goons put our Angular CMS JSON config on a publicly accessible endpoint.
WITH A GOOGLE MAPS API KEY.
JUST CHILLING IN PLAINTEXT.
Though I'm relieved it wasn't my fault, my faith in humanity is still somewhat diminished. 🤷♂️
Oh, and it's only Monday. 😎
Hey, Root? How do you test your slow query ticket, again? I didn't bother reading the giant green "Testing notes:" box on the ticket. Yeah, could you explain it while I don't bother to listen and talk over you? Thanks.
Hey Root. I'm the DBA. Could you explain exactly what you're doing in this ticket, because i can't understand it. What are these new columns? Where is the new query? What are you doing? And why? Oh, the ticket? Yeah, I didn't bother to read it. There was too much text filled with things like implementation details, query optimization findings, overall benchmarking results, the purpose of the new columns, and i just couldn't care enough to read any of that. Yeah, I also don't know how to find the query it's running now. Yep, have complete access to the console and DB and query log. Still can't figure it out.
Hey Root. We pulled your urgent fix ticket from the release. You know, the one that SysOps and Data and even execs have been demanding? The one you finished three months ago? Yep, the problem is still taking down production every week or so, but we just can't verify that your fix is good enough. Even though the changes are pretty minimal, you've said it's 8x faster, and provided benchmark findings, we just ... don't know how to get the query it's running out of the code. or how check the query logs to find it. So. we just don't know if it's good enough.
Also, we goofed up when deploying and the testing database is gone, so now we can't test it since there are no records. Nevermind that you provided snippets to remedy exactly scenario in the ticket description you wrote three months ago.
Hey Root: Why did you take so long on this ticket? It has sat for so long now that someone else filed a ticket for it, with investigation findings. You know it's bringing down production, and it's kind of urgent. Maybe you should have prioritized it more, or written up better notes. You really need to communicate better. This is why we can't trust you to get things out.
"Hey, Root, someone screwed up and now all of our prod servers are running this useless query constantly. I know I already changed your priorities six times in the past three weeks, but: Go fix it! This is higher priority! We already took some guesses at how and supplied the necessary code changes in the ticket, so this shouldn't take you long. Remember, HIGH PRIORITY!"
1. I have no idea how to reproduce it.
2. They have no idea how to reproduce it.
3. The server log doesn't include queries.
4. The application log doesn't include queries.
5. The tooling intercepts and strips out some log entries the legendary devs considered useless. (Tangent: It also now requires a tool to read the logs because log entries are now long json blobs instead of plain text.)
6. The codebase uses different loggers like everywhere, uses a custom logger by default, and often overwrites that custom logger with the default logger some levels in. gg
7. The fixes shown in the ticket are pretty lame. (I've fixed these already, and added one they missed.)
8. I'm sick and tired and burned out and just can't bring myself to care. I'm only doing this so i don't get fired.
9. Why not have the person who screwed this up fix it? Did they quit? I mean, I wouldn't blame them.
Why must everything this company does be so infuriatingly complicated?11
My last wk93 story, the time we discovered school faculty was spying on students and we uncovered student's deepest secrets.
I call it, kiddiegate.
So if you've read my past rants you've noticed I did some pretty childish and reckless stuff with my highschool's systems when I was younger, but nothing compares to this thing.
After resetting the sysadmin account pwd on some machines it occurred to me I could write a keylogger to capture teachers Moodle accounts and so on, I decided to try it out on a regular lab computer first.
Imagine my surprise when I found a hidden keylogger already installed! I couldn't believe it but then I thought, what if other PC's have it? So I recruited my mates and teached them the process to check if a PC had been infected...ALL PCs were, over 30 computers we checked had been logging for over 3 months! That damn sysadmin! >:[
We were shocked and angry, but then I thought "hey. . . My work has been done for me, better take advantage"
So we did, we extracted each log and then removed it from the PCs along with the keyloggers. There were hundreds of records and then one day we started snooping into the fb accounts of some students (we shouldn't have) we uncovered so many nasty, shocking secrets...
One of the school's lady's man had a drunk one nighter with one of our gay friends, the most secluded and shy guy was sexting like crazy with 15 chicks at the same time, things like that...we promised to never say a word and deleted the logs.
After that we didn't do much and continued highschool as every teenage minor should, getting drunk and avoiding responsibilities, though we could never see many of our classmates the same way. The sysadmin was fired shortly after I graduated, no reason was stablished.
I want to clear out we were minors and laws in my country weren't clearly stablished at the time plus no harm was ever done. I don't condone hacking or any kind of illegal activity, just thought I'd share.6
Root: Fleshes out missing data in some factories. Tests affected code and finds the change breaks some specs (but shouldn’t).
Root: Reaches out to spec author.
Root: Messages thundercunt (the ticket’s code reviewer) on slack about the specs and the reaching out. No response.
Root: Works on another ticket while blocked.
Root: Logs off.
Root: Talks with spec author chick in the morning. Decide to pair on specs later.
TC: Still no slack response.
Root: Gives update in standup. Mentions factories and broken specs. Mentions pairing with spec chick.
TC: Still no slack response.
Root: Pulled off tickets in favor of prod issue. Gets ignored by everyone else diagnosing prod issue. Investigates prod issue by herself. Discovers prod issue isn’t from bad code, but bad requirements — code works as requested. Communicates this with details. Gets ignored by people still diagnosing prod issue. Tries again. Gets ignored. Gives up. Works on non-blocked tickets instead.
TC: Still no slack response.
TC: Comments on PR telling me I broke specs (how did I not notice?), that I need to reach out to spec chick and work with her, and that I can’t resolve the ticket until it’s fixed and passes code review.
TC: Still no slack response. (21 hours later at this point)
TC: Logs off. Still no response (25 hours at this point)
Ignoring the prod issue for the moment…
I broke specs. No shit.
I need to talk with spec chick. No shit.
I can’t resolve the ticket. No shit!
Bitch, I told you all of this 21 fucking hours prior, and again 3 hours prior during standup. But no, I clearly “don’t communicate” and obviously have no bloody clue what I’m doing, either, so I need everything spelled out for me.
And no, I didn’t resolve the fucking ticket. Why the fuck would I if it still has pending changes? Do you even check? Ugh!
