Hacker Newsnew | past | comments | ask | show | jobs | submit | bhnmmhmd's commentslogin

This has happened to me. I was learning Haskell until finally the pressure of "how will this help me pay bills" persuaded me to stop and learn Python instead. Even with Python there's the pressure of "I can't make an online sideproject by it" which has tempted me at times to switch to JS, but I've resisted that so far.


Good thing you resisted. You absolutely can do an online project with Python :)


This.

I studied Electrical Engineering and in my experience, the only thing my classmates cared about was the technical aspects of the field, and often they couldn't see the big picture. Great talents of course, but more like robots, not even capable of understanding the humanitarian aspects of life. Such a waste.

Disclaimer: It's just my experience, not necessarily true for all engineers and engineering schools.


That's one part of it, I'm sure. Another part is that if you pipe up, you risk getting shunned: loss of income, loss of status, loss of being "part of the cool kids". Sure, you can get another job (that pays less) but you'll probably not work at the same cutting edge technology, and when you tell relatives or friends who you work for, their eyes don't light up.

If somebody offered me five to eight times as much as I make today with a huge boost in status and other positive side effects, and all they ask is to quiet down those silly principles, I'm not so sure I'd say no. "You can still do good in 5 years when you've made enough money to be set for life", I'd probably tell myself.


<quote>If somebody offered me five to eight times as much as I make today with a huge boost in status and other positive side effects, and all they ask is to quiet down those silly principles, I'm not so sure I'd say no. "You can still do good in 5 years when you've made enough money to be set for life", I'd probably tell myself. </quote>

The worst thing is, sticking to those principles just means someone else will take the job.

Coming out of university, I had one job offer (it was a bad down market at the time). My principles made me say no. I didn't want to work on guidance systems for military applications, thank you very much. That choice has cost me a great deal in earning potential. Even so, I would find it hard to blame someone for compromising their principles.


Did they not pay any attention during the humanities portion of their education? That's part of why it's there.

Disclaimers: I'm not an engineer, but my CS degree is ABET accredited. I also took a few more Archaeology, Cog Sci, and Philosophy classes than strictly necessary.


Another personal data point.

When I was in college I did have to take philosophy 101. It and all of the other core classes such as biology and English were nothing more than annoying hurdles. I do enjoy learning about philosophy and such, but that is not at all what happened while I was trying to get my degree.

Though to be totally honest, to a large extent I considered the entire degree earning process to be little more than an about hurdle to entering the job market. I spent practically zero time reflecting in any of it during the process.


Personal data point:

Did a Bachelor's + Master's degree in CS in Europe. Only around 5% of my coursework had (and could!) be outside CS. And you could fill that with things like Game Theory 102 or Copyright Law 101 or something without issue.


That's more like the typical engineering school format. I went to a Liberal Arts university, nearly half my classes were general humanities and sciences. The rest were in my major (also required an additional 18-21 credits for the ABET accredited degree option).


This makes sense, as you get your broad education before university. School focusses on a well-rounded education, whereas universities let you specialise. I realise this is a bit of a contrast to the US education system, with majors and minors, etc.


TL;DR:

Sell it per download.

Why?

If it is to be run in areas that rarely have access to the internet, then your target market isn't probably interested in buying a license or any thing that even remotely makes the app hard to use.

Go easy on your customers. From the moment they visit your website to the moment they have the app up and running should not take more than 3 clicks. Just my 2 cents.


> The internet is based on trust.

Maybe that's why we need to move towards blockchain-based networks since trust is not required on blockchain, only proof.


Then people have to trust the people with the time, energy, and interest to know how to develop the stuff behind it. All it does is shift the trust to people who are even less accountable.


Thanks, I need a curated list of all such websites :)


Actually it's not just that. I believe sometimes people forget what enormous achievement it was to have _one_ centralized repo for all the apps a fragile, mobile device can have. Without AppStore, the dark ages of Symbian suddenly come into mind. And sure, Apple has had a number of flaws when it comes to AppStore policies and regulations, but part of Apple's sternness and rigidness is well justified, given the myriad of apps it has to serve, and safety and security of billions of devices that run those apps.


It’s arguably an advantage for users. But not necessarily for developers.


It's a complicated issue to be sure, but developers also benefit from a platform that has high trust among users. I think users are probably much more likely to try out new apps and even pay for them if they know ahead of time that the apps been vetted, the payment processor is secure (and already has your details), etc.

Not that there isn't crap on the iOS App Store, but I don't think it's controversial to suggest that the Good to Bad App ratio is better there than on the web as a whole.


And for some devs and some apps that’s worth giving up a 30% cut and dealing with Apple’s processes and restrictions. For some it won’t be, if they can distribute the app any other way.


I was wondering, has anyone here applied cluster analysis techniques for anomaly detection?

I read a paper that used it for insurance fraud detection, but I don't know what other fields are using clustering to detect frauds and abnormalities?

I'd be grateful if someone can help.


Yes, tons of that.

See this - using K-Means clustering for anomaly detection in web traffic:

https://imgur.com/a/qWMgUo0

Using DBscan clustering for anomaly detection in healthcare claims data (detecting doctors who anomalously prescribing opioids). Using public CMS data set from 2015.

