Ask HN: What weird or hard problems are you trying to solve?

I hope this counts as weird enough ...

PROBLEM:

If time travelers from the future were to visit you, it would be difficult for them to quickly prove their authenticity.

SOLUTION:

Temporal passwords. At the start of each year, you devise a new password. You commit the password to memory, but you never write it down or divulge it to anyone until Dec. 31, when you submit it to the Temporal Password Registry, which publishes it and promotes its dissemination.

RESULT:

Prior to their visit, time travelers from the future can look up your temporal password in the registry for the year in which they plan to visit you. Their ability to communicate a password that you have not yet shared with anyone provides evidence that they are actually from the future.

PROMO:

https://temporal-password.pressbin.com/index.html

Education.

Schools are functionally no different from part-time prisons. You must attend daily under penalty of law.

Many teachers are plain awful.

They rely on students being additionally taught by their parents. That forces parents to go to school with their children, which perpetuates the vicious cycle. The teacher recalibrates the class to the students who either understood everything the first time or had supplemental education and the rest languish.

Schools assign homework that is not easy to do when the student hasn’t fully grasped the concept. That burns time they could use to get better.

So... I am implementing Math common core in software. The first part is an automatic homework solver for math. Once we solved the student’s homework, we can teach them how to do it with generate problems. Crucially, there will be multiple perspectives and an ontology of topics do the student can backtrack to where they got lost in class weeks ago.

After we are good with math, we’ll do the same with English. It will probably not go too deep, but it will let students obtain the missing foundation of knowledge.

I’m working on a new way to talk online, with the goal of killing cancel culture, increasing understanding, and basically calming down current radicalization. Picture Reddit, but with the ability to anonymously share ideas with other people in your social circles.

My theory is that most reasonable people stay off social media, so places like Twitter end up filled with unreasonable narcissists. At the same time, discussing politics on a semi-anonymous forum like Reddit is pointless, who cares if someone on the Internet is wrong. But maybe there’s a better way of communication, something new, that lets you talk with people you actually know.

Existing mediums for note-taking (Evernote, Notion, Roam Research) are not sufficient for doing knowledge work over long periods of time. The functions these incumbents serve are primarily as “stores of knowledge” that we save because it’s interesting in the moment but never read through or “scratchpads” that we use once and never get rid of, which end up cluttering our information space like a junk drawer full of shopping lists and knick-knacks.

Serious thought involves more than just collections and associations: mastery requires repetition, creativity requires serendipitous discovery, and productive output requires flow states. It’s also a matter of acknowledging the fact that “units of knowledge” do not exist on their own: all knowledge is embedded in context (or “deeply intertwingled”, in the words of Ted Nelson), and without context, metaphor, and nuance, we cannot form meaningful connections. By baking these attributes into the medium itself, it’s possible to build an information space that’s simple to explore, can surface information when you need it, can augment the mind’s naturally ability to form connections, and can get out of the way the rest of the time.

I am trying to reinvent the web.

Replacing HTML/JS/CSS with a language called ALFI. It is stupidly simple in its design but still very powerful. Similarly to HTML you use it to create widgets, place them, and define their behavior. It is humanly readable like HTML but line-based instead of markup-based. Instead of nesting it uses references. This allows it to be streamed.

A big difference is that the language itself doesn't allow styling (like CSS), the downside being you get less flexibility but the upside being it will render correctly on any display with any resolution.

For this I have also written a new type of web browser called NAVI which takes ALFI code and produces (somewhat) beautiful widgets and renders them using OpenGL.

Source for both ALFI and NAVI: https://github.com/jezze/alfi

My own ALFI website: http://www.blunder.se/

You need NAVI to actually browse blunder.se properly. otherwise you will just see ALFI code. Also this is still very early so all features are not done yet.

Let's solve Harmony!

The first scientific experiment was conducted by 5th century BC Pythagoreans. They wanted to show that the basis for musical consonance was math. From that, they inferred that harmony in math accounted for the harmony of the cosmos. This integration of math+physics was very forward thinking.

But, if we fast forward to the present, we still don't have a complete scientific explanation for the basis of consonance and dissonance. Really! To make my own contribution, I've been running psychophysical experiments to investigate why consonant chords that are mathematically slightly dissonant actually sound much better than chords with perfect mathematical consonance. I've been gathering data with sounds but also with haptic vibrations and with visual flicker frequencies. This multisensory approach is fun because it produces visible rhythmic entrainment in the brain, as seen with EEG. My goal is to contribute to a general theory of neural resonance and harmony in human experience.

Why does this matter? Happiness is great, but I'd argue that what we really want is personal and global harmony. Note that harmony isn't sameness, it is unity in variety -- the resolution of conflict and dissonance into an integrated wholeness. We want inner harmony with our selves, harmony in our relationships with others, harmony in society, harmony with technology and harmony with nature. Happiness is individualistic but harmony involves the pleasure of virtue. I hypothesize that harmony can help set a better objective function for the future of humanity.

