What kind of project / prompts - what’s working for you? /I spent a good 20 years in the software world but have been away doing other things professionally for couple years. Recently was in the same place as you, with a new project and wanting to try it out. So I start with a generic Django project in VSCode, use the agent mode, and… what a waste of time. The auto-complete suggestions it makes are frequently wrong, the actions it takes in response to my prompts tend to make a mess on the order of a junior developer. I keep trying to figure out what I’m doing wrong, as I’m prompting pretty simple concepts at it - if you know Django, imagine concepts like “add the foo module to settings.py” or “Run the check command and diagnose why the foo app isn’t registered correctly” Before you know it, it’s spiraling out of control with changes it thinks it is making, all of which are hallucinations.
I'm just using Gemini in the browser. I'm not ready to let it touch my code. Here are my last two prompts, for context the project is about golf course architecture:
Me, including the architecture_diff.py file: I would like to add another map to architecture_diff. I want the map to show the level of divergence of the angle of the two shots to the two different holes from each point. That is, when your are right in between the two holes, it should be a 180 degree difference, and should be very dark, but when you're on the tee, and the shot is almost identical, it should be very light. Does this make sense? I realize this might require more calculations, but I think it's important.
Gemini output was some garbage about a simple naive angle to two hole locations, rather than using the sophisticated expected value formula I'm using to calculate strokes-to-hole... thus worthless.
Follow up from me, including the course.py and the player.py files: I don't just want the angle, I want the angle between the optimal shot, given the dispersion pattern. We may need to update get_smart_aim in the player to return the vector it uses, and we may need to cache that info. We may need to update generate_strokes_gained_map in course to also return the vectors used. I'm really not sure. Take as much time as you need. I'd like a good idea to consider before actually implementing this.
Gemini output now has a helpful response about saving the vector field as we generate the different maps I'm trying to create as they are created. This is exactly the type of code I was looking for.
I recently started building a POC for an app idea. As framework I choose django and I did not once wrote code myself. The whole thing was done in a github codespace with copilot in agentic mode and using mostly sonnet and opus models.
For prompting, I did not gave it specific instructions like add x to settings. I told it "We are now working on feature X. X should be able to do a, b and c. B has the following constraints. C should work like this." I have also some instructions in the agents.md file which tells the model to, before starting to code, ask me all unclear questions and then make a comprehensive plan on what to implement. I would then go over this plan, clarify or change if needed - and then let it run for 5-15 minutes. And every time it just did it. The whole thing, with debugging, with tests. Sure, sometimes there where minor bugs when I tested - but then I prompted directly the problem, and sure enough it got fixed in seconds...
Not sure why we had so different experiances. Maybe you are using other models? Maybe you miss something in your prompts? Letting it start with a plan which I can then check did definitly help a lot. Also a summary of the apps workings and technical decissions (also produced by the model) did maybe help in the long run.
I don't use VSCode, but I've heard that the default model isn't that great. I'd make sure you're using something like Opus 4.5/4.6. I'm not familiar enough with VSCode to know if it's somehow worse than Claude Code, even with the same models, but can test Claude Code to rule that out. It could also be you've stumbled upon a problem that the AI isn't that good at. For example, I was diagnosing a C++ build issue, and I could tell the AI was off track.
Most of the people that get wowed use an AI on a somewhat difficult task that they're unfamiliar with. For me, that was basically a duplicate of Apple's Live Captions that could also translate. Other examples I've seen are repairing a video file, or building a viewer for a proprietary medical imaging format. For my captions example, I don't think I would have put in the time to work on it without AI, and I was able to get a working prototype within minutes and then it took maybe a couple more hours to get it running smoother.
Also >20 years in software. The VSCode/autocomplete, regardless of the model, never worked good for me. But Claude Code is something else - it doesn't do autocomplete per se - it will do modifications, test, if it fails debug, and iterate until it gets it right.
I think back on the ten+ years I spent doing SRE consulting and the thing is, finding the problems and identifying solutions — the technical part of the work — was such a small part of the actual work. So often I would go to work with a client and discover that they often already knew the problem, they just didn’t believe it - my job was often about the psychology of the organization more than the technical knowledge. So you might say “Great, so the agent will automatically fix the problem that the organization previous misidentified.” That sounds great right up until it starts dreaming… it’s not to say there aren’t places for these agents, but I suspect ultimately it will be like any other technology we use where it becomes part of the toolkit, not the whole.
>You seem to be under the impression that the word "screening" means TSA can do whatever it wants.
I assure you TSA thinks it can do whatever it wants. I say this as a white male and have certainly heard even worse stories that my own of egregious violations from people with other demographics.
Except that it appears one of the primary reasons this has become a thing is that the Feds are angry at states like Washington that don't verify citizenship when issuing driver's licenses. The whole point was that Washington (as an example) wanted to make sure people were able to get an identification and driving with a license (IE: some degree of documentation, had achieved some degree of driver's education and testing somewhere along the line...) regardless of their immigration status - and that pissed off the Feds. So it shouldn't be related to citizenship but that's part of how we got here.
