I love it and use it for personal projects and internal tools. I tend to combine it with https://pocketpages.dev/ which gives me file-based routing and nice templates.
Ah, and Pocketbase has automatic database migrations, so all schema modifications can go into version control.
I even hacked a Gemini protocol server into it, so that I can browse my personal knowledge graph using Lagrange.
Talking of cheap and powerful devices one can also look at Chinese UZ801 4G LTE (Qualcomm MSM8916) dongles. They cost like only $4-5 and pack quite impressive HW: 4GB eMMC, 512MB RAM, actual 4G modem sometimes with 2 sim switching support. Since it's actually old Android SOC there is even GPU and GPS in there. And a lot of work was already done on supporting them:
Google Play Services is a dependency for some apps, and GrapheneOS allows for people to take steps to protect their privacy while still being able to use those apps.
First, with GrapheneOS google play services run in a sandbox like any other app. (play services have more privileged access in vanilla android)
It also works well with a multi-user setup. The default account in Android is the "owner account" and in GrapheneOS (and AOSP) you can use the owner account to create multiple distinct user accounts on the device. Then, you can only install google play services in one user account. Google play services won't start if you're not logged into that user account.
Google play services won't have visibility into your other user accounts and what you're doing there. And even in your account with play services installed, there's a bit more privacy because of the sandboxing (although I believe google play will know all of the apps installed in that user account)
Edit: I am a web security researcher and longtime user of GrapheneOS and have always been impressed by the features, frequent security updates, and focus on usability, security, and privacy. They've upstreamed numerous security improvements to Android and other open source projects (so if you're running Android, they've probably made your phone more secure!).
Can authors of such articles at least cite Dijkstra's "On the foolishness of "natural language programming"." which appeared eons ago ? Which presents an argument against the "english is a programming language" hype.
Fleshy meat sacks on a space rock eating one another alive and shitting them out on a march towards inevitable doom in the form of a (likely) painful and terrifying death is a genius design, no?
I think the genre has gotten smaller, but higher quality. Games are more a work of love than a marketing hype.
Thimbleweed Park, Kathy Rain, whispers of a machine, Disco Elyseum, Unusual Findings, Sexy Brutale, Darside Detective, blacksad, obra-dimm, takes two, Orwell, unheard, shadows of a doubt, 12 minutes…
Cloudflare -- Free for most services
OVH Cloud -- Free and unlimited
Scaleway -- Free for most services
Great:
Hetzner 20-60 TB / mo per instance $1.08
Not bad:
Linode 1-20 TB / mo per instance $5.00
Oracle Cloud 10 TB / mo $8.50
A bit much:
Backblaze 3x the amount of data stored $10.00
Bunny CDN -- $10.00
DigitalOcean 100 GB - 10 TB / mo per instance $10.00
UpCloud 500 GB - 24 TB / mo per instance $10.77
Vultr 2 TB / mo for most services $10.00
Uh...
Fly.io 100 GB / mo $20.00
Are you actually serious?
Microsoft Azure 100 GB / mo $78.30
Amazon Web Services 100 GB / mo $92.16
Railway -- $100.00
Zeabur 10-100 GB, depends on plan $100.00
Google Cloud Depends on service $111.60
Screw you guys:
Render 100 GB - 1 TB, depends on plan $300.00
Vercel 100 GB - 1 TB, depends on plan $400.00
Netlify 100 GB - 1 TB, depends on plan $550.00
(We use Netlify and have well over 1TB of monthly traffic. They're insanely expensive for what they are. As soon as we have roadmap time to revisit it, we'll move away.)
I'm starting to think of cloud as less of an asset and as more of a liability. We can leverage them for temporary scale, but in no way will we tie ourselves to a particular vendor.
- Have crystal clear Mathematical foundations, as in why this formula/method the way it is, rather than being able to solve college/HS test problems. Really solid footing in Differential Calculus and Linear Algebra is necessary.
- Know the Statistical language that you learn from a basic college-level Stat 101 course. Be able to translate normal sentences into those using Statistical notation, and be able to read easily. Also, know basic Statistics.
- You already know programming, I assume. Learn Python if you don't know already. It's really easy.
- There are a number of paths you can go from there. Here's what I did.