Harmony was also the objective function for the first deep learning neural network, Paul Smolensky's Harmonium.

Finally, harmony is also a central theme in classical philosophy. The concept had a massive influence in the Italian Renaissance and in the English Scientific Revolution.

I recently put together a reader for understanding Plato's views on Harmony. Comments are welcome:

https://docs.google.com/document/d/1lqXpXgWI5YMBCz1O0gCmrEwz...

I'm trying to find a way to generate stories using tropes as the building blocks.

There's a sample at https://random-character-generator.com/

There's an unreleased version, which focuses more on how to portray characters, rather than just what they are. For example, instead of saying "energetic", they'll be pacing about a bit.

I might just pivot it into a story/plot tracker for writers, and use it to fill out the blanks rather than generating full characters from scratch. Where the community can add in their own templates and tropes. An author can decide that they have a character who is stoic, cynical, and sarcastic, and the tool will generate a background story, how to portray the character, what conflicts they get into with other characters.

I am building a database for human movement. Right now each exercise or pose or movement is indexed by its name - Downward dog, squat, handstand, etc. But this gets confusing when multiple names apply to the same movement. The true identifier is the motion of your limbs in space. I want to encode that motion/position so that if you try and upload L-sit handstand and also L-sit, it tells you that you are try to upload a duplicate (except for the arms). Furthermore, hopefully each movement can be uniquely encoded into an ID that could be used at a webURL. If you did this you could also compute the similarity between two movements to indicate that one is a progression step to train the other.

I don't have a computer science background so encoding an compression are new to me, but Im a good hacker and I can quickly get things like open pose working. Im trying to complete this in the next 3 months. Wish me luck.

Writing a computer program is often the best way to convince yourself that you really understand a problem. And an already-written computer program is often a good way to document the behavior of a complex system because you can play with it to see what happens in all the edge cases.

So I'm writing curricula that use computer programs as the primary teaching tool. One is for computer science, where the idea is that anyone who can read some python can pick up all the important ideas from a formal CS education without sitting through a year or more of preliminaries. Over time I'm planning to add smaller sections on more advanced topics.

The other curriculum is theoretical physics. There's already a good book that does classical mechanics [1] in scheme. I've hired some postdocs to learn scheme and code lessons in general relativity, statistical mechanics and so on. I do the lessons, solve the problems, and then we talk about what worked and what didn't. I work on this about ten hours a week. After a couple of years I should have knowledge roughly equivalent to an ABD physics grad student, plus teaching material that can take anyone else to the same level from modest beginnings.

I'm looking for collaborators on this project so don't be a stranger. Twitter/email is in my profile

[1] https://groups.csail.mit.edu/mac/users/gjs/6946/sicm-html/bo...

Automating cloud architecture creation.

I'm building a "catalog" of architectures that you could use to create a complete cloud architecture on your AWS, GCP or Azure account in less than one minute.

So, for example, you could create a docker-based architecture with CI/CD, auto-scaling, zero downtime deployment, SSL, load-balancing, high availability and MongoDB in less than one minute in your own AWS account.

It's like Terraform with the user-friendliness of Heroku.

It's very hard because every providers have different APIs and concepts so you have to start from scratch for each.

I love working on it because cloud-computing may have so much impact in some organizations like biotech startups or NGOs.

Improving the performance of fountain codes and applying them securely to peer 2 peer file sharing.

A fountain code is an almost magical algorithm that can split a file of size n up into a (practically) infinite stream of blocks of size b, such that collecting any n/b blocks out of the stream can reconstruct the file.

Applied to p2p file sharing it effectively can eliminate rare pieces as well as the need to communicate which pieces people have. Related topic here is homomorphic hashing.

Unless I find something better before September my master thesis will be on this topic.

I've been mulling over an idea that is essentially a combination of personal ID, secure digital authentication and online communications all baked into one.

There's a EU directive instructing on how citizens should be able to identify online with eIDAS. In my country, you can use eIDAS to authenticate in basically any governmental agency portal, but you can't get any eIDAS enabled auth method as a citizen. The current way of authenticating is done via bank accounts or a paid extra mobile service that requires a non-prepaid mobile contract.

This is a relatively huge issue. First off, the Finnish government pays the banks for each auth any user does when they for example want to log into their medical records etc. It's a few million euros a year just for verifying the users.

There's also obviously issues with whom the banks serve, there has been some cases with them not taking foreigners or people with bad credit as customers, making it impossible for them to authenticate themselves.