And what the fuck with that prod issue? I’m literally giving you the answer. fucking listen! Stupid cunts.
Why is it all of the women I work with are useless or freaking awful people? Don’t get me wrong, many of the men are, too, but I swear it’s every single one of the women. (Am I awful, too?)
I can’t wait to leave this sewer of a company.
Oddly still a good day, though. Probably because I talked to recruiters and sent out my resume again.16
micromanager: "Quick and easy win! Please have this done in 2-3 days to start repairing your reputation"
ticket: "Scrap this gem, and implement your own external service wrapper using the new and vastly different Slack API!"
slack: "New API? Give me bearer tokens! Don't use that legacy url crap, wth"
prev dev: "Yeah idk what a bearer token is. Have the same url instead, and try writing it down so you don't forget it?"
Slack admin: "I can't give you access to the slack integration test app, even though it's for exactly this and three others have access already, including your (micro)manager."
Slack: "You can also <a>create a new slack app</a>!" -- link logs me into slack chat instead. After searching and finding a link elsewhere: doesn't let me.
Slack admin: "You want a new test slack app instead? Sure, build it the same as before so it isn't abuseable. No? Okay, plan a presentation for it and bring security along for a meeting on Friday and I'll think about it. I'm in some planning meetings until then."
This job is endless delays, plus getting yelled at over the endless delays.
At least I can start on the code while I wait. Can't test anything for at least a week, though. =/18
A little over a year ago I was in the hiring process with a new company and countered their initial offer. I was told by the CTO that it was no problem and they would get back to me soon.
A couple days go by and I'm then informed that they're hiring a new IT director and would like me to interview with him as well. It felt kinda lame since I'd already been offered the job but I rolled with it.
When I showed up to the office for an interview I tried to call and let them know I was there and couldn't get a hold of anyone. 30 minutes later I get a call from the CTO saying they couldn't find the new IT director and when they got him to answer the phone he said he had left early and would call me to do a phone interview.
Obviously the whole experience so far has been pretty lame but I stuck with it because I knew the CTO personally. I did the phone interview and quickly realized this dude was a prick, and would be a terrible boss, but I spoke with the CTO again who told me to stick with it and eventually I did get the job.
Fast forward about a month and it's clear the new director is trash. He literally bragged about firing a dude over an accidental outage (wtf!?).
He had the technical experience you'd expect of a junior help desk and his management skills were pretty clearly sub-par.
He was also, for whatever reason, completely unable to communicate with the only woman on our team. When assigning work he would always feel the need to ask if she could 'handle it' rather than just assigning it to her like it's done for everyone else. He was pretty clearly sexist.
The whole team hates this dude by this point but he's somehow managed to woo the executives into thinking he shits gold.
I was helping him set up a Python venv on his machine when I noticed another VPN client installed which certainly piqued my interest. After a bit of digging it was clear he was using company time and company equipment to continue working for his previous employer.
We turned over logs and he was fired the next day. He tried to add me on LinkedIn afterwards and I have never declined something quicker.
Moral of the story is don't be a dickhead.1
So apparently due to an extremely talkative x input driver and an error in a certain app, I've been running an emergent keylogger on my computer for half a year. On every keypress event, the driver would call the app, the app would segfault, the driver would log the incident including the event to /var/log and then crash, and the app would restart the worker. I noticed this when I started wondering why /var/log is over 100GB in size.13
Worst hack/attack I had to deal with?
Worst, or funniest. A partnership with a Canadian company got turned upside down and our company decided to 'part ways' by simply not returning his phone calls/emails, etc. A big 'jerk move' IMO, but all I was responsible for was a web portal into our system (submitting orders, inventory, etc).
After the separation, I removed the login permissions, but the ex-partner system was set up to 'ping' our site for various updates and we were logging the failed login attempts, maybe 5 a day or so. Our network admin got tired of seeing that error in his logs and reached out to the VP (responsible for the 'break up') and requested he tell the partner their system is still trying to login and stop it. Couple of days later, we were getting random 300, 500, 1000 failed login attempts (causing automated emails to notify that there was a problem). The partner knew that we were likely getting alerted, and kept up the barage. When alerts get high enough, they are sent to the IT-VP, which gets a whole bunch of people involved.
VP-Marketing: "Why are you allowing them into our system?! Cut them off, NOW!"
Me: "I'm not letting them in, I'm stopping them, hence the login error."
VP-Marketing: "That jackass said he will keep trying to get into our system unless we pay him $10,000. Just turn those machines off!"
VP-IT : "We can't. They serve our other international partners."
<slams hand on table>
VP-Marketing: "I don't fucking believe this! How the fuck did you let this happen!?"
VP-IT: "Yes, you shouldn't have allowed the partner into our system to begin with. What are you going to do to fix this situation?"
Me: "Um, we've been testing for months already went live some time ago. I didn't know you defaulted on the contract until last week. 'Jake' is likely running a script. He'll get bored of doing that and in a couple of weeks, he'll stop. I say lets ignore him. This really a network problem, not a coding problem."
IT-MGR: "Now..now...lets not make excuses and point fingers. It's time to fix your code."
IT-VP: "I agree. We're not going to let anyone blackmail us. Make it happen."
So I figure out the partner's IP address, and hard-code the value in my service so it doesn't log the login failure (if IP = '10.50.etc and so on' major hack job). That worked for a couple of days, then (I suspect) the ISP re-assigned a new IP and the errors started up again.
After a few angry emails from the 'powers-that-be', our network admin stops by my desk.
D: "Dude, I'm sorry, I've been so busy. I just heard and I wished they had told me what was going on. I'm going to block his entire domain and send a request to the ISP to shut him down. This was my problem to fix, you should have never been involved."
After 'D' worked his mojo, the errors stopped.
Month later, 'D' gave me an update. He was still logging the traffic from the partner's system (the ISP wanted extensive logs to prove the customer was abusing their service) and like magic one day, it all stopped. ~2 weeks after the 'break up'.8
Me passing time on the weekend
Random call from unknown number
Turns out it's the manager
M: hey , how is your weekend going ...
Me: nothing much ... Whatsup ?
M : yeah well , we wanted to push some minor adhoc fixes as some clients wanted it urgently
The Devops folks need developer support . Can you pitch in and monitor
Me : I'm not aware of what changes are going , i don't think i can provide support
M : don't worry it's minor changes , it's already tested in pre prod , you just need to be on call for 30 mins
Me : ugh okay .. guess 1 hr won't hurt
M: thanks 👍🏽
Me: *logs in
*Notices the last merged PR
+ 400 lines , implemented by junior dev and merged by manager
*Wait , how is this a *minor* release...
*Release got triggered already and the CI CD pipeline is in progress
*5 mins later
*Pipeline fails , devops sends email - test coverage below 50%
Manager immediately pitches in ...
M: hey , i see test coverage is down , can you increase it ?
Me: and how do u suppose I do that ?
M : well it's simple just write UTC for the missing lines ... Will it take time ?
Me : * ah shit here we go again
Yeah it will take time , there are around 400 lines , I am not aware of this component all together
Can you ask junior dev to pitch in and write the UTC for this
*Actually junior dev is out on a vacation with his girlfriend
M : well he's out for the weekend , but
as a senior dev , i expect you to have holistic understanding of the codebase and not give excuses ,
this is a priority fix which client are demanding we need this released ASAP
Me : * wait wat ?
I ended up being online for next 3 hours figuring out the code change and bumping up the UTC 🤦🏾9
I was on vacation when my employer’s new fiscal year started. My manager let me take vacation because it’s not like anything critical was going to happen. Well, joke was on us because we didn’t foresee the stupidity of others…
I had to update a few product codes in the website’s web config and deploy those changes. I was only going to be logged in for 30 minutes to complete that.
I get messaged by one of our database admins. He was doing testing and was unable to complete a payment on the website. That was strange. There was a change pushed by our offsite dev agency, but that was all frontend changes (just updating text) and wouldn’t affect payments.
We don’t want to enlist the dev agency for debugging work, especially when it’s not likely that it’s a code issue. But I was on vacation and I couldn’t stay online past the time I had budgeted for. So my employer enlists the dev agency for help. It’s going to be costly because the agency is in Lithuania, it was past their business hours, and it was emergency support.
Dev agency looks at error logs. There are Apple Pay errors, but that doesn’t explain why non Apple Pay transactions aren’t going through. They roll back my deployment and theirs, but no change. They tell my employer to contact our payment processor.
My manager and the Product Manager contact Payroll, who is the stakeholder for our payment gateways. Payroll contacts our payment gateway and finds out a service called Decision Manager was recently configured for our account. Decision Manager was declining all payments. Payroll was not the person who had Decision Manager installed and our account using this service was news to her.
Payroll works with our payment processor to get payments working again. The damage is pretty severe. Online payments were down for at least 12 hours. Our call center had logged reports from customers the night before.
At our post mortem, we had to find out who ok’d Decision Manager without telling anyone. Luckily, it was quick work. The first stakeholder up was for the Fundraising Dept. She said it wasn’t her or anyone on her team. Our VP of Analytics broke it to her that our payment processor gave us the name of the person who ok’d Decision Manager and it was someone on the Fundraising team. Fundraising then starts backtracking and says that oh yes she knew about it but transactions were still working after the Decision Manager had been configured. WTAF.
Everyone is dumbfounded by this. How could you make a big change to our payment processor and not tell anyone? How did our payment processor allow you to make this change when you’re not the account admin (you’re just a user)?
Our company head had to give an awkward speech about communication and how it’s important. The web team can’t figure out issues if you don’t tell us what you did. The company head was pissed because it was a shitty way to start off the new fiscal year. Our bill for the dev agency must have been over $1000 for debugging work that wasn’t helpful.
Amazingly, no one was fired.5
We've had a bunch of flaky tests in our repo for a while now that no one could be bothered to fix; we'd just re-run ci until it's green. Today I looked into it and I was inspired to make this meme, because I lived through it.
Adding logs to investigate just lowered the fail rate making investigation more difficult. I do have an idea of what it might be though so, we'll see tomorrow.3
I was pressued to shift the blame.
We received an angry email from a customer that some of their data had disappeared. The boss assigns me to this task. This feature is relatively new and we've found some bugs in the past in here. I go through request logs, search the database, run some diagnostics, etc. for about 5 hours and I cannot find the problem. I focus on the bugs that we've had before but they don't seem to be the problem.
I tell the boss "sorry but I checked XYZ and I can't find the problem. I'm out of ideas." But the boss wanted answers by the end of the day. They did not want to admit to the client that we couldn't figure out what's wrong.
By now I was more pressured to find an answer, find something or someone to blame it on, not exactly to find the real solution. So I made up some BS:
"Sometimes, in HTML forms, the number inputs allow you to change the number by scrolling. We have some long forms where the user has to scroll. Perhaps the focus remained on the number input, so when they scrolled down they accidentally changed the number they meant to input."
The boss was happy with that. We explained this to the customer, and there's now a ticket to change type="number" to type="text" in our HTML forms and to validate it in th backend.
A week later another customer shows us a different error. This one is more clear because it had a stack trace, but I realise that this error is what caused our last error. It was pretty obscure, mind you, the unit tests didn't detect it.
I didn't tell the boss that they were connected tho.
With two angry clients in two weeks, I finally convinced the boss to give us more time to write more unit tests with full coverage.
Today I found out when a user logs in in this piece of crap: 59 calls to the api just to get user permissions 😤
For the love of god, I spent 2,5 hours debugging why Minecraft from the windows store doesn't work...
The game just shows a red message telling you it didn't work.
I checked the logs, nothing just warnings
I re-installed the game, nothing, same error
Updated java and all parts of the store, nothing....
Obviously I had to install Something called the "xbox identity Provider"... You know... On a PC... For a distinctly PC game to work... Installed by the store... And the provider is also on the store... But it doesn't auto-install with the game
Ever since you migrated to the Microsoft Auth the login experience is awful (I ranted about that already)
How about you do the bare fucking Minimum of an User experience and Install the fucking dependencies when I re-installed something your fucking store??!!!
The fucking bare minimum that every package manager ever created fucking has as a basic requirement?! Are you kidding me?
Rename your fucking services so they make sense and please don't waste everyone's time by having both shitty logs and no dep management for your own apps... Fucks sake15
I left my previous company because my tech leadership was insensitive and agressive.
However, I am in a start-up right now and CTO is a nut job.
He creates random Slack threads and keeps messaging me like crazy. The co-founders have shut him down multiple times and yet his only success metric is "number of deliveries".
The other day something broke in production and teams were discussing about resolving the bug in one of the Slack channels.
CTO literally wrote this and I wish I was making this up, "let us not look at the logs and trust our code to work fine."
I was baffled and confused. I realised me leaving my previous organisation because of such tech leadership was a stupid decision.
Crows are black everywhere.5
Not the worst, but probably the only one I can sort of explain & not get into trouble for NDA breach..
Umm.. here it goes.. wrong id returned from db procedure, tried to do something on db with that id and got exception that the id doesn't exist. Instead of checking why the procedure returns nonexistent id, he just wrapped everything in try catch without any logs.. & of course, didn't tell anyone about this.. o.0
I know, I know, code review could have prevented this, but holy fuck..
Guy's cv had more experience than I have now, so at the time, I didn't think I'd have to check every line of code he wrote, especially not for shit like this.3
You know modern cars, they have these computer thingies that tell you when something isn't working with little warning lights.
How useful !
"Take me to repair shop!" it says, and even sets the SatNav route.
Of course, the place might be closed, but still, its trying to help. :-)
Anyhow, by chance just happen to be there getting said car serviced..
Mention the several warning thingies that sprang up on the way in..
Which took twice as long as a normal service, so I was hopeful they was fixing things !
Though every time I go and ask how things are, magically its just been finished and I haven't been waiting for no good reason because no one remembered I was waiting..
No, they didn't fix any of the faults...
Why I asked without getting angry..
Because the diagnostic computer said there wasn't any..
But there was !
Come back when the fault returns they said..
If the fault disappears before their computer gets plugged in, they will just say there isn't a fault..
Apparently on the car there is no fault logging, its either, a fault right now, or no fault at all..
This might explain why a few months ago all the brakes seized up ( Its less than 2 years old, it shouldn't do that ! ), if some computer part is playing up..
So, I'll get my own car diagnostic computer and wait for it to play up, and maybe get some more error codes/etc. to pass on to the car fixing place !
Today's lesson, logs are important !
Also, just because a computer says there isn't something wrong with something, doesn't mean there isn't, so go and check it physically !
And, the customer is always right !
Previously had an issue with a part that had worn out, asked for it to be replaced.
Went to pick up the car, asked if the part had been replaced.
No it hadn't !
They thought it wasn't worn out !
I asked, did they look at it ?
No they didn't was their reply..
I told them, if you take it off, you can see its worn out.
I watched them take it off, ( After much struggling, to which I remarked that yes, when I took it off to look at it, I had similar trouble ! ) they then saw it was worn out and put a new one on !
They then struggled to put the new one on, which I also mentioned I had the same trouble.
Being as it was my first time taking off one of those parts, you could be forgiven to think I was just a beginner.
But you might think a professional would be able to do a better job..
You just can't get the staff these days !14
At the institute I did my PhD everyone had to take some role apart from research to keep the infrastructure running. My part was admin for the Linux workstations and supporting the admin of the calculation cluster we had (about 11 machines with 8 cores each... hot shit at the time).
At some point the university had some euros of budget left that had to be spent so the institute decided to buy a shiny new NAS system for the cluster.
I wasn't really involved with the stuff, I was just the replacement admin so everything was handled by the main admin.
A few months on and the cluster starts behaving ... weird. Huge CPU loads, lots of network traffic. No one really knows what's going on. At some point I discover a process on one of the compute nodes that apparently receives commands from an IRC server in the UK... OK code red, we've been hacked.
First thing we needed to find out was how they had broken in, so we looked at the logs of the compute nodes. There was nothing obvious, but the fact that each compute node had its own public IP address and was reachable from all over the world certainly didn't help.
A few hours of poking around not really knowing what I'm looking for, I resort to a TCPDUMP to find whether there is any actor on the network that I might have overlooked. And indeed I found an IP adress that I couldn't match with any of the machines.
Long story short: It was the new NAS box. Our main admin didn't care about the new box, because it was set up by an external company. The guy from the external company didn't care, because he thought he was working on a compute cluster that is sealed off behind some uber-restrictive firewall.
So our shiny new NAS system, filled to the brink with confidential research data, (and also as it turns out a lot of login credentials) was sitting there with its quaint little default config and a DHCP-assigned public IP adress, waiting for the next best rookie hacker to try U:admin/P:admin to take it over.
Looking back this could have gotten a lot worse and we were extremely lucky that these guys either didn't know what they had there or didn't care.
I used to think that I had matured. That I should stop letting my emotions get the better of me. Turns out there's only so much one can bottle up before it snaps.
Allow me to introduce you folks to this wonderful piece of software: PaddleOCR (https://github.com/PaddlePaddle/...). At this time I'll gladly take any free OCR library that isn't Tesseract. I saw the thing, thought: "Heh. 3 lines quick start. Cool.", and the accuracy is decent. I thought it was a treasure trove that I could shill to other people. That was before I found out how shit of a package it is.
First test, I found out that logging is enabled by default. Sure, logging is good. But I was already rocking my own logger, and I wanted it to shut the fuck up about its log because it was noise to the stuffs I actually wanted to log. Could not intercept its logging events, and somehow just importing it set the global logging level from INFO to DEBUG. Maybe it's Python's quirk, who knows. Check the source code, ah, the constructors gaves `show_log` arg to control logging. The fuck? Why? Why not let the user opt into your logs? Why is the logging on by default?
But sure, it's just logging. Surely, no big deal. SURELY, it's got decent documentation that is easily searchable. Oh, oh sweet summer child, there ain't. Docs are just some loosely bundled together Markdowns chucked into /doc. Hey, docs at least. Surely, surely there's something somewhere about all the args to the OCRer constructor somewhere. NOPE! Turns out, all the args, you gotta reference its `--help` switch on the command line. And like all "good" software from academia, unless you're part of academia, it's obtuse as fuck. Fine, fuck it, back to /doc, and it took me 10 minutes of rummaging to find the correct Markdown file that describes the params. And good-fucking-luck to you trying to translate all them command line args into Python constructor params.