-- IBM Data Science Professional Certificate (not deep at all, but lays out the landscape well; did it in a week)
-- Machine Learning for Absolute Beginners by Oliver Theobald which you can finish in an evening.
-- Machine Learning Specialization by Andrew Ng on Coursera.
-- Deep Learning Specialization by Andrew Ng on Coursera.
-- fast.ai course.
- Learn PyTorch really well. I suggest Sebastian Raschka's book.
Now from here, you can chart your own path. You can choose NLProc, Vision, RL, or something else.
I went towards Vision. And I do Edge AI as hobby.
I was in the last year of college as a Physics undergrad, when I was hired to do Vision modelling/research for a non-flashy company in 2021. Finishing my CS Master's next month and starting to look for PhD. I worked in the same company for the ~2.5 years.
EDIT: If you want a job in big tech, grind Leetcode, and learn about system design, study Machine Learning systems, and be able to design them. Chip Huyen has a good book as I hear. 6-7 rounds of interview is common in Meta/Google. DL hackathon awards, open source contributions are significantly helpful.
As n gets bigger, it goes from 0 to 1.
When n equals x, it is 0.5
As n gets bigger, the difference between n and n+1 gets smaller
For two sufficiently large n's, the results are equal.
Say somebody told you about a new cafe in town and that it is completely awesome. The best cafe ever. What probability do you assign to it really being an exceptionally awesome cafe? If your x is 3, then the probability after one person praised it is 25%:
1/(1+3) = 0.25
And if another person told you about that cafe being awesome, the probability becomes 40%:
2/(2+3) = 0.4
And after 3 people told you the cafe is awesome, chances are 50% it really is:
3/(3+3) = 0.5
The changes in probability are pretty strong at the beginning. But after 1000 people reported about the awesome cafe, the next report makes almost no difference anymore. It only ups the probability from 0.997008 to 0.997011.
By changing x from 3 to 4, your formula becomes more "suspicious", by changing it from 3 to 2, it becomes more "gullible".
I wonder if this formula has a name already. If not, "the trust formula" might be a candidate.
It’ll teach you all the history behind the OS decisions we take for granted now. I was delighted to learn that early OS’s used to have only one global lock. Sounds silly, but it’s extremely effective — HN does the same.
It has a bunch of chapters presenting these kinds of “games” user programs can play to get an advantage against the schedulers.
Also it’s $10 for instant access to a pdf. I was amazed. Buying a book is always such a hassle, except this time it was 30 seconds.
It’s not much for some parts of the world. But I’m well enough from this that I even took the time to build a small calendar app (https://lowtechguys.com/grila) from which all the funds will go to my brother’s college costs so he can stop working 12h/day jobs.
Before this I tried creating paid web services but none took off. I realized I actually don’t use any indie web product after 8 years of professional coding. I’m only using web products from big companies like Google, fly.io, Amazon etc.
Desktop apps on the other hand, most that I use and love are made by single developers.
With the ascent of Apple Silicon, and the ease of SwiftUI, this has the potential of bringing a modest revenue while also being more fulfilling than a corporate job.
In case you’re curious how the code looks for something like that, here’s a small open-source app that I built in a single (long) day, which has proven to be useful enough that people want to pay for it: https://github.com/alin23/Clop
Main thing is don’t expect too much. You will not experience that same feeling of comfortable camaraderie again for a long time.
Make an effort to listen to people. Keep your ego in check - avoid saying “Well at my old company we did …” - just listen to what people are doing here and now.
I would also say try to maximise informal 1-on-1 time with other people, ask them about their work, their opinions, etc. If you are in a larger group don’t opine, just listen. If you have questions, don’t interrupt the group - use it as an opportunity to go ask the relevant person after and build a 1-1 relationship.
Try to keep a smile on your face. Avoid dark humour or joke-complaining.
Edit: Say yes to everything, at least once, for the first year. Any time anyone invites you to a drink or a talk or a meeting, just say yes.
Honestly, I've found that most ML tooling is overly complicated for most ML projects.
I use a paperspace VM + Parsec for personal ML projects. Whenever I've done the math an hourly rate on a standard VM w/GPU is better than purchasing a local machine and the complexity of a workflow management tool for ML just isn't worth it unless you are collaborating across many researchers. As an added bonus, you can re-use these VMs for any hobby gaming you might do.