The current EU directives also indirectly require that the banks should provide a bank customer the possibility to authenticate without needing to have a banking account (which costs money), but to my knowledge this still isn't possible. I pay around 20 euros a month just for the luxury of having an account, not everyone can afford that on top of other bills.

Auth services are not accessible for impaired users.

It's also basically impossible to manage who has essentially the power of attorney and over which matters, for how long etc. Either you have to give them your login info (good luck resetting your SSN) or try to use the services over the phone and somehow convince the other side that you have permission to manage things for another person.

There's no ways of authenticating who is using your accounts online and actually verify the users.

Basically, my idea is combining biometrics, PGP and having the government running the identity management themselves. This would have added benefits of basically enabling hashed throwaway addresses and info for use online while providing a free and accessible way of authenticating strongly online.

I'm working on an fully open source Physical Vapor Deposition System which is capable of producing thin flim solar cells, Ex(CIGS,CDTE) at efficiencies in the range of 15- 20%. The system is designed to produce one 10 watt cell for every 45 minutes. I could go into further depth here but for obvious reasons It would take me a while to explain everything.

Government – Citizen interaction.

Specifically, I'm working with a local pro-refugee organization in a densely immigrant populated region in Spain. There's a complex chain of steps that you have to go through in order to acquire citizenship. Only people with access to good lawyers are able to deal with all the bureaucracy of the process, without mentioning other problems (missing obscure expiry dates that reset your process, language-related problems, local government workers not actually knowing/willing-fully ignoring migrants' rights...).

There's a good network of volunteer lawyers working on this issue, but its not scalable. I'm working on a platform that would allow migrants to solve their own situation, by crowdsourcing the knowledge of lawyers on a case-per-case basis and offering a simple interface in their language to track open processes & discovered the ones they need to go through and how.

As an abstraction for this, I've been thinking on how we could improve citizen/government communication. A small use case / example for this could be refugee camps. My previous experience here is that they are small, disconnected communities with a top-down type of organisation towards the camp organisers. It shouldn't be hard to provide real-time tools for connecting both, potentially leading to things like asking for their needs, managing their legal situation, or even allowing for voting & self-governing.

How can we document "human society" in 1000 pages or less? I've been casually researching for a while, and will eventually write a guide on going from zero to the moon.

Step one is survival, basics of hunting, first aid and farming sort of stuff. That volume would end around homesteading and self-sufficient living.

Volume two would be establishing society in larger groups than a family unit. Things like job specialization (N roles for N people instead of 1/N of each role for each person), establishing trade (currency, weights & measures, supply chain), government (mostly what not to do, what to protect, how to adjudicate disagreement), public works ("roads are a good idea") and their ilk. Also medicine beyond first aid and basic care.

Volume three would be advanced STEM topics, getting from a functioning society to... more. Not even the sky's the limit. It should include blueprints for things we take for granted like refrigeration, telecommunication and birth control. It will include all the basics of physics, chemistry and biology required for smart people to fill in the gaps and launch a human to the moon and back.

I want to super-nerd-out about this, and publish it on Tyvek or something exotic so it'll last through decades of wear and tear (and water-logging and more), and include a ruler on the spine and it's own weight documented for reference.

Dentistry: can we replace dental x-rays with infrared? Can we build optical panographs just using the reflections from a dental mirror? How can we monitor patients' oral health over the long term more often than just an annual visit to the dentist, and does that improve oral health outcomes?

Machine learning: what happens when we replace Euclidean metrics with p-adic ones? Distance is fundamental to so many algorithms (least squares regression; nearest neighbours; anything involving gradient descent). How do those algorithms behave over completely foreign metric spaces?

PROBLEM(S): I (well, the nonprofit I lead) am trying to solve for the general problem of municipal public policy but, as you might expect, it's a series of discrete, linked problems. Some of these problems have thousands of people focused on a solution; in other areas, it's virtually greenfield.

SOLUTION: The space doesn't lack for good research and policy recommendations, but it has historically lacked (on the right and left), non-screedy, nonpartisan voices that can be trusted when policymakers look for solutions. We're attempting to fill that space.

WHAT WE WANT: Ultimately? We want every major American city to work better for the people who live and work there. It'll look different from community to community and our job isn't about applying a cookie-cutter approach. Instead, we want to get a wider range of ready-to-implement tools in front of the decision makers, and educate engaged citizens that solutions exist.

A new Web.

Today's web is a collection of applications that largely provide a frontend for browsing data. The applications and data they contain are silos: there is no easy way to separate the data from functions and compute across datasets. Every application must (re)invent its own UI for querying and displaying data.

But if the web is actually a collection of datasets, why don't we have a web browser for consuming and interacting with arbitrary structured datasets?

We can model most popular sites (HN, Instagram, Twitter, Amazon etc) as a collection of hyperlinked JSON records. Let users adjust how these records are displayed. Provide a universal way to query and navigate any dataset and invoke associated functions (eg the upvote function for an HN post).

Full separation of data and functions instead of application silos is necessary to achieve general AI compute in the future.

Example: can you email Mark a summary of the top 5 most popular HN articles 3 days before our meeting?

Document software design 10x better.

The foundation is flowcharts, with support for individual layers distinguishing levels of abstraction, and scenarios for exploring use-cases. From there:

- Live data. We look at metrics on dashboards but it doesn't put into perspective how they relate to each other. Imagine seeing on your flowchart of servers, that one worker has an anomalous CPU reading, and you can click into that to see the individual readings of the running services on it. (rudimentary version: https://app.terrastruct.com/diagrams/1404897320)

- Automatic generation and sync of diagrams. Having access to sources like AWS account and version control to create and keep in sync diagrams of your infra, db schemas, UML classes, etc.

- Collaborative editing, seamless integrations with written documentation, linking directly to code where appropriate, version control, etc.

So much of software can be better understood visually. Still early on, if you're interested in learning more, https://terrastruct.com. And would love to chat (email in profile) with anyone with ideas.

Dreams.

I've been journaling my dreams for years and I'm working on an app that makes it easier to (visually) map them out & find patterns: https://oneironotes.com/

I like the idea of accessing other (inner) dimensions during sleep, like an explorer (an "oneironaut"). The problems to overcome are related to capturing and recollecting experiences that only take place in the mind. You asked about the weird stuff...

Human-computer context switching.

Which is why I started making Cleave: An application that lets users persist OS state as a "context" - saving and loading open applications, their windows, tabs, open files/documents and so on.

Started because of frequent multitasking heavy work with limited resources.

Made it because I wanted to switch between studying, working, reading, looking for an apartment, etc. without manually managing all states or consuming all resources.

Open Beta (macOS) as soon as I finish license verification and delta updates, but I keep getting sidetracked...

https://cleave.app

I'm trying to replace SQL by building a language that compiles to SQL. It has first-class functions, nicer syntax, better type-system, introspection, and other things you would expect from a modern language. But in the end, you still get SQL's performance, and the ability to use it with dozens of database engines.

Can a programming language based on Willard's Self-Verifying Theories [0] self-interpret in the limited sense established by Brown and Palsberg [1]?

Which is to say, what is the relation between a sub-Peano logical system that can prove itself consistent, and is complete, thus bypassing the 2nd Incompleteness Theorems, and self-referential decision problems posed in the programming language that corresponds to that logical system?

My suspicion is that such a language could provide very strong guarantees on the auditability of self-modifying code.

[0] https://en.m.wikipedia.org/wiki/Self-verifying_theories

[1] https://www.semanticscholar.org/paper/Breaking-through-the-n...

Trying to find a solution to the maximum-entropy probability distribution Q(x,y,z) constrained to reproduce the marginal distributions P(x,y), P(y,z), and P(z,x) from some other distribution P(x,y,z).

It is known that Q takes the form Q=a(x,y)b(y,z)c(z,x) for some functions a,b,c to be determined by solving the system of equations:

P(x,y) = sum_z Q(x,y,z)

P(y,z) = sum_x Q(x,y,z)

P(z,x) = sum_y Q(x,y,z)

It's not clear there exists a general closed-form solution. Iterative algorithms are known. This type of problem comes up in a number of interesting contexts. For instance, testing for non-trivial multi-variable interactions in dynamical systems such as neural networks or spin networks, performing joins on probabilistic databases, constructing reduced models of probability distributions, and in some cooperative game theory problems.

Examples: https://www.princeton.edu/~wbialek/our_papers/schneidman+al_...

http://vldb.org/conf/1987/P071.PDF

https://doi.org/10.6028/jres.072b.019

https://www.mdpi.com/1099-4300/16/4/2161 Edit: formatting

How do I satisfy the needs of non-technical middle managers in engineering to appear as if they understand and control their shops, while minimizing damage to those shops?

Information overload.

My current side project is https://feedsub.com. Right now, - it's not great - I started by building an simple tool for getting regular updates from RSS feeds, but longer-term I want to turn this into a system which can absorb all the data streams you're interested in (news, stocks, weather, social, communities) and give you dials (filters, curation, signals, etc..) to surface a healthy amount wherever you want (SMS, email, web, RSS, chatbots, etc..).

The crux of the problem is endless scrolling feeds we're sucked into 24/7, which is why I based my MVP on email.

My current solution is trivial on a technical level. Honestly, my biggest problem is thinking about the problem on a non-technical level, balancing this with working life and branding, since my software and vision are very far apart right now.

(TBH, this isn't nearly as hard a problem as some of the others here - but I enjoy the ideas/feedback I get from communities like HN)

Building a relational lang (https://github.com/Tablam/TablaM) I allow myself to get derailed in how provide a nicer "linq/iterator" protocol that work inside rust and outside in the lang (so, how I write them in rust is close to what the user could write in the lang).

The regular iterator protocol, as today in rust, make hard to do stuff like JOINS, GROUP BY and other fancy stuff (because you need to decompose the computation in a partial state machine. This is hard even for a developer, impossible to ask for a regular data user). Also, you need to duplicate all that for async (with streams) and other abstractions...

I'm now trying to understand transducers (https://www.reddit.com/r/rust/comments/gqiyej/potentials_adv...) and stumble upon effects:

http://mikeinnes.github.io/2020/06/12/transducers.html

that look to my the clean way I wish to use.

But what look easy in python/f#/etc in rust is HARD to do. So I'm in a kind of limbo :)

(It's a germ of an idea but I hope to work on this!)

Problem: Public schools in India don't do justice to students. Private schools in India charge a bomb but most of the money ends up in the hands of the "owners" and not enough to teachers (For reference, an average primary school teacher earns lesser than what an Uber driver earns) .

The solution : A network of "not-for-profit" schools where the fee structure is reasonable ( can't be free ), but the profits are shared amongst the people who make the schools run. Think "community banks" but for schools. I can't solve the problem for everyone but hope to set a good example by attracting the cream of teachers. It's time the teachers got their due.

Since I live in Tokyo, it’s about a 3 - 4 hour drive to various ski resorts. I want to snowboard every weekend. So I’m trying to find the best* ski resort I can head to next weekend.

Best is a mix of

- the ones I enjoy most

- but not one that I just visited last weekend

- where there has been recent snowfall / is predicted to snow

- that I have may have a season pass for

- where costs of highway tolls, petrol, hotels can be optimised

- driving time

- whether I’m going alone or with friends

- don’t have anything important at work on Monday

Eventually, I’d like to turn it into a kind of friend finder / social network thingy but for snowboarding.

Given the earth's population trajectory and the reduction in fertile arable unpolluted lands, there is a coming crisis in the distribution of food. How do we distribute food to high density Asian urban populations efficiently, minimizing needless motor vehicle trips, packaging and spoilage, when convenience purchasing is on the rise and average household sizes are shrinking? Our answer is a network of robotic service locations with automated stock-keeping and a shared, wholly owned logistics network plus personalized direct from fresh ingredient preparation.

Fully offline, fully searchable copies of personal data (email, tweets, calendar, etc.), English Wikipedia, IMDb, OpenStreetMap (tiles, routing, points of interest), geocoding. Fully offline and state of the art speech recognition. Fully offline voice assistant with almost complete coverage of the most common usage.

Bounds for the best possible designs for optical devices: well-studied [0, 1, 2, 3, 4], yet really hard.

More specifically, whenever you give a designer a design spec, it is always worth asking, how good is the best possible design for this spec? And, of course, can the designer actually achieve it, or something close to it? This is the question here.

In this scenario, the design spec is the optimization problem (what you want to optimize), the designer then gets to choose how to best approach this problem. In this case, you want to give a number that states, independent of how this problem is solved, what is the best any designer (no matter how smart or sophisticated, how much computational power they have, etc) can hope to do. In many cases giving such a number is actually possible! (See below references.)

-----

[0] https://pubs.acs.org/doi/10.1021/acsphotonics.9b00154 (PDF: http://web.stanford.edu/~boyd/papers/pdf/comp_imposs_res.pdf)

[1] https://arxiv.org/abs/2002.00521

[2] https://arxiv.org/abs/2003.00374

[3] https://arxiv.org/abs/2002.05644

[4] https://arxiv.org/abs/2001.11531

Representing musical and audio (sample) time in ways that maximises reversible conversion between the two of them. Musical time is typically spoken of (in western culture) in terms of "bars" (or "measures") and "beats". The relationship to audio (sample) time is defined by a "tempo map" which defines the number of beats per minute and the number of beats per bar.

The mapping between the two is monotonic, non-linear, and can be stationary. If the tempo map is allowed to contain ramps (accelerando and ritardando in music speak), there are implicitly exponential sections in the function that maps between them.

Using floating point arithmetic leads to errors that have immediate effects. A musical time that should be considered to be at sample N is instead considered to be at sample N-1 or N+1.

It's surprisingly hard to do it correctly.

Designing a combat robot for the UK Antweight division, which is only 150g max weight. (or 175g for some groups).

Despite this tight weight budget, I intend to build something rather interesting, but it is causing me to spend a lot of time in Fusion designing the parts along with slicing and reslicing 3D printed parts to shave partial grams of components to save a bit of weight.

Solving climate change of course!

I am partly obsessed with and researching / writing about how you could make carbon-free heating cheaper than fossil fuel based heating. If you can make geothermal systems cheaper than a natural gas furnace, then homeowners would have the same economic incentives as drivers, where operating an EVs is far cheaper and cleaner than ICEVs.

Private file access between computers. Even in 2020 it appears that building a connection between machines (macOS, Linux and Windows) requires a miracle if you don't like to share everything with cloud providers.

cryo file mangager is an attempt to get rid of the hassle: https://cryonet.io

Everyone seems to ignore the only critically significant hard problem: how do we mitigate the negative impacts that humans have had on the planet? Right now it looks like we are careening towards a quick extinction of the human race and most other species and we are not doing anything about it. True, the magnitude of the problem is daunting. True, it may be too late as we are past the tipping point. True, it is easier to ignore the risks that it is to confront them.

Maybe the real hard problem is getting people to pay attention to what is happening in their world and to forecast what the impact of their actions will be in the next few decades.

I am working on my own on a system that creates long lasting human relationships. It is a solution to loneliness.

I used to run match.com's research division. I left to pursue this project.

I want to give everyone a digital identity. In some countries (including mine) basically everyone has an e-ID which we use to sign in to things like government services, banks, payment providers and much more. This is absolutely essential to everyday life and many startups are built around it.

Unfortunately, many countries don't have useful e-IDs and the ones that do are limited to that one country. I want to create a single digital identity which works for everyone, for all applications, across borders. The basic features are:

- App based with no special hardware necessary.

- Privacy friendly with the user always fully aware of what data they are revealing.

- Simple to integrate for developers. It's a standard SSO flow over OAuth/OIDC.

I'm currently calling it Pass: https://getpass.app. If anyone wants to have a chat about digital identities you can reach me at fabian (at) flapplabs.se

Problem: A lot of professional jobs spend a shitload of time in front of the computer doing nothing else than googling or research on websites, taking notes (or transforming that knowledge to another medium or file) and repeating that all over the next day until they come up with a conclusion.

Idea: A web browser for self-automation that learns semantics and sentiment of content on the web, whilst trying to respect references or articles’ sources to recorrect the ground truth.

Solution: Still far away from it, but am implementing a peer to peer Web Browser that can share its information (or states) with trusted peers. Trying to implement a recordable, editable and repeatable GUI for everything, which is a lot harder than it sounds.

[1] https://github.com/cookiengineer/stealth

All kind of feedback appreciated!

Not tried yet but I’d love to make a JS front end library to “bring your own storage” so that if I provide you with an app online (think SPA) it can then save the data to where he used wants (s3, Dropbox, local disk using extension, etc.)

Another similar idea is a very simple rest protocol so that you can save to server and then make it easy to self host it.

I like the idea of people building apps as web pages without needing to worry about the server and the user owning their data but having the convenience of a cloud like solution where you just visit the site, log in and work.

Let’s say you have multiple people who want to share their location with each other, but to different degrees of precision. Some people might be willing to share their street address, while others only want to share the city that they live in. Some only want to say “Northern New Jersey”, or “Southern France”. Some might want to just share their timezone.

How do you store this information in a database, in a geographically meaningful way? How do you represent it on a map?

I’m working on a formula that factors ML model run-time cost (speed / memory-usage) into its performance evaluation. I generated a very simple scoring function from economic first principles. Now, I’m trying to test performance on cumbersome transformer models published by Google / FB / Microsoft etc in order to prove that many of these models are not “state-of-the-art” if run-time cost is taken into account.

Knowledge.

Working on building a self-hosted app that would alow you to save, organise and search your knowledge in one center.

It would contain information like notes, bookmarks (it would download the links contents) and in general provide a programmable, opensource interface to preserve the info you'll find useful and even sync with external apis to save your online presence locally (think reddit posts, hn links, etc...)

Stackoverflow doesnt have a good recommender system that can suggest questions to me that I can answer. I am trying to build one using their data from the data they post online.

I’m trying to figure out how to ship DNA effectively. It’s really inefficient right now, so I’m genetically engineering bacterial spores to make it a lot better. Not sure if people will adopt, but at least it’ll be 10x better than what is currently available!

I’m trying to map relational algebra onto Rust’s type system. If I’m successful, I’ll have a bunch of collection types with different performance characteristics that are all drop-in replacements for each other.

Trying to predict where electrons go in a molecule, but using a classical model. This can be done with supervised machine learning - you can use quantum mechanics to get lots of labeled data - but it's a tough problem, because chemical physicists have very high standards for accuracy.

We're writing too much code. So I'm making stuff that will help us write less code.

Like a vim-like editor that translates your spec into generated code while you also see it on the go to fix any issues. Think yeoman on generics and steroids.

I'm writing a book about ADHD: https://adhdpro.xyz/

Unfortunately there seems to be a lot of infantilizing in and around the ADHD sphere. Either we're treated like helpless children or we are encouraged to lower our expectations/goals.

