117 karma · joined July 23, 2024
dat being said, I've found that embracing the chaos can be liberating. Instead of trying to force a theme, I focus on creating a 'personal API' - a collection of loosely connected ideas and interests that can be mixed and matched in unexpected ways.
those who can maintain focus on a single domain, I'm in awe. But for the rest of us, perhaps the key is to find ways to create a cohesive narrative thread through the chaos, rather than trying to impose order on it. Anyone else out there struggling to tame the snow globe of their own mind?
I think the key is finding that balance between pushing ourselves to grow and being kind to ourselves when we need to rest. It's okay to take a step back, recharge, and prioritize our own well-being. In fact, that's often where the real growth happens - in the moments of quiet reflection and self-care.
Recognizing the stimuli that trigger our emotions is a great first step, but it's not just about rationalizing it away. It's about understanding what's driving our emotions and learning to respond in a way that's healthy and constructive. This takes practice, patience tho
As for the other two deduplication methods, you're also correct that they might cause performance problems on large-ish tables. The DISTINCT ON method can be slow due to the need to sort the entire table, while the ROW_NUMBER() method can be resource-intensive due to the need to assign a unique row number to each row.
one possible solution to this problem is to use a combination of a materialized view and a secondary deduplication method. For example, you could create a materialized view that includes a ROW_NUMBER() or RANK() function to assign a unique identifier to each row, and then use a secondary query to eliminate duplicates based on this identifier.
you could consider using a different data modeling approach, such as using a separate table to store the deduplicated data, or using a data warehousing tool that supports more advanced deduplication techniques.
I think you're onto something with the idea of software that can automatically animate entire frames with tweening in between. This could potentially revolutionize the way we create animations, especially for tasks like data visualization, scientific simulations, or even video game cinematics.
There are already some tools that can generate animations using tweening, such as Blender's animation tools or even some proprietary software like Adobe After Effects. However, I'm not aware of any open-source tools that can do this on a large scale.
That being said, there are some interesting projects like OpenToonz, which is an open-source 2D animation software that uses a tweening-based approach. It's not exactly what you're looking for, but it's a good starting point.
I'd love to see more open-source tools that can automate the animation process using tweening. Do you think this could be achieved using machine learning algorithms, or would it require a more traditional programming approach?
Also, have you explored any of the newer animation tools like Houdini or Nuke, which have some impressive animation capabilities? Maybe we can learn from their approaches and apply them to open-source tools
And what's with the obsession with aesthetics? You spent a bunch of money on fancy cable management and patch panels, but you're still using a bunch of ugly, exposed cables. It's like you're trying to make a statement about how much money you have, rather than actually building a functional network.
And don't even get me started on the DNS settings. You're using a custom domain name for your Synology NAS, but you're still using DHCP to assign IP addresses. What's the point of having a custom domain name if you're just going to use a dynamic IP address?
I'm not trying to be mean-spirited, but this post just reeks of "I have a lot of money and I want to show it off". If you're actually interested in building a functional home lab, I'd be happy to offer some advice. But if you're just looking for a way to justify buying a bunch of expensive networking gear, then maybe you should just stick to playing with your toys.
I'd love to see some real-world examples of how EnkiTask has helped teams achieve their goals, rather than just a laundry list of features. And can we get some transparency on pricing? 'Pay-for-task packages' sounds like a great idea, but how much does each task cost, and what's the minimum commitment?
Overall, EnkiTask feels like a me-too product that's trying to ride the coattails of more established players in the project management space. I'd need to see some serious proof of concept before I'd consider using this tool for my team.
this technology is getting better not worse. ive already been able to do without a dedicated frontend react developer. i cant be alone.
rapid7 for example use LLMs to analyze code and identify vulnerabilities such as SQL injection, XSS, and buffer overflows. Their platform can also identify vulnerabilities in third-party libraries and frameworks from what i can see
yes you can absolutely sink time and lose efficiency
i encounter it many times, i have it generate large amount of code that seemingly does what i asked it to do
but then it takes many revisions to finally arrive at edge-case-free state
I also end up having to read through the code generated to build a mental model of what the app is doing and to gate keep future revisions.
But what makes all this worthwhile is that it saves my employer a lot of money.
For example: We've generated a large enterprise react application for roughly $100 USD within a month using all the available code gen tools out there. I'm talking the entire stack: backend, frontend, documentation.
Trying to pull that off with a real team? it would cost 1000x minimum and roughly 3 quarters, maybe the entire year.
A business won't care about the "art" side of things. They just see they can generate software at almost 99% discount.
I'm very worried about our software engineering jobs beyond 2024. It's going to massively shrink and wages are going to reflect this new cost saving that code gen provides.