The majority of ML methods train quickly on a single large modern GPU for typical academic datasets. The scaling beyond 1 GPU or 1 host leads to big model research. While big models are a hot field, this is where you would need large institutional support to do anything interesting. A model isn't big unless it's > 30 GB these days :)
Even in a typical industrial setting, you'll find the majority of scientists using various python scripts to train and preprocess data on a single server. Data wrangling is the main component which requires large compute clusters.
https://www.howhotwillitget.com, I made an overview of future climate scenarios of each country worldwide with the latest predictions by the World Bank, maybe this is what you are looking for. :)
With a light dose, it’s just been mild, being silly and giggling at colors and nature. Pleasant, but just kind of a fun drug honestly.
With a moderate dose, you tend to be introspective and consider your own life and relationships in different ways. Feels like therapy to be honest. Hard work. Emotionally draining. There’s usually a lesson at the end. Overall it is good for me and makes me a better, more empathetic person.
With higher doses in a so-called bad trip, it’s sometimes been just extremely confusing and not at all fun, to the extent that I’d forget my own name or what I’ve taken and basically just all memories. It’s very frustrating to be in that state, constantly trying not to fly away but not able to ground yourself. It’s a little disturbing, but not horrific. I think the reason for this is that I am afraid to lose my ego, or die, and I would actually benefit from a slightly higher dose so I'm unable to resist.
With really high doses though, it’s been a truly beautiful, enlightening experience. I still think of it sometimes. It’s very difficult to really explain in words, but it feels like kind of extreme empathy with all life.
Buddhists would call it going egoless. I just felt very at peace: we all live, we all die, and it’s all okay. I’m not particularly religious or spiritual at all but that experience really changed me. It feels like life is in all these different forms, and you just happen to be one particular variation of it, but you could easily be any other living thing, so you just feel love and empathy for all life. And you know it’s finite and you’re not afraid of death. I would still attempt to avoid death, but I was just at peace with it if it was inevitable. Also, when you come back to reality you’re utterly, utterly convinced that you’ve experienced something profound, possibly, despite all your rationality, there’s a message form god inside this mushroom. It's so strange coming back, because you have to repiece together your life: "I am... a human... I live in a city called San Francisco... I am an engineer...". It’s almost like the mushroom is laughing at you: “good luck going back to your normal life now.” You’re left with “wow? That was that?!” It’s annoying honestly, because I’m not religious at all and I know it sounds crazy to even say that out loud. Of course, that feeling does slowly fade away like a distance dream or memory.
I think in general mushrooms strip away your individual sense of self and zoom out your perspective, which is why with a small dose you consider your relationships and with a larger dose you can sometimes get scared because you lose your sense of self but you're still trying to cling on. But with an even larger dose (obviously it's not necessarily the dose, but it's a rough correlation) you blast through even further, and the distinction between you and everything else melts away, which is a truly profound spiritual experience. In all cases though, seeing fractals and cool stuff is very unimportant.
It wasn't a big serious project or anything, but for my first income generating project ever I made a small single player word game for Android that I should have tried to monetise more. Android had only been out for a couple of years at the time so it was easier to get noticed then.
The game was free with an ad-banner during gameplay. I was making a decent amount in ad clicks and also had players emailing me for a paid version that just disabled the ads. Instead of jumping on this to release a paid version and expand the game with more features, I ran into perfectionism issues + decision paralysis releasing any changes.
The game blew up for a few weeks and I got something like 0.5 million downloads eventually but didn't do anything to monetise it more so felt like I missed a big chance.
It was a great feeling checking analytics though to see that there was literally years of collective gameplay time being logged, and I've had several emails and reviews from players saying they've been addicted to playing the game for years.
My biggest lessons are probably:
1) Doing/releasing something is always better than doing nothing so don't let decision paralysis get in the way. "Don't let perfect be the enemy of good".
2) Don't fear releasing or making changes because you might get bad reviews. You'll never please everyone and even the perfect app will inevitable get some brutally unfair and weird/crazy reviews.
3) Some people are obsessed with word games!
Anyway, as a lockdown project for fun, I started making a new web version for mobile + desktop. :)
Ah, and Pocketbase has automatic database migrations, so all schema modifications can go into version control.
I even hacked a Gemini protocol server into it, so that I can browse my personal knowledge graph using Lagrange.