Great to see Boeshield in this thread - so much of what's happening in this thread is the wrong product for a particular application. As you point out, Boeshield is a great product for protecting cast iron
Boeshield has a tendency to increase friction though unless buffed really hard.
Lanolin based coatings (fluid film, et al) don't have this issue.
Of course, i live in a super-humid place these days, so i have to control humidity anyway. This doesn't stop rust, but it means i can worry a lot less about which coatings and how often.
Ahem. There are neighborhoods in the US where you leave nothing in your car because otherwise your car will become a target. It's often "the rule" in these places that you also leave the doors unlocked because that way "they" won't break your window trying to get in. They open the door, see there's nothing of value to steal and move on. In other places in the US it's (still but fading) normal to leave your car doors unlocked because "everybody knows everybody and no one would steal from each other." Code switching is knowing which of the neighborhoods you are in and how to adapt.
I totally get and respect the perspective of the parent poster, I'm just keeping it real that the US is generally not a high-trust society. If it were, we wouldn't have disclosures and disclaimers and limits of liability for everything we do all day long.
>I'm just keeping it real that the US is generally not a high-trust society.
Completely false, you mean Urban areas are not high trust.
I live in a place (In the US) where kids walk to school, don't lock bikes and our downtown has free umbrellas to take and give back whenever there is rain.
Educate me: How is the Canyon, Ranger, or Frontier not a modern equivalent to the S10? All small(ish) trucks available in a two door or extended cab configuration with basic options.
Small pickups could be pretty fuel efficient. The problem is not CAFE standards but the fact that zero Americans buy small trucks, because the entire market for new vehicles in the US is people who are financially illiterate and easily marketed to and making them buy $80k brodozers is more profitable than a $30k S10
Half these people still choose to buy the vehicle they do for insane and superficial reasons like "It's got a Hemi", like my uncle, even though Hemispherical combustion chambers haven't been state of the art or even good ICE technology in decades.
They're substantially larger in all around size. Like comparable to a Dodge Dakota. A Maverick or Santa Cruz is comparable to a historical Ranger or S10, with the caveat that they're only available in one cab and bed configuration.
Maybe I'm an outlier here (but I don't think so...) in that CarPlay is an absolute non-negotiable. I don't care (and don't really want...) it to handle climate control, but music, podcasts, weather, messaging, phone, and navigation? Heck yes. The built-in systems are bollocks and 99% of the planet has already committed to Android or Apple for these features in the rest of their outside-the-car life, so the dumbest thing any auto manufacturer could do is push against the tide.
I don't want any ties to a phone manufacturer or OS, honestly. I'll have my car a lot longer than I have my phone. Just give me decent sound and an charge/aux connection. The rest I can do on my phone with a mount.
> I'll have my car a lot longer than I have my phone
Doesn't that make CarPlay/Android Auto a good thing? Provided the car supports both platforms, it means you can change phones during your car's lifespan without having to worry about losing features, and you get new phones as your phone upgrades without having to change your car.
I don't know, what if a third mobile OS is developed and gains traction (or is just one that I prefer) but it isn't supported by the car. What if future releases of iOS or Android are incompatible with a 10-year-old version of CarPlay or Android Auto?
CarPlay and Android Auto were engineered from the start to be as agnostic to the car's hardware as possible. Your car's stereo is really just a dumb screen (i.e. just a display and input/output interface) with the phone doing most of the rendering + a few other things (i.e. providing some car instrumentation, like fuel remaining, if the manufacturer enables it) - the hardware requirements aren't really strict from a performance standpoint (minus CarPlay Ultra, and even then, that's just a tighter integration).
Not just push against connecting your own phone, but also charging you for the features your phone could already provide (for free or for a cost you're already paying as a phone user). GM with $10/month data plans, and Toyota (per the article) with $200/year plans to access navigation.
They might try. But nobody is going to pay for a sub par experience.
Some executives will get a big bonus for pushing this out, and later another for reverting back.
I won’t buy a car that does not have CarPlay. And I know many who are the same. My phone is a centerpiece of my life (gahh I hate saying that), my car however is not.
How about Android but implemented by the car company? (Literally using apps like Google Maps but the main app menu is the car company's skin). I believe this is in effect what GM is doing? That satisfies your argument of using Android or Apple.
That doesn't solve my issue at all, though it might be fine for some. I use my iPhone for everything - maps, podcasts, music, and of course phone and messaging. When I get in my car, I want it to instantly become my mobile office connected to my iPhone. Building those features into the car, regardless of the technology, does nothing for me if it duplicates rather than synchronizes with my out-of-car life.
Yeah but that problem happens with companies like Tesla as well given their oldest cars are dog slow now on the latest OS. I have heard some reports of more bugs being introduced on the oldest models because of less QA (not fully sure though)
The only way out of that is to ignore the infotainment and use your own device like with an AUX port.
Back when I was a software developer, I needed a Mac Book Pro or Mac Pro. But as a Realtor, an iPad makes for an excellent laptop. Extremely portable and does everything I need in a mobile productivity device. For many people, it is absolutely everything they need in a computing device and gets better with each release.
