VS has the “Publish” functionality for direct deployment to targets. It works well for doing that and nothing else. As you said, CI/CD keeps deployment IDE agnostic and has far more capabilities (e.g. Azure DevOps, GitHub Actions).
BLE and WiFi are power hungry on any platform. LoRaWAN or zigbee are better for low power, low bw comms. Kinda sacrifice simplicity & ubiquity of support in that you need a gateway/router of some sort, but they are solid solutions with long range and high penetration.
I have to say I prefer Claude for code and diagnosis. ChatGPT in CoPilot often replies with essentially, "these are the things I would be looking at if I wanted to troubleshoot this", to which my muttering frustration is, "yeah obviously - no sht - now do it.", whereas Claude just does it (e.g. looks through the code, provides analysis as to where the problem is or the nature of the problem, and returns a practical and immediately useful response). You could say that's a prompting problem (and you'd be right) but if Claude is lower drag than ChatGPT in this context and others, that's a value proposition.
However, I really do* prefer ChatGPT for generalized but as-applied conversation about architecture decisions, pros/cons of different approaches, patterns. Claude can do this too, but I do prefer ChatGPT for this.
the dashes and the auto capitalization are awful for technical writing. The ctrl-z becomes painful and annoying very quickly. Would that Word supported markdown.
This is a ridiculous maladaptive behavior. Word has been replacing dashes forever, consequently it has been unintentionally ubiquitous in business writing forever. That this character ever became a heuristic for AI is silly.
beside the point, but over 50 = proto-boomer? You mean para-boomer, maybe? Gen X is <=60, I believe, so you referring to the cusp boomer/genx I think..
I do not think that word means what you think it means. Proto boomer would be pre-boomer or at least the very early boomer, and chronologically that makes zero sense. Understand what you're trying to say though -- Quasi, Para, Pseudo, Demi would each be more appropriate prefixes. A Demiboomer could be in the Monster Manual.
Are you bearish on the shovel seller? Is now the time to sell out? I'm still +40% on nvda - quite late to the game but people still seem to be buying the shovels.
Personal opinion, I'm bearish on the shovel seller long term because the companies that are training AI are likely to build their own hardware. Google already does this. Seems like a matter of time for the rest of the mag 7 to join. The rest of the buyers aren't growing enough to offset that loss imo.
FWIW, Nvidia's moat isn't hardware and they know this (they even talk about it). Hardware wise AMD is neck and neck with them, but AMD still doesn't have a CUDA equivalent. CUDA is the moat. As painful as it is to use, there's a long way to go for companies like AMD to compete here. Their software is still pretty far behind, despite their rapid and impressive advancements. It will also take time to get developer experience to saturate within the market, and that will likely mean AMD needs some good edge over Nvidia, like adding things Nvidia can't do or being much more cost competitive. And that's not something like adding more VRAM or just taking smaller profit margins because Nvidia can respond to those fairly easily.
That said, I still suggested the parent sell. Real money is better than potential money. Classic gambler's fallacy, right? FOMO is letting hindsight get in the way of foresight.
What's the old Rockefeller clique? When your shoe shiner is giving you stock advice it is time to sell (may have heard the taxicab driver version).
It depends on how risk adverse you are and how much money you have there.
If you're happy with those returns, sell. FOMO is dumb. You can't time the market, the information just isn't available. If those shares are worth a meaningful amount of money, sell. Take your wins and walk away. A bird in your hand is worth more than two in the bush, right? That money isn't worth anything until it is realized[0].
Think about it this way: how much more would you need to make to risk making nothing? Or losing money? This is probably the most important question when investing.
If you're a little risk adverse or a good chunk of your profile is in it, sell 50-80% of it and then diversify. You're taking wins and restructuring.
If you wanna YOLO, then YOLO.
My advice? Don't let hindsight get in the way of foresight.
[0] I had some Nvidia stocks at 450 and sold at 900 (before the split, so would be $90 today). I definitely would have made more money if I kept them. Almost double if I sold today! But I don't look back for a second. I sold those shares and was able to pay off my student debt. Having this debt paid off is still a better decision in my mind because I can't predict the future. I could have sold 2 weeks later and made less! Or even in April of this year and made the same amount of money.
I have absolutely no clue whatsoever. I have zero insider information. For all I know, the bubble could pop tomorrow or we might be at the beginning of a shift of a similar magnitude to the industrial revolution. If I could reliably tell, I wouldn’t tell you anyway. I would be getting rich.
I’m just amused by people who think they are financially more clever by taking conservative positions. At that point, just buy ETF. That’s even more diversification that buying Microsoft.
The thing about the UniFi platform is it iteratively improves. Years ago you couldn’t manage NAT rules or DNS from the GUI, though there were workarounds to modify iptables at the command line and preserve customization across upgrades.
Now days, static routes, SNAT/DNAT, and DNS are all in the management interface. So.. things improve, and every time I’m back using EdgeRouters, Extreme, or Juniper elements I miss the low friction of managing UniFi stacks.
Agreed that if you need VRFs for example, DC power, and are working through similar complexity requirements, Ubiquiti is the wrong stack. I’d say Ubiquiti is not heavy weight, but it seems to address 90% of SMB setups.
> The thing about the UniFi platform is it iteratively improves.
That's a very charitable and positive spin on "was expensive the day you bought it and got all the functionality you expected years later".
I'm fine with things getting better over time. I am a lot less understanding when you ship a device in 2024 and it still has trash IPv6 support but don't worry because "we'll fix it via an update coming soon!"
That is something that should have been there from day 1.
Well, that’s kind of how I write software so I can’t cast stones.
I would agree with the desire for full feature sets out of the gate, but UniFi stack is hardly expensive. Further, you know the balance of limitations and advantages when you opt to purchase it instead of more traditional mainstays with decades of compounding feature buildout.
I've always said that Unifi handles well enough the 10% of networking configuration that 90% of users need. If you're in that other 10% of admins who need something more complex then it's not the right pick, but in a great many cases it's strongly planted in "good enough" territory.
VS has the “Publish” functionality for direct deployment to targets. It works well for doing that and nothing else. As you said, CI/CD keeps deployment IDE agnostic and has far more capabilities (e.g. Azure DevOps, GitHub Actions).