It's hard to explain unless you've been open about having ADHD, or have been part of the community. The zeitgeist revolves around accepting and maintaining a status quo. The problem I'm trying to solve is how to build a growth/thriving lifestyle despite an ADHD diagnosis.

I'm building cad software for designing atomically precise objects.

I don't expect atomically precise manufacturing (at least in the form of molecular nanotechnology) to show up for at least a decade or two.

But the software to design things that can be built using nanotech can be written before the technology to synthesize them exists.

Online learning, particularly personalised learning aimed for self education and continuous learning. I'm trying to model the knowledge space as a graph, index learning material (free online resources right now then user generated content) on this graph and then provide different ways to navigate this space. I'm trying to address the "best way" for someone to learn a specific concept, and to help people identify the knowledge they miss/are looking for. The project would be open, collaborative and a non profit btw.

Over the past few months, I had somewhat as "writing a documentation for Google News RSS URL patterns" as my task.

There is no such official documentation.

Python package: https://github.com/kotartemiy/pygooglenews

Blog post: https://codarium.substack.com/p/reverse-engineering-google-n...

I've been tinkering in the reverse-engineering space. My problem amounts to reusing compiled binaries by combining them in novel ways. That is, I would like to take algorithm/subsystem X from software Y, combine it with something else. The goal is to have a library of components which I may be able to combine. It has lead me to investigate a few technologies I have been meaning to invest time into, like llvm, qemu. There are a few projects which combine these as well as related technologies like DECAF, radare2, DynamoRio, mcsema. The hard problem which I am facing, but by no means have an effective solution is of having to extract semantically the essence of the program in spite of ISA, and to find a balance between emulation and readapt-ability (i.e abstract out the code that is dependent upon some base-address assumption. The value on the surface seems counter intuitive to the investment, especially from my roots in SWE where one can have a hard enough time trying to accomplish that with source-code available. Although the application is broad, I've focussed intermittently on video games. I believe this is where some value lies, as a finely-tuned subsystem can be the heart of a franchise.

I am working on a system to parse the English language using a hand-written compiler and then store the IR in a database, so all human knowledge can be searched and queried in all its facets (who said what when and where) using English language. I believe that a database is the key to NLU, and machine learning is mostly useless for true NLU, because machine learning currently has no good way of interacting with a database AFAIK, and without the knowledge of a full database of human knowledge and knowing who said what it's impossible to truly understand human language. Storing all human knowledge in a neural network just isn't practical anytime soon.

I wrote a new programming language, Eek, because it was impossible for me to handle the complexity of doing all with current languages (that lack built-in support for asynchronous database access and parsing). So far the first generation of the programming language is working, but as an interpreter written in TypeScript, and I wrote an English language parser with it, and a simple database. Now I am working on a better, LLVM-based implementation of Eek. I started this thing about 3 years ago, and it will take some more years before this will be even demoable...

UDP DDOS mitigation. It'll be important for protocols like HTTP3/QUIC where decrypting every packet is costly, and malicious packets can't be identified until after the application tries to decrypt them.

The idea is to assign a random 8 or 16 byte number (DdosID) to each connection. The DdosID is unrelated to any other identifier; like the HTTP3 connection id. The client puts the DdosID at the beginning of every UDP packets data section (raw; no encryption or compression.) Packets that have a valid DdosID are processed. Packets with an invalid DdosID are dropped. New connections use zero as the DdosID. Any client can use zero to re-establish a connection or if something goes wrong.

If attackers make random DdosIDs it will result in packets that are dropped before decryption. If a valid DdosID (or hundreds of valid DdosIDs) are used, the volume of packets with that DdosID will make the attack obvious, and the DdosID can be invalidated and the packets dropped.

Attackers will need to spam new connection attempts (DdosID zero) to deny service. New connection attempts could be dropped, or a small percentage allowed when the application can handle them. The rate of allowed new connection attempts could be adjusted very easily and quickly, but a high rate of bad packets would still deny new connections. Existing connections to continue to work.

The challenges are mostly creating functional thresholds: what packet volume is too high? how long is a DdosID valid? Should a load balancer record the IP and Port and to force a new DdosID if it changes?

I would also like to see a way that applications and load balancers can use a DdosID system without it being dependent on the specific application; DdosID should not depend on the protocol it's protecting.

## Problem

Half of the Indian farmers do not have access to institutional/formal credit. They end of borrowing from local loan sharks at extremely high interest rates. The vicious cycle continues as they do have access to markets favoring their produce's selling price, and end up being exploited.

Most banks, despite their mandated percentage of loan for agriculture, are not too keen (and unable) to work closely with farmers.