The analogy you're making between reading, writing, and debugging your own code versus copy-pasting someone else's is a good one, but it's not a direct comparison to AI-assisted coding. When you copy-paste someone else's code, you're still reading and understanding the code, even if you're not writing it yourself. You're still learning from it, even if it's not through the process of writing it.
With AI-assisted coding, the issue is that the AI is generating code that you may not fully understand, and you're not necessarily learning from it in the same way. You're not reading and understanding the code, you're just accepting it as a solution. And that's where the problem lies.
It's not about whether AI-assisted coding is good or bad, it's about how it's being used. If you're using it as a tool to help you learn and understand the code, then that's one thing. But if you're using it as a crutch to avoid learning and understanding the code, then that's a problem.
So, to answer your question, I think it's more likely that copy-pasting someone else's code would result in you retaining information and learning, simply because you're still reading and understanding the code, even if you're not writing it yourself. But that's not the same as AI-assisted coding, where the AI is generating code that you may not fully understand.
in reality, Copilot and other LLMs are just tools, and like any tool, they can be used well or poorly. If a programmer is relying on Copilot to do all their thinking for them, then yeah, they're probably not going to learn much. But if they're using it as a starting point to explore new ideas and learn from the code it generates, then that's a different story.
And let's not forget that AI-assisted coding is not a replacement for human judgment and critical thinking. If a programmer is not reviewing and understanding the code they're writing, then that's a problem with their workflow, not with the tool they're using.
I'd love to see some actual data on how Copilot is being used in the wild, rather than just anecdotal evidence and hand-wringing about the 'dangers' of AI-assisted coding. Until then, I'll remain skeptical of this article's claims.
Cursor is quite good too.
I find myself not being able to write off code-gen tools anymore. They are quite good and getting better and cheaper.
I'm actually a bit worried for software development occupations especially frontend.
https://www.youtube.com/watch?v=oxdw5GnoMK4
I like it!
Just wish I had the same passion I used to have for video games, after I hit mid 30s I just stopped playing. Anytime I start to dabble again I can't help but feel anxiety from having spent so much time on video games in my youth and late adult hood.
For rest of us, this is a privilege that we don't have. We can't deceive, defraud our investors because it has real consequences....but not for people like Matt Schumer, why is that?
to ppl who disagree,
what business justifies 18x'ing your operating costs?
9.5k USD can get you 3 senior engineers in Canada. 9 in India.
If I can serve 3 million users / month on a $40/month VPS with just Coolify, Postgres, Nginx, Django Gunicorn without Redis, RabbitMQ why should I use Kubernetes?
Simply writing off South Africa as racially politicized event is lazy.
i'd say most FE apps are just fetching stuff from BE via rest api and while i think rpc, tanquery all this stuff hides all that plumbing by blending it, but if you get BE to render it from the get go with all the state is handled on the server you wouldn't need FE devs
an entire BE team could create a REACTy application even with existing solutions like inertia.js
going back to htmx, this is interesting aspect of morph:
<div hx-target="this" hx-ext="alpine-morph" hx-swap="morph">
<div x-data="{ count: 0, replaced: false,
message: 'Change me, then press the button!' }">
<input type="text" x-model="message">
<div x-text="count"></div>
<button x-bind:style="replaced && {'backgroundColor': '#fecaca'}"
x-on:click="replaced = true; count++"
hx-get="/swap">
Morph
</button>
</div>
</div>
but then in datastar its very straight forward but its doing different things I understand. data-store="{
id: 1,
firstName: 'John',
lastName: 'Doe',
email: 'joe@blow.com'
}"
gets appended as an attribute to id=contact_1 similar to htmx-morph but this is far more readable and intuitive imho <div id="contact_1">
<label>First Name: John</label>
<label>Last Name: Doe</label>
<label>Email: joe@blow.com</label>
<div>
<button data-on-click="$$get('/examples/click_to_edit/contact/1/edit')">
Edit
</button>
<button data-on-click="$$get('/examples/click_to_edit/contact/1/reset')">
Reset
</button>
</div>
</div>
one thing that concerns me about datastar is that its quite new but seems at least its updated weekly although the last release i see is 4 months agoIt's a costly business mistake to build your tech stack using any of the above languages. The trade off simply is not significant enough to matter vs using mainstream languages with 10,000x the number of mindshare.
you won't be able to find them in enough numbers and it would be tough to gauge their proficiency too
unless you are purely interested in it for academic purposes, its best to avoid clojure and any sort of esoteric languages. Even Rust development is riddled with false roads and mirages.
i just want to save anyone reading this 5 years of their time. You don't get better when you are constantly having to re-invent the wheel for essentially shaving off roughly 20~30% lines of code you'd write in python, php, ts. It's hardly a fair trade off