"But PTH, you're overreacting!". No, fuck you, I'm not. Guess whose code broke today because of a 4th number version bump. Yes, you are reading correctly: My code broke, because of a 4th number version bump, from 220.127.116.11, to 18.104.22.168, introducing a breaking change. Why? Because apparently, upstream decided to nest the OCR result in another layer. Fuck knows why. They did change the doc. Guess what they didn't do. PROVIDING, A DAMN, RELEASE NOTE. Checked their repo, checked their tags, nothing marking any releases from the 3rd number. All releases goes straight to PyPI, quietly, silently, like a moron. And bless you if you tell me "Well you should have reviewed the docs". If you do that for your project, for all of your dependencies, my condolences.
Could I just fix it? Yes. Without ranting? Yes. But for fuck sake if you're writing software for a wide audience you're kinda expected to be even more sane in your software's structure and release conventions. Not this. And note: The people writing this, aren't random people without coding expertise. But man they feel like they are.6
Come on guys, use those JSON schemas properly. The number of times I see people going "err, few strings here, any other properties ok, no properties required, job done." Dahhh, that's pointless. Lock that bloody thing down as much as you possibly can.
I mean, the damn things can be used to fail fast whenever you misspell properties, miss required properties, format dates wrong - heck, even when you want to validate the set format of an array - and then libraries will throw back an error to your client (or logs if you're just on backend) and tell you *exactly what's wrong.* It's immensely powerful, and all you have to do is craft a decent schema to get it for free.
If I see one more person trying to validate their JSON manually in 500 lines of buggy code and throwing ambiguous error messages when it could have been trivially handled by a schema, I'm going to scream.18
Still having problems with samba doing weird shit. Now looking through the logs and
Hm... *that* doesn't seem right 🤔3
4 f*ing years since the app is published and f*ing QA team still can't read f*ing logs and find out why the f*ing backend messed up and f*ing blaming it on the frontend in a f*ing scenario that has nothing to do with the f*ing REST API. F**********CK!7
I dont understand the Log4j vulnerability.
Isnt the ability to execute code a feature they added so that you can add dynamic data to the logs?
If it is a feature then isnt it written in the documentation?
Is the problem that a lot of companies forgot to sanitize the input before logging it?23
In the 90s most people had touched grass, but few touched a computer.
In the 2090s most people will have touched a computer, but not grass.
But at least we'll have fully sentient dildos armed with laser guns to mildly stimulate our mandatory attached cyber-clits, or alternatively annihilate thought criminals.
In other news my prime generator has exhaustively been checked against, all primes from 5 to 1 million. I used miller-rabin with k=40 to confirm the results.
The set the generator creates is the join of the quasi-lucas carmichael numbers, the carmichael numbers, and the primes. So after I generated a number I just had to treat those numbers as 'pollutants' and filter them out, which was dead simple.
Whats left after filtering, is strictly the primes.
I also tested it randomly on 50-55 bit primes, and it always returned true, but that range hasn't been fully tested so far because it takes 9-12 seconds per number at that point.
I was expecting maybe a few failures by my generator. So what I did was I wrote a function, genMillerTest(), and all it does is take some number n, returns the next prime after it (using my functions nextPrime() and isPrime()), and then tests it against miller-rabin. If miller returns false, then I add the result to a list. And then I check *those* results by hand (because miller can occasionally return false positives, though I'm not familiar enough with the math to know how often).
Well, imagine my surprise when I had zero false positives.
Which means either my code is generating the same exact set as miller (under some very large value of n), or the chance of miller (at k=40 tests) returning a false positive is vanishingly small.
My next steps should be to parallelize the checking process, and set up my other desktop to run those tests continuously.
Concurrently I should work on figuring out why my slowest primality tests (theres six of them, though I think I can eliminate two) are so slow and if I can better estimate or derive a pattern that allows faster results by better initialization of the variables used by these tests.
I already wrote some cases to output which tests most frequently succeeded (if any of them pass, then the number isn't prime), and therefore could cut short the primality test of a number. I rewrote the function to put those tests in order from most likely to least likely.
I'm also thinking that there may be some clues for faster computation in other bases, or perhaps in binary, or inspecting the patterns of values in the natural logs of non-primes versus primes. Or even looking into the *execution* time of numbers that successfully pass as prime versus ones that don't. Theres a bevy of possible approaches.
The entire process for the first 1_000_000 numbers, ran 1621.28 seconds, or just shy of a tenth of a second per test but I'm sure thats biased toward the head of the list.
If theres any other approach or ideas I may be overlooking, I wouldn't know where to begin.16
The amount of energy spent to just write ‘Hi’ and click a send button is so big that we should consider banning of sending hi messages.
Instead of just saying “Hi!” we are now using analog to digital preprocessors that convert it to bunch of 0 and 1 to send it over communication layer and deliver it to other human being that will convert it from digital to analog by reading it but that is simple.
By sending message using phone we also:
- save it to local phone
- convert it to couple protocols
- transmit it over air so make connection to internet provider services that would generate logs on this provider as well as whole routing table before it gets to the target person
- save it on messaging provider disk
- probably be processed by filters by provider, sometimes be reviewed or listened by third parties and also processed in bulk by artificial intelligence algorithms
- finally delivered to target phone and saved there where that person would just change this text to their inner voice and save it
- sometimes encrypted and decrypted
- sometimes saved on provider
- sometimes saved on phone manufacturer cloud backup
- don’t get me started on people involved to keep this infrastructure in place for you just to say hi
There are also some indirect infinite possibilities of actions for example:
- emit sound and light that can lead to walking from one room to other
- the floor in your house is destroyed cause of it so you need to renovate your floor
- sound can expose your position and kill you if you’re hiding from attacker
- sound can wake you up so you wake up in different hours
- it can stop you from having sex or even lead to divorce as a result simple hi can destroy your life
- can get you fired
- can prevent from suicide and as a result you can make technology to destroy humans
and I can write about sound and light all day but that’s not the point, the point is that every invention makes life more complicated, maybe it saves time but does it really matter ?
I can say that every invention we made didn’t make world simpler. The world is growing with complexity instead.
It’s just because most of those inventions lead to computer that didn’t make our world simpler but made it more complicated.1
So I made an update to my React Native app. I changed UI of a couple of screen, added a few animations here and there, refactored how my graphQL resolvers work in the backend(no breaking changes), changed how data gets loaded into the database etc.
It worked in dev so I figured hey let's deploy it. Today is(was because it's now 3am but more on that later) a national holiday so no one goes to work so no one will use my app so I have an entire day to deploy.
I started at 15:00(because i woke up at 13:00 lol). I tested the update once again in dev and proceeded to deploy it to prod. I merged backend to master, built docker images, did migrations on the db, restarted docker-compose with new images. And now for the app. I run ./gradlew assembleRelease and it starts complaining that react-native-gesture-handler is not installed. Ugh, rm -rf node_modules && yarn install. It worked. But now gradlew crashes and logs don't tell me anything. Google tells me to change a bunch of gradle settings but none of them work. Fast forward 5h, it's around 20:00 and I isolated the issue to, again, react-native-gesture-handler. They updated from 2.2.4 to 2.3.0 which didn't fucking compile. 2 more hours passed (now 22:00) and I got v2.3.1 working which fixed the problem in 2.3.0 but made my app crash on startup. YOUR FUCKING LIBRARY GETS 250K WEEKLY DOWNLOADS AND YOU DONT EVEN BOTHER CHECKING IF IT COMPILES IN PROD ON ANDROID?! WHAT THE FUCK software-mansion?
After I solved that, my app didn't crash. Now it threw an error "Type errors: Network Request Failed" every time I fetch my legacy REST API(older parts use rest and newer use graphql. I'll refactor that in the next update). I'll spare you the debugging hell i went through but another 5h passed. Its 3am. My config had misspelled url to prod but good for dev... I hate myself and even more so react-native-gesture-handler.3
Someone please explain to me how error messages such as
"Something went wrong" or "Critical error" are valid and provide little to no follow up explanation in the GUI, Logs, or client logs.
I get that not all error cases can be displayed on a GUI, but at least have decent error handling. Especially if your $8+ billion company.1
Long time no see devRant. This rant is dedicated to an MQTT implementation we use. Mosquitto, mqtt.js - FUCK YOU.
I spent the last fucking 30+ hours trying to find why the bloody fuck the stupid server / client won't connect to the shitty mqtt broker. From changing all possible config, enabling & disabling specific code nothing abso-fucking-lutely works.
But then it will randomly decide to connect to the fucking broker, not causing any issues at all. And each fucking day when I wake up again and think to myself: oh today I can actually leave when it is still somewhat bright outside - NOPE. Because guess what? The fucking shitty abomination doesn't work anymore.
I just love these types of problems that are almost impossible to debug because the only logs you get is: "SERVER disconnected". It's impossible to get a proper reason out of this shit show, it's just turned into randomly guessing what the error could be (and especially where it could be).
And each time I got it to work, tested it and let the testing team know that they can start testing it will just stab me in the back and be like "fuck you, I'm not working any more". Luckily it's not like the deadline is next week... otherwise work is great, trust me.13
A large pool of application instances' is writing logs to the same physical file. No way to distinguish which instance wrote which line.
Welcome to hell
We're being asked questions. We're replying that we cannot help unless logging is fixed. Noone's bothering to fix this mess and instead returns tickets with requests to investigate more.
Retarded WSL crashed which means I lost all my fucking logs
How do people even work on this retarded OS? Is it just a scam for parasites to pretend they work and leeche on society using their precious little intellectual property bullshit12
A guy using our tool that automate rest api calls requests a feature to add the request body in the logs.
He was using “get” calls with no request body as a proof of missing feature.2
My day couldn't start in a worse day.
We are having a demo this week and I worked yesterday after hours to get the product ready. Tested everything and we were all set for the demo content.
Today we installed the new version with my fixes and nothing worked. Today's version should be the golden version to prep the demo! Obviously everyone starts looking at me as to why nothing works both worried and eager to help. But I got so stressed I just wanted to dig a hole.
Luckily after going through the logs a colleague of mine pieced together something he heard about another colleague on another location we have submitted a fix (without telling our location) that f**ked up the whole system.
Luckily we reverted it since the system was better without it and got it stable again but after all of that I had to go rest because until we found the answer I was starting to think I couldn't get one thing right. I think it was the most stressful moment in 5+ years on this job4
As a tech lead i sometimes find it very hard to defend developers for no fault of theirs.
Management is completely incapable of noticing hard data like git logs or action items updated on an excel and seems to have an idea that the devs are incompetent , but the ba that sets impossible goals and crap business documentation is competent.
Should i just let the project and juniors burn.3
Stakeholder: Users are unable to buy tickets on the website. IT says Azure’s health check is showing an unhealthy status.
[It’s Sunday. Web Engineering is not on call so no one sees this right away.]
Stakeholder: IT restarted the Azure website twice, but users still can’t place orders.
Me: There was never an issue with the Azure site. That health check is inaccurate. There is a rewrite rule that sends the Azure supplied domain to our custom domain. The Azure health check doesn’t like that so it returns an unhealthy status. The problem is the ticketing server that the website has to communicate with. The ticketing server is overwhelmed and can’t handle more requests. IT should have checked the ticketing server’s logs. This has happened before and it’s never been an Azure issue. It’s a ticketing server issue.
Stakeholder and IT: Oops 😅
JFC. Stop trying to make this web engineering’s problem. Stop trying to make it look like engineering dropped the ball. The ticketing server has experienced this issue multiple times. The ticketing server is maintained by a different team. The website’s symptoms are always the same and there are steps you need to take before you make the decision to restart the website, which will cause the website to show a blue screen of death that says 503 service unavailable for a few minutes. And we have a switch to shut off all transactions. Why do you not want to use it when it’s clear the website can’t process transactions???3
Alright boys.. calling in my networking friends for help..
Recently switched my ISP and got a fibre optic installed (100Mbps).
Thr ISP provided a new TP-Link router which supports 5GHz as well as 2.4GHz.
Some of my devices support 5GHz and connect to that network which works flawlessly.
However, my phone does not support 5GHz and hence, have to connect on 2.4GHz.
Somehow, the main router as well as the access point, are not functioning well for 2.4GHz. Whenever the connection is established, it would work fine for a minute or two before the networks starts disconnecting.
Restart the device Wi-Fi and it works for few moments and the cycle repeats.
I am not sure of what is causing this issue.
For the records, the access point is an old D-Link router. Why I mention this? Because funnily whenever the access point cable is plugged into the main router and I login to the router, the system logs me into the access point router (D-Link instead of TP-Link).
Can someone please help me resolve this issue?
Fun fact: The D-Link was a giveaway by one of my dR friends @Bigus-Dickus8
Debbuging options that no customer uses since it makes the logs unreadable.
Formatted the logs and in 3 Years Not one customer used the feature or asked for it.
An automatic tool (like smartgit) for our internal use. Not one uses it, instead they still complain about git2
- yo bro do you have some time ?
- quick cause I'm taking a dump
- I think I have been hacked, got black screen kernel panick, linux freeze seldomly I have to reboot, no internet connexion
- save your stuff and reinstall linux
- I don't have enough stockage to backup
- Then buy one and save, probably either OS is fcked up or you have some hdd problems
Time that it will take: ~30min to reinstall whole shit
Peace duration: ~2years
Later on the same day
- I can't log into windows
- Did you change the password ?
- Yes but it does not work anymore
* looking at shit
* logs successfully. Reason: interface changed after automatic update.
* wait some more so fucking windows fucking starts
* Desktop is ugly as fck.
* Some stupid settings messed up (like high contrast set, black theme or so)
aunt (the same)
- I can't log into my (other) laptop either
* wait more more more
Guess what: automatic updaaaates. Freezes 100%cpu
* Being a very experienced user: wait before reboot because this suckass os will probably fail to boot otherwise
* Blackscreen with a percentage: Installing updates...
* Blackscreen with a percentage: Installing updates continuing...
* finally boot (feels like a miracle windows succeeds lol)
* still slow
aunt now sleeps
* look at running process and install programs
* sees shits like camera recognition (vendor installed), candycrush
* occasionnaly get adds
time lost: 2h
peace duration: ~3month
FFS I am a dev, not a fucking trash lover
It is already pain to fix someone os, but windows is the cream of cream
It brings no ease of use for novice user
It is so insanely slow
It has stupid settings set up by default!!!!!!!! Who FFS wants candycrush and ads
The maj are so fcking hazardous. It is 2022 pretty much the same as 15y back then. Updates take fucking eternity. And needs reboot. and are not even finished!!!
I swear I am gonna stretch my ass and install linux and any fckin other toolsuite needed so they can use Micro$$ word, which is the only fucking usecase they need windows for in the first case anyway
I SO wish this OS would die
I mean, even more than safari8
Anyone tried converting speech waveforms to some type of image and then using those as training data for a stable diffusion model?
Hypothetically it should generate "ultrarealistic" waveforms for phonemes, for any given style of voice. The training labels are naturally the words or phonemes themselves, in text format (well, embedding vectors fwiw)
After that it's a matter of testing text-to-image, which should generate the relevant phonemes as images of waveforms (or your given visual representation, however you choose to pack it)
I would have tried this myself but I only have 3gb vram.
Even rudimentary voice generation that produces recognizable words from text input, would be interesting to see implemented and maybe a first for SD.
In other news:
Implementing SQL for an identity explorer. Basically the system generates sets of values for given known identities, and stores the formulas as strings, along with the values.
For any given value test set we can then cross reference to look up equivalent identities. And then we can test if these same identities hold for other test sets of actual variable values. If not, the identity string cam be removed, or gophered elsewhere in the database for further exploration and experimentation.
I'm hoping by doing this, I can somewhat automate the process of finding identities, instead of relying on logs and using the OS built-in text search for test value (which I can then look up in the files that show up, and cross reference the logged equations that produced those values), which I use to find new identities.
I was even considering processing the logs of equations and identities as some form of training data perhaps for a ML system that generates plausible new identities but that's a little outside my reach I think.
Finally, now that I know the new modular function converts semiprimes into numbers with larger factor trees, I'm thinking of writing a visual browser that maps the connections from factor tree to factor tree, making them expandable and collapsible, andallowong adjusting the formula and regenerating trees on the fly.7
Is there an ios app that records my gps logs for last n day(cyclic buffer)? Privacy is also important: data shouldn’t leave my phone: no internet access.9
Two years ago we took over this project which has been a nightmare to maintain. It's a set of netcore 2.1 webapps running on an on-prem windows machine. Everyone who has worked on it so far has quit, leading to two episodes of it being passed on with near zero handover.
Its function is fairly simple, so naturally we have been nagging to redo it and cloudlift it.
I was finally given one week to see how far I'd get, and had a poc running in Azure after one day; 4 apps in clean net6, SSO, and managed identities. The only thing lacking was setting up the authentication for third parties.
And... they still don't want "something new" when the old one works. Back to IIS and debugging windows event logs.1
What the hell is the point of this small projects team spending 2-3 months on developing extensive logging system for an internal application for inside and outside customers to use if your application isn’t going to log any of the fucking errors. Sure you write the failure status to the database, but it just says failure with an even more vague explanation than microsoft’s errors. “An error occurred”. No shit, that’s why I’m looking in the logs and database to debug the application to get these files on their merry way so our company can stay in compliance with the state, feds, and not pay out the wazzoo in fines. All our other applications state where the error occured such as “failed to connect to the email server”, why can’t this one.
In most businesses, self-proclaimed full-stack teams are usually more back-end leaning as historically the need to use JS more extensively has imposed itself on back-end-only teams (that used to handle some basic HTML/CSS/JS/bootstrap on the side). This is something I witnessed over the years in 4 projects.
Back-end developers looking for a good JS framework will inevitably land on the triad of Vue, React and Angular, elegant solutions for SPA's. These frameworks are way more permissive than traditional back-end MVC frameworks (Dotnet core, Symfony, Spring boot), meaning it is easy to get something that looks like it's working even when it is not "right" (=idiomatic, unit-testable, maintainable).
