HNHacker News
TopNewBestAskShowJobs

yallpendantools

495 karma · joined March 19, 2023

Ever since I started studying CS/Software Engineering, the holy grail was a rigor in this discipline comparable to that of aerospace engineering. Repeatable builds, predictable processes. Hence, it is ABSOLUTELY MADHAT BONKERS to me that we are now "fine" with LLM-ing away our industry.

Fuck off if you push AI code without review. Especially fuck off if you merge to master code with undue diligence and wild abandon. Y'all crazy.

submissionscomments
yallpendantools··on Institutions try to preserve the problem to which they are the solution
I'm gonna have to play advocate here for The Institution over The Individual. The jobs you mentioned (pest control, basic car maintenance) exist because a lot of people can't be bothered to learn how to do it properly.

(Or, to put it more cynically: a lot of people are idiots. Like, a lot. Cue the quote about the average man and the thought that half of humanity is dumber than that...)

A lot of things can go wrong with the $10 spray you bought from Home Depot. You could end up spraying it where you're not supposed to and at best you end up poisoned in the ICU, at worst you contaminate your area's water supply. You could spray it on a lazy Saturday afternoon but you forgot about your dog who loves to lick the floor; at best you end up with a very expensive vet bill, at worst your dog then licks your kids in the face and you end up with a dead dog and a dead kid.

To be clear, I'm not saying your gut feeling is wrong; I'd probably do the same, honestly. But it most certainly doesn't apply to everyone.

Further, oil and filter changes might be super easy, barely an inconvenience but you could end up not resealing and tightening a valve or a nut enough and the worst time to find that out is when you're doing 100KPH on a highway. Don't even get me started about people who think they could save money by using olive oil where they are supposed to use a specific type of coating grease or lubricant; after all, they buy olive oil from the grocery once every month so, as a car maintenance item, it's "basically free".

These people are not only a danger to themselves. They are a danger to everyone they share a road (or a residential area) with.

Of course, having an industry around these tasks doesn't eliminate the possibility of these dumb outcomes but having "professionals" who are have read the fabulous manual and are regulated put them head and shoulders above J. Handyman Smith when doing said job. Emphasis on regulated, we give that far less credit than it deserves. You might have also read the fabulous manual but if you are not regulated, not beholden to a specific standard or process, how do I even check you didn't cut corners? If you mess up somewhere, how can you and everyone else even begin to assess the magnitude of your fuck-up and thereby respond appropriately?

yallpendantools··on Why is no one making a new version of old Facebook?
I've personally considered doing something along these lines but one thing simply stopped me on my tracks: moderation.

I feel like twenty odd years ago you can just open the gates to a web community and get by with little moderation; you can still cultivate a community from which participants derive some value. Nowadays, moderation would have to be no less than someone's full-time job. And I believe this has been the case for a good few years #before# generative AI came into the scene. Maybe for a good while you can hold down the fort with technical moderation solutions like fail2ban or Captcha but I bet you will run into problems within your first half a million users or so.

Maybe projects like Mastodon have more robust solutions to this but, still, moderation will be a problem if and when you make it big, just like your server costs.

Finally, don't even get me started about CP, which is a whole other level of low and rotten. Maybe if you restrict your community to niche topics it would (a) deter malicious outsiders and (b) be self-moderating to a degree. The thing is, a niche discussion forum old Facebook ain't.

yallpendantools··on Read the Chronicles of Narnia in Publication Order (2018)
> I can't imagine how jarring the transition between _The Magician's Nephew_ and _The Lion, the Witch, and the Wardrobe_

> Nephew feels more like Lewis' earlier science fiction novels, while Wardrobe has a fairy-tale quality

Will vouch for this. I started Narnia in my early teens. I knew Wardrobe to be a classic but the edition the bookstore carried had Nephew labeled as the first book. Honestly? I felt kinda short-changed at what I got. Nephew set-up this expectation of being a myth-heavy story, something in between Harry Potter and LOTR/Earthsea in terms of world-building. Jackson's LOTR was at peak popularity at the time too so I really lapped Nephew up.

Honestly though, I loved every book in the series. I can understand why people might have issues with Caspian---maybe the weakest book in the whole Chronicle IMO---and looking back now the Christian imagery might be heavy-handed at times. But as children's/early teens' fantasy series, the stories and the world are entertaining. Which, as a teenager, was all I really needed of the books. Maybe Lewis' only fault is that he peaked too early with Wardrobe, setting up expectations for the rest of the series.

yallpendantools··on Neocities – Developers API
Git and HTML both deal with the (let's face it) complicated data structure we all know as trees.

The difference: You can fuck up an HTML tree and most browsers can still display your page[1], you can still get your message across. You fuck up a git history tree and you can somehow summon Cthulhu.

Not to mention, even if you can manage to stay away from branch confusion, just dealing with diffs is already a huge barrier for newbies.

[1] See "Where HTML 'beats' C" - Computerphile https://www.youtube.com/watch?v=-csXdj4WVwA

yallpendantools··on Neocities – Developers API
THIS! I've been maintaining a Neocities website for almost three years now. While, being a software engineer at heart, I still ultimately versioned the offline copy of my website via Git(lab), it's liberating to not have to wrestle with a git workflow just so I can hide my works-in-progress.

Also, in my case, a matter of branding. While I'm not trying to promote my content (a bunch of short stories and illustrations) so heavily, it just felt wrong for them to reside in Github/Gitlab subdomain. The closest I would've considered Git* pages is if I was doing sci-fi stories and that is still a "maybe". At the time I started my project, I was also worried I might run afoul of some content policy these Git platforms might have given that my website is not exactly for their demographic.

yallpendantools··on You Win or You Learn
Cool find! I'm trying to post a content-focused side project in as many grassroots platforms as I can. My main site is already in Neocities, for example. I think I could work with this too!

That said, I wish they featured some random pages or categories from their homepage. I'd like to see what communities already exist around the topic of my project.

yallpendantools··on Spring 6.1 now compatible with virtual threads and JDK 21
Let me throw-in my 2c worth. I started my career in consulting with an in-house but Spring-inspired framework. I hated it, turned to other jobs that used other tech stacks; nowadays I largely consider myself a Pythonista. Lately, after about a decade from that first gig, I've finally decided to learn Spring properly and I've made good progress with it. Mind you, I decided to do so only because it seems like a good career move to know Spring in this economy.

So, my honest evaluation: learning Spring is frustrating and not an enjoyable experience at all. Just the sheer verbosity of things hurts my Pythonista soul. It's not just that Java is inherently more verbose than Python, it's just that "the Spring way" forces you to create three classes with little to no shortcuts around it; even the most opinionated Python/Node frameworks I've encountered wouldn't be that draconian. Coming from Python/Node/PHP/serverless, the framework's philosophy seems incomprehensibly monolithic.

All the layers of abstraction (and their resulting stack traces) make for an extremely confusing debugging experience especially for noobs.

*BUT*, I admit, judging a technology for its learning curve is not fair. I just had to vent my honest opinion on the experience.

I'm now at that point where I think I've grokked "the Spring way". That is, my Google searches on how to do X in Spring/troubleshoot a problem are smarter now. I've actually grown to appreciate it. (I'm pretty sure the Spring Gods would make me pay for the sheer hubris of this statement and will make me eat my words in the next feature I come up with in my side project.)

JpaRepositories are powerful magic. If Spring has any advantage over the frameworks I've come to love, it's this. JpaRepositories compensate for what Spring/Java lacks in brevity. Oh, also, not a Spring-specific advice but Lambdas and method references have made Java easier to write and has definitely deflated the number of classes I've had to create otherwise. Some would say that's just lipstick on a pig but I prefer to see it as an oasis in a desert.

Lastly, don't take this as advice for the backend you are building for a new start-up but I really see Spring's strength in that consulting CRUD-software market. If I can define specs in a legally-binding contract before writing a single line of code, if I can bang out a schema at the start of development and not worry about migrations, I would actually choose Spring. All the "corporate-rep" of Java won't hurt for these types of projects either.

But on the other hand, for environments where I need a lot of flexibility and lack the leisure of time almost always, I feel like Spring would be a development death march.

Misc notes:

1. I'm not actually learning vanilla Spring. I'm using Spring Boot alongside Vaadin. Overtime I've come to distinguish better between the two but maybe some percent of my frustrations in the learning phase is on Vaadin, not really Spring.

2. I made extensive use of ChatGPT in learning Spring. Even for such a boring/common tech, it's hilarious what ChatGPT gets wrong. But it also gets things right, giving advice in a level of clarity that blog posts and SO posts couldn't have provided. I would recommend anyone trying to learn Spring (or maybe any other framework/technology) to give ChatGPT a shot too. Even when it gets things wrong, the feedback loop that resembles actually talking to someone helps with the rubber ducking.

Again, all 2c worth of opinions from someone not even a year in this framework. Caveat lector!

yallpendantools··on AWS networking concepts in a diagram
Can someone actually read the image? I'm on desktop and I can't get it in a readable resolution. https://miparnisariblog.files.wordpress.com/2023/03/aws-netw... is still a web page and the image is, sure, larger, but no more legible. :(
yallpendantools··on LastPass users locked out due to MFA resets
I'm a developer by profession and I almost didn't switch from LastPass after their breach last year.

Simply put, after all the reports of last year's breach, I assessed how vulnerable I am. First, my LastPass settings were such that I shouldn't be too affected by their breach; among other things in their self-assessment report, I had the "new" healthy default of 600K iterations. Also, the three most important accounts forming the basis of my online identity were never on LastPass and had unique passwords.

(And yeah, I understand that the security issue isn't purely on technical merit but also a social question of LastPass' reputation as a company. But on a personal level, I didn't really care that much. Moving on...)

Hence, on a personal basis, I didn't see much reason to switch out. The alternative would be the hassle of evaluating a new password manager, exporting data from LastPass, setting up the new password manager on my devices, importing my pre-existing vault, tweaking the new password manager so it behaves as I expected, etc. I know I'm playing the world's smallest violin with this grievance but that's really how it was. I think there was also a confluence of other factors why I didn't want this hassle on my plate at the time (e.g., I remember this was end of last year and I'd rather focus on my holiday arrangements).

I did reach out to family members whom I might've recommended LastPass to in the past though, and advised them to switch out. I didn't believe they could make the same self-assessment that I did.

In the end, I did switch to Bitwarden though. I did go through the hassle as I thought I would but articles like this make me glad I did. The decisive factor for why I did it anyway was that I realized that I might have some passwords/keys in my vault that I use professionally so, out of professional prudence, I switched. Were I not a developer, I might not have had this factor at all.

yallpendantools··on Pynecone – Performant, customizable web apps in pure Python
Awesome! Now that I've seen this framework, it tempts me though...

I have this personal side project which in the wee days of me career I wrote in CherryPy; I love Python and have been exploring frameworks. I didn't really get a lot written in it because eventually I realized the project is frontend-heavy and I was never in the mood to write the JS for it.

Recently I had the need for this side project again so I thought, why not rewrite it using a framework that would avoid me having to deal with the madness that is the Node.js ecosystem? For one reason or another I ended up with Vaadin.

I've nothing against Java. If it works, it works. I got over the initial bump that is Spring and I got a "core" backend running, with tests too. My grievance with it is the usual Java fault of needing to write a lot of boilerplate just to shoehorn my app into the framework's philosophy. (Man, part of my initial consideration was even the tooling, editor, autocomplete, and all that, because damn me if I try Spring without these modern conveniences!)

And this...this is all the things that attracted me to Vaadin---but in Python! I am so tempted to rewrite again. My experience and your documentation should make this a two hour job, right? Right?

("Haha, I'm in danger," my side project chuckles like Ralph Wiggum, staring blankly at https://www.commitstrip.com/en/2014/11/25/west-side-project-...)

yallpendantools··on Artist refuses prize after his AI image wins at top photo contest
> I see no reason why editing his image with AI-assisted tools should automatically disqualify the artist, unless the contest terms prohibit that, of which I see no mention in the article.

He didn't edit an image so much as generate it. His work is closer to a digital painting (from a Wacom tablet, say), than a color-corrected, cropped, then filtered photograph. Everyone else started their process with capturing light from a camera.

yallpendantools··on Epic Games, others accuse Sundar Pichai of violating retention obligations
Caveat: I did not read the embedded pdfs but I did read the article, and nothing there even alludes to anything like a retention policy.

Even the wording of the accusation does not even imply a policy:

> Like Mr. Pichai, other key Google employees, including those in leadership roles, routinely opted to move from history-on rooms to history-off Chats to hold sensitive conversations, even though they knew they were subject to legal holds

Routinely opted is not a policy. A policy is formal, not an option you routinely take.

I'll wager even the best lawyers will be hard-pressed to prove this routine is primarily to render discovery impossible. In fact, on face value alone it will be hard to objectively prove any purpose to this, unless more info/evidence/data/witnesses come up.

The text that follows my quote

> they did so even when discussing topics they knew were covered by the litigation holds in order to avoid leaving a record that could be produced in litigation.

makes assumptions that you can only speculate on, not prove. Sure, they were discussing topics covered by the litigation and then pivoted to private chat within nine seconds. Does it follow that the part of the conversation they moved off-record (and hence we are not privy to) is also definitely covered by litigation and could affect the outcome of the trial? Not necessarily (although likely, alas courts don't settle for likely).

yallpendantools··on Epic Games, others accuse Sundar Pichai of violating retention obligations
Your analogy uses loaded terms (co-conspirator) and makes use of biased character portrayals (alone by a sketchy bridge at 2AM, don't wear a wire).

Even in this case, unless the court issued an order that forbids physical meet-up/contact between the accused, I don't see anything the court can hold against them. Meeting at a bridge at 2AM is not a crime, and doesn't mean they are avoiding anything other than other people.

yallpendantools··on Epic Games, others accuse Sundar Pichai of violating retention obligations
But in this case, an "irrelevant topic" can range from something exclusively Google's business (like details of a new product) to something uber-mundane or even personal.

So, really, unless the legal hold on Google is a blanket, encompassing order to preserve all internal comms, I find it difficult to be objectively certain they are at fault here.

yallpendantools··on Epic Games, others accuse Sundar Pichai of violating retention obligations
But how can you say the conversation that happened off-record does pertain to the matter at hand? How can you preserve something that was not created in the first place?
yallpendantools··on Epic Games, others accuse Sundar Pichai of violating retention obligations
> Turning history off when you are under a retention order is likely to be equivalent to destroying records, in the eyes of a judge.

> "routinely opted to move from history-on rooms to history-off Chats to hold sensitive conversations, even though they knew they were subject to legal holds"

Nah, I think that's too far. There are simply too many legitimate reasons for Google to turn history off. What if they are switching to a topic irrelevant to the case and, therefore, not covered by discovery? You can only prove they had a sensitive conversation they wanted no record of, but you can't prove that that sensitive conversation is under the "legal hold" they were in. Even if all the context of the on-record chat seem to imply a crucial bit is coming up, you can't really prove the crucial bit did come up. It's all based on speculation.

> If they wanted to be truly off the record, they should have met in person or made a phone call.

So... `spichai: let's take this to an ephemeral chat` will be viewed negatively but `spichai: let's discuss this in person` is completely normal? I'm not buying it.

yallpendantools··on How America Manufactures Poverty
Maybe don't present a problem as "American" when the problem is global? The problem is not so much that America manufactures poverty as much as that the consumerist(/capitalist) economic system the majority of the world depends on manufactures poverty. The article is just listing out American symptoms of a global disease.
yallpendantools··on How America Manufactures Poverty
I agree with you but unfortunately beggars can't be choosers. America speaking English, strength of the USD, all contribute to making the US the most attractive (if not the only) option. When you're poor and just taking chances, you make the best of what you're presented.
← PreviousPage 6 of 6