4 out of 8 top anomalies (doctors) were later actually convicted of crimes or gone into all sort of troubles with DOJ:

https://imgur.com/a/6wFWTg5

https://imgur.com/a/f721ndb

(Splunk Enterprise + free apps was used to ingest data and build all this logic and dashboards)


Thank you so much, it really was helpful.


I mostly agree with you. But really, who needs +16GB RAM on their laptop?! I think the need for more RAM is correlated with the need for more CPU clock, more powerful graphic card, and a better Motherboard, all of which result in a huge decrease in battery life (which is supposed to be a laptop's main strength over heavy, power-consuming PC).


I need the RAM. Network modeling, network simulation, dev installations of SDN platforms. This is a developer and programmer forum, right?


It's the reason I end up always looking at gaming laptops. I don't really need a GPU at all, it's nice but not needed. However yeah, Cores & RAM that's what I need.


I was about to buy a new MBP this year but held back because I already have 16GB of RAM and want more, actually need more. I often have the problem that my current MBP runs out of memory and when this happends it costs me like 30 to 60 minutes to get back to the point where I was when it ran out of it. So yeah, there are some of us that need more RAM.


> But really, who needs +16GB RAM on their laptop?!

I need it on whatever machine is my development machine. Clojure REPLs are memory intensive. I'm sitting at 7GB in use right now, mostly thanks to Chrome, even without anything major running.

I could probably squeeze under 16GB if I was aggressive about shutting down server applications (e.g. Apache and MySQL) I'm not using at that moment and only starting them up when I need them, but that's just adding more chores to my workflow.

Having more than 16GB of RAM just means I never need to worry about it. It's one less headache to deal with.


As a developer, I need a machine that's powerful enough to develop on; being light and having battery life is a nice-to-have but not essential. If that means I have to buy a bulky gaming laptop rather than a sleek portable one, that's what I'll do. At the moment I can just about manage with 16GB, but it's definitely the limiting factor on my current machine (surface book 2, which is capable of some pretty good gaming).


For me, large datasets used for analytics. I don’t want to use a desktop, I want to do this stuff in a coffee shop, and in my hotel. The difference between in-memory and disk based operations is, understandably, very significant.


As (mostly) a designer, even I'm pushing RAM usage at times. Some GPU wouldn't hurt either.

yet I don't feel like switching away from Mac is a realistic option for me. There are a ton of apps on which I depend. Can't exactly stick them in a docker image.


I would think most developers would like at least 16gb


From Game Theory perspective, "not using Linux" is actually the game's equilibrium. Because if you have to choose between a proprietary OS (Windows, macOS) and an FOSS (GNU/Linux, BSD, ...), you'd have to choose the former in order to increase your own utility. If people said they're gonna start using Linux from now on, that wouldn't work, since at least some are willing to "unilaterally deviate" from that decision and use Windows/macOS to gain a boost against others.

If you want people to start using FOSS, you'd have to offer them something they can't have on non-FOSS alternatives.


The offer is there. It's freedom, and most people don't want it.


A 'free' computer that doesn't actually run any of the tools you need is worthless.

Makes a good doorstop, I guess.


Indeed. You are free to modify it to run those tools, or create alternate tools that do run (maybe not so much after Oracle v Google). But ain't nobody got time for that.


This is such a bullshit argument.

I don't want to spend my time making a toaster - I just want toast.

Also this is why FOSS software continually reinvents the wheel, badly. Want some BS program? There are a million of them. Want an actual tool that requires deep talent and domain knowledge? Outside of compilers, if it's FOSS, it is almost certainly garbage.


I think you are imagining the argument you want to see. Your comment is not responsive to what I actually wrote.


It's was about time we moved on to the next energy-storage technology for mobile devices. I can't believe in 2018, battery is still the main issue of all flagships.


Won't it always be an issue? I have a feeling they'll just put in more and better power-hungry hardware and software and we'll be back in the minimally acceptable battery life range.


Right. And if they don’t do that, then the phone will just get thinner. We already have incredible energy density compared to a decade ago, but that budget is used up by shaving millimeters off the thickness.


> used up by shaving millimeters off the thickness

I wish they didn't do that, I would gladly carry around a slightly bigger phone if it had much better battery life.


I have an LG V20 with a 6700mAh battery. Goes for DAYS with moderate use, and will easily make it all day with extremely heavy use. I managed to kill it exactly once, with 4 hours of screen-on nav + steaming spotify, another 3 hours of youtube, and probably some misc usage. It's fantastic, and I keep wishing that somebody would release another phone with a user-replaceable (and upgradeable) battery so that I could move on. I actually bought this V20 after the V30 and S9 were out specifically for the removable battery. sigh, I blame apple.


Many phone models do have expansion options like this, even iPhones. I for one am appreciative of “low power mode” so I get to choose efficiency over app freshness.


Yeah, but that (1) feels like the hardware equivalent of DLC, and (2) the resulting unit is much thicker than it would be if the battery were designed into the device from the start.

I get why some people prefer thinness over battery life. Not all use cases are the same. Just wish that there was an option when choosing a model to go with svelte or chunky.


A Jevon's paradox on battery life. Hmm, yea.


I think the issue is still all the background activity. My phone (moto x pure) lasts multiple days in airplane+battery saver mode. It was 30% charged after using it for GPS/maps and occasional music on a 4-day backpacking (hiking) trip.


Put the battery we have now into a Nokia flip phone from the 90s. It’ll probably run for weeks.


Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: