HN user

starky

1,244 karma
Posts0
Comments513
View on HN
No posts found.

This is a good way of putting it. People who advocate for over-engineering don't consider cost-benefit vs not doing it. This is where FMEA analysis comes in when developing products, you assign values for severity, occurrence, and detection to determine which potential issues need extra effort implemented, but just as important is identifying the potential issues that aren't worth putting effort into to prevent over-engineering.

I've recently had to deal with a problem where the parent company has dictated 100% production testing for a spec that we meet for the spec sheet but customers practically don't ever need. But since the parent company test their products this way they say we have to. I've been very vocally against this because it has delayed us shipping products that are otherwise ready, prevented starting new products, and added a lot of production cost when our RMA rate without the test is something on the order of 0.01% and most of those are due to installation errors, not true failures.

But for me the take away still is: “hardware is as hard as you make it”.

I dislike this statement. Hardware is as hard as the product dictates it needs to be. A 25 component PCBA and a clamshell of 2 injection molded parts is about as simple of a product as you can make. Heck, most people would just buy an off the shelf clamshell for that type of product.

That doesn't work for the vast majority of products, building a product with 20 COTS parts plus 60 custom tooled parts, and 4 complex PCBAs is what is hard. Getting everything to fit together, pass testing, arrive on time from dozens of suppliers, etc. it where the complexity comes in. Then consider that hardware is cash intensive, you need to pay ahead of time for all the tools, you need to pay to buy the individual parts and warehouse them somewhere while you build the products you hope are going to sell. If something goes wrong there, then your expensive parts are just sitting there waiting for replacement components before you can ship. Then even when you ship, that revenue goes into buying the next round of parts you are going to receive in 3 months. This is where many hardware projects fail as they run out of cash before the project breaks even.

Seriously, microSD cards are the second worst hardware component to have to deal with (first being labels). For being a "standard" you end up with cards that are subtly but significantly different from each other, the slots are ALL poorly designed, and microSD cards are just inherently unreliable. Plus with the current flash shortage, the major manufacturers don't care at all to support any issues with the product.

That is awesome. I worked at a bowling alley as a mechanic in high school (Brunswick A2) and got a sense of how stupid expensive the replacement equipment was. At one point another employee spilled their coffee over the keyboard used to control the score-keeping system, apparently the replacement was something like $400. What fancy proprietary keyboard did you get for that $400? The keyboard cable replaced with a phone plug.

People think this doesn’t matter, but GoPros are used all over in aerospace. If we replaced the brand with Insta360, that puts a big attack vector all over the place.

I'm actually surprised to hear that GoPro is used in aerospace applications. I assumed that they would use cameras from companies like Teledyne that have a more industrial/military customer base. That is to say, there are almost certainly options for domestic camera manufacturers if required.

You are right about China. The West (especially Apple) taught China how to manufacture while losing the ability ourselves and for some reason didn't consider that they would figure out how to design products themselves and bring the entire product chain within the country.

Individuals at the end of the day mostly care about buying the cheapest thing that does the job. They don't consider where it comes from. Even those that do care have limits in how they will compromise (there is a reason why I bought Bambu Labs 3D printer even though I would have preferred a Prusa).

It seems like every recommendation algorithm gets into a state that is broken eventually. For the music streaming services, I find they end up in one of two states, either it shows me the exact same artists and songs over and over again even if I skip their songs constantly, or I'll play a song once out of curiosity or accident and it decided that that genre is all I am interested in listening to for the next month. Is it really that hard to map artists

I got so frustrated with Tidal recently that I finally sat down and finally setup a media player on Linux to play my locally saved music (most of which is from What.cd).

The functionality is literally built into both Android and iOS. For example, I've got some of my self-hosted services linked on my home screen. On Firefox for Android you just browse to the site, click the 3 dots, and then under more its "Add to Home screen", Chrome works similarly.

This is China's massive differentiator than anywhere else. They have impeccable supply chain management. You will find that certain items are almost always made in the same city. Shenzhen (and surrounding area) is just the hub for electronics manufacturing. Within the city there is the main output, and then a bunch of smaller manufacturers making all the input components within driving distance of that factory which allows it all to work incredibly efficiently. This makes manufacturing in China so much faster than elsewhere.

I'd even go so far to say that Shenzhen is probably the wrong place to build the electric train motor above. I bet there is some 2nd or 3rd tier city somewhere in China that specializes in just that. For example, the only reason I'm aware of the city of Yueqing is because I did a project with pushbuttons once and that is the pushbutton city in China.

It depends on how you look at it.

I've traveled all around Asia visiting different factories for projects over the years. After some time, factories just tend to blend together, big rooms filled with machines where materials come in one end, and are modified to be more like the final product at the other end. The rooms might look/feel different, whether it is the dirty floors and heat of a die cast grinding room, the green floors of an injection molding factory, or even a clean room you have to squeeze yourself into a bunny suit for (They never have large enough suits for me), its all just equipment in a room to transform material.

A factory is also a set of people and processes, which is what truly matters. If you don't have good people with expertise and the processes in place you don't have a chance at getting good product. You build relationships with the people at the companies and they will move mountains to help you. You build a truly excellent relationship with a factory, and they will anticipate your requirements and sometimes ship you perfect parts at T0.

I've got plenty of examples of factories that are fantastic, and ones that have been horrible from my time in working. The good ones all have the same differentiator, people that we have built a good long term relationship with. The bad ones almost always have a constant rotation of different people you have to talk to.

In simulations, their approach solved 99% of Wordle puzzles, while the traditional method solved just 90%.

This seems wrong to me, getting a 98%+ solve rate for Wordle is pretty common.

Tourism is a huge benefit for many places and brings lots of money into their economy. But at the same time over-tourism can significantly harm locals and the environment.

My photo example is just a example anecdote that proved a point, being a major city having a very localized area get busier is pretty inconsequential, but if its a lake view somewhere it definitely draws a risk of irreparably damaging the environment.

Usually explained by a different time of week/year/month

I hadn't thought about that and checked the dates of the photos I took on both trips. Coincidentally apparently I was there within a week of exactly 10 years apart!

While the folks like Bourdain did romanticize tourism, and TV shows are known to regularly kill businesses that don't know how to manage the increase in customers, they did at least project the proper attitude of embracing the differences in other cultures when they traveled to places.

The current issues with tourism are significantly more to do with "influencers" and social media. Many places are overrun by people that are just there to get their photos and have zero interest in engaging with the culture or treating locals with respect.

Its shocking how different some places have gotten due to "influencers". Last year I was in Kuala Lumpur for a few days and took the person I was with to a bunch of the places I had visited when I was there a decade ago. It struck me when walking around a couple places that there are photos I took during my first trip that would simply be impossible to get today because of the number of people in the way.

Yet when I switch between home and work I have to fully restart my laptop about half the time in order for it to even detect the monitors. I also find this feature has an issue with certain programs (Obsidian in particular) where it opens the window almost off screen.

A good portion of it is just that I'm more comfortable with Solidworks and Creo, so if I'm modeling something complicated I'd rather use those so the tool isn't in my way. Otherwise the main thing is the surfacing tools in other software are more mature and feature filled.

I made the switch about 1.5 years ago and haven't looked back. The only software I use frequently that doesn't have a good enough equivialent available on Linux is photo editing and CAD. Even for CAD, if it is simple enough, I'll do it in Onshape rather than booting up my Windows VM.

I asked it since I wasn't sure if I was just my brain being the odd one out here.

I tend to just remember topics and search for them when needed. For example, I rely completely on Outlook's search to find the email from 5 years ago that I remember X person sending (and I mention Outlook specifically because surprisingly Gmail is terrible at this)

How many people actually find utility from a Zettelkasten system?

I just can't bring myself to go to the effort of documenting a thought and adding links/tags unless it is something I predict that I will need sometime in the future and won't just remember. Due to this, my Obsidian vault is pretty much a collection of a bunch of temporary to-do lists and then some folders with specific reference information. If I'm linking thoughts together I'm doing it real time in my head, anything else takes me too far out of my thought process.

I can see it if you are a person working in academia or a writer where you may be generating concepts that you want to link together in the future. But as someone that does project type work, I'm following too much of a defined process to see any benefit.

Massdrop's original version is essentially what you are describing, they were a middleman that ran drops where you would pay for something upfront and once done they would collect the money and order it from the manufacturer before shipping it to you in a couple months. They used to sell a wide variety of things, but over time they moved more and more to their collaborations and Drop branded goods in the headphone and keyboard spaces and essentially became just another online retailer.

Due to some comments the CEO made regarding a few Trump decisions there are a number of people that believe that the company supports the current US government and are looking for any link they can to prove it.

This is true, and it is also true that the maintainers of Wayland have done a terrible job of developing the replacement. It is mostly good enough now to replace X11, but based on what I've seen reported about different features, they frequently let "perfect be the enemy of done" when it comes to implementing critical features. I mean, just look at the drama around remembering the position of a window, its absolutely ridiculous that after years they haven't picked a "good enough" direction and implemented it.

A lot of success in working with suppliers in China (and really anywhere in Asia) is in building a relationship with them where they know exactly what your expectations are and holding them to it until they understand that it is just easiest for them to do it right to start.

I've got suppliers who I can send a difficult part to and know that I'm going to get exactly what I expect, faster and cheaper than just about anyone else. It took a few years to get to that point, but these few vendors make it really hard to go with anyone else, much to the chagrin of the sourcing team who rightly recognize it a risk to rely on just a few suppliers.

Once you get to a certain type of supplier you end up running into the problem where their processes are such that they won't do anything without you clearly documenting it. They simply refuse to make any assumptions on your behalf. They can be so frustrating when you are used to the other way of doing it. I simply cannot answer some questions because I'm so used to my other suppliers just doing it correctly and haven't ever asked about it.

Plus to my knowledge NVidia drivers are still an issue.

This is an overblown issue. For the most it is a case of installing the proprietary driver package following the instructions for your distro. One more step than AMD which just works, but not really any more difficult than installing another package.

This is frankly one of the areas of Linux that I'm most disappointed about. I tried a number of these options for local players and found them all to be terrible. Especially considering that I really like the specific layout I've used for well over a decade on Foobar2000. I actually ended up installing Foobar2000 through Wine because of my frustration, only to find that it wouldn't start reliably.

Someone told me early in my career that the longer you work in an office, the more Office Space transforms from a comedy to a documentary. They weren't wrong...

Long time SolidWorks user here with experience in other programs. Frankly, SolidWorks is one of the easiest pieces of CAD software to use, being much more flexible in how things are done compared to a lot of other programs. That said, it is incredibly powerful software, and while someone can learn how to use it in a week, it takes months or years to be actually proficient.

My big tip if you can't find a button there is always the search bar. Just search the command you are looking for, it will even show you where the button is located for next time. That said, they don't move things around that much from year to year, I'm surprised if you can't find a command in a tutorial made in the last 10 years.

The features you are talking about sound like you want to be doing surfacing, which is definitely a more advanced modeling technique that I only recommend trying to learn once you understand the basics and can predict how the software wants you to model something.