They then use components as if they were simple HTML elements injecting the initial state via attributes (props), skip event handling and immediately add state store libraries (Vuex, Redux). They aren't aware that updating a single prop in an object with 1000 keys passed as prop will be nefarious for rendering performance. They also read something about SSR and immediately add Next.js or Nuxt.js, a custom Node express.js proxy and npm install a ton of "ecosystem" modules like webpack loaders that will become abandonware in a year.
After 6 months you get: 3 basic forms with a few fields, regressions, 2MB of JS, missing basic a11y, unmaintainable translation files & business logic scattered across components, an "outdated" stack that logs 20 deprecation notices on npm install, a component library that is hard to unit-test, validate and update, completely vendor-& version locked in and hundreds of thousands of wasted dollars.
I empathize with the back-end devs: JS frameworks should not brand themselves as "simple" or "one-size-fits-all" solutions. They should not treat their audience as if it were fully aware and able to use concepts of composition, immutability, and custom "hooks" paired with the quirks of JS, and especially WHEN they are a good fit.
This guy has been a “php programmer” for 3 years now and yet he doesn’t know where to find Apache logs, is it weird that I am finding this really strange?11
What are your use cases for noSql dbs? I haven't really found a reason beyond stuff like chat messages or logs, but even those tend to work perfectly fine with SQL.
I imagine they're pretty good for prototyping, but haven't really tried them out for that yet. Perhaps for cases where you're handling billions of records?9
TL;DR I have to bump a Redis cluster from t3.medium to m6g.large just to get enough network bandwidth even though I have no need of the extra memory.
Debugged an interesting issue today.
I am adding Elasticache to a project to reduce strain on the single node postgres DB.
Deployed a Redis replication group with 2 shards, with multi-AZ replication for resilience.
Everything was going well. We arent caching that much atm so was barely using 100Mb of memory.
Suddenly, when our US region comes online, latency skyrockets and the logs are full of Jedis timeout errors.
Still no issue with memory or node CPU.
The cause? Arbitrary network bandwidth throttling by AWS. The app currently processes about 3,000 requests per second so we were exceeding Amazons random ass allowances which arent documented anywhere.1
Thanks, management, for letting us use AWS glue but not letting us have a dev endpoint so we need to determine what the fuck is going on by reading logs and divination if there is a deep problem
#Suphle Rant 9: verbatim exception scare
In multiple rants, I've bitched about laravel stealing suphle features. By some very weird coincidence, it appears I've been given a taste of my own medicine. Let me explain:
We're having a chat this morning on a laravel group chat when someone says he uses their notification component a great deal. Curious, I ask him what he uses it for since I only used it sparingly during my laravel days. To pry an answer out of him, I ask whether he uses them for sending app error alerts to a slack channel, and he responds with an eerily familiar term. I quickly look it up and the results on the docs are chilling: errors can be sent to bugsnag (which suphle has an integration for), sentry and Co. Errors can either be broadcast or disabled. Specific kinds of errors can be caught. My heart sunk. My brother called for something while I was going through it and I was struggling to pull myself together
Their exception component is almost identical to mine and I'm only just realising. It's shameful that I'm just learning about functionality present since 5.8. I thought my creation was novel. BUT! The good news is, the implementation differs
Too many errors went unnoticed during my time there because error broadcasting is optional. Since none of my colleagues read that part of the documentation, we were firefighting by pulling and wrangling production error logs. This informed their abolishment in suphle altogether
A relatively minor difference is in the APIs –their philosophy makes significant use of global functions, violating SRP, etc.
But the most important difference, that still cheers me up, is that they only catch known errors. Suphle has a construct for isolating calls to a decorated service. Any unforeseen error to occur during its execution will do a series of things before control is returned to the caller
How do I deal with this;
Edge case hiccup on production, no errors in the available logs(very shallow logging), no access to the production server, issue unreproducable on staging and a manager that want me to fix it AFTER I already said that im kind of sailing blind and can't do much without logs or access, and already looked at it with another dev who also has no idea what is going on3
I have a server that's is happy being full. Well atleast that's how I think of it ... I deleted all the logs and anything I thought was taking up space the server will go from 93GB space ussd out of 120Gb. To 119GB used in least than 10 minutes... am tired I fell like just turn it off physically in the server room and going to sleep .. f%$$%k this . I have cleared logs like 10 times now4
I have the following scenario with a proposed solution, can anyone please confirm it is a secure choice:
- We have critical API keys that we do not want to ship with the app because de-compiling will give access to those keys, and the request is done before the user logs in, we are dealing with guests
- Add a Lambda function which accepts requests from the app and returns the API keys
- Lambda will accept the following:
1. Android app signing key sha1
2. iOS signing certificate sha1
- If lambda was able to validate them API keys are sent back.
- Can an attacker read the request from the original (non-tampered) apk and see what the actual sha1 value is on his local network?
- If the answer to the question above is yes, what is the recommended way to validate that the request received is actually from the app that we shipped and not from curl/postman/script/modified version of the app11
I spent hours trying to figure out why a specific library couldn't be found on my system. I finally tore apart the ./configure script to no avail. I did a full text search of the source code and found a `config.log` file mentioning it was having trouble with a sub-dependency. I had failed to install OpenSSL.
Despite already having a few years of professional experience dealing with Linux servers, I still, to this day, confuse, which environment file gets sourced and when...
There's /etc/profile, /etc/bashrc, ~/.bash_profile, ~/.profile, ~/.bashrc
I think it's... Bashrc for interactive shells, profile for login shells.
But then I have examples like "ssh user@server 'echo $var'" that... Don't source any of the files!
You can enable user environment files for SSH that get sourced whenever a user logs on through SSH (~/.ssh/environment / environment specified for a key in ~/.ssh/authorized_keys)
Is there some sort of master environment file that gets sourced *every* time, no matter what kind of shell starts?1
Moengage is one of the worst analytics software I have ever worked with...
Integrating it into a react website is a pain in the ass, they don't have a npm package, you need to add a script tag to html file.
It also has a wierd bug that the service worker they mentioned in the documentation doesn't work when the debug logs are off.
Aaaargh. Now I have to make a service worker handler to import this service worker and see if it works...