## Solution

By leveraging technology, and international connections, we help farmers by giving them access to credit at a favorable interest rate. We work with international community, such as the Japanese, to lend their capital on our platform, thus helping the underserved farmers. In turn, the Japanese investors get returns much higher their national savings interest rates.

Banks too can benefit from our ability to deploy their cash.

## Why Us, Why Now

I have struggled to articulate the philosophy/motto on how to help the farmers, as they are the most exploited populace. I'm zeroing in on -- "BE KIND".

If you want to hear more, contact me and I will send you our Executive Summary and/or Pitch Deck. We are fund-raising.

I'm trying to do a historiographical quality history of computers, the kind of scholarship you might see with say the French Revolution or the Roman Empire.

What more, my history efforts ENDS around 1937; it's the BC part of computing that set the stage. The build up to computing.

Episode 0 lays it all out (http://comphistpod.com/introducing-the-computer-history-podc...)

I'm doing it in podcast form because it's 2020. This is going to be multiple years and I'm totally fine with that.

I divide the history up into multiple facets each with separate timelines.

Currently I'm going all the way through "electric communications" from the electric spark up through relay networks talking about switching, encoding, error correcting, signalling, all the important developments along the way.

In 2021 I'll close that out and do the same for storage (music boxes, looms, etc) and computation, starting with clocks, pascal's mechanical calculator and going forward from there, mechanical registers, overflow, adders, etc...

Each one is going to take at least a year or so.

I'm already about 2 months in to recording, about 6 months into the project. This week will be du fay, boze, desaguliers, watson, and the electric wire. Next week will be leyden jars, this is a long, slow project, and AFAIK it has never been done before.

I'm doing the odd episodes as a timeline and the even episodes as diversions and discussions to keep things entertaining and light.

This is the first time I'm talking about it publicly. It's at http://comphistpod.com

Simply put, earning a living. I didn't travel the conventional path of high school > university > job, and so I'm struggling to get my shit together, the concept of earning a living (outside of making minimum wage) is so foreign to me that I really don't know how I'm going to do it in a reasonable (< 5 years) amount of time.

I’m working on a way to design free standing domes using mortar-less bricks. The idea is to build a cut-list for each brick, such that when stacked, there are no gaps, and the entire thing is held in place with just gravity (maybe the first layer is permanently attached to the ground)

I’m looking for a good framework to simulate the physics and visualizations.

I am attempting to bring 100% client-side speech recognition to the web:

https://github.com/jaxcore/bumblebee

Although the first release is not officially out yet, the NodeJS code is working and you can install the development version of the app server and try out the hello world app locally.

The solution involves running Mozilla DeepSpeech inside an Electron desktop application with a websocket server and client API that NodeJS scripts can interact with, to receive speech recognition results, utilize "alexa" style hotword commands, and text-to-speech. The electron app handles all the heavy stuff, and you just use a simple API.

A web browser extension can also make use of this API to bring these capabilities to web sites, but that part isn't finished yet.

JHipster style templating for SAAS onboarding design patterns.

UX description language for forms that respects high level constraints. Compiles to desktop browser, phone browser, and Alexa layouts.

Solving the complexity of matrix matrix multiplication by brute forcing the lower bound with semigroup combinatorics.

DSL for linear logic.

Ending Iowa’s criminalization of “annoying” speech. (Iowa Code 708.7)

Exposing Polk County Iowa Sheriff Kevin Schneider torturing inmates with denial of basic medical care.

Exposing pure nepotism corruption between Iowa Attorney General Chief of Staff Eric Tabor and his sister Iowa Court of Appeals Judge Mary Tabor (mom of @ollie).

Exposing that the prosecutor on Tracy Richter’s murder trial had relations and ended up marrying the daughter of his star witness Mary Higgins - and that the blood spatter expert Englert is a known fraud who wrongfully convicted David Camm and Julie Ray Harper.

I will be honest. I really don't get it. How come I am working harder, and more hours, and studying more but I don't earn more ? - It should have been a direct correlation. Shouldn't technology and automation and all of this good stuff make us work less but earn more ?

Loneliness! We have really lost our sense of community with those around us so I’m trying to see if we could supplement that with strangers on the internet.

Getting time to do anything in between kids, work and whatever else is expected from me is the actual hard part for me.

The technical aspect is mostly fun in my case :-)

Making leaking safe(er).

After the twitter account of DDOSecrets got shut down (due Blueleaks), this got me thinking: How would you leak / provide data but are not directly attributable. (At least like a retweet - not your tweet, just amplified.).

And how to add some resilience and protection to the distribution, since there were indicators that the torrent and download of leaked data was being attacked.

So far, I have come up with an encryption matryoshka: you distribute leaks without telling whats in and enable gradually a few to look inside until they are public.

All thats missing is a better document describing it and a command line tool to help to walk through the multi level encryption ... so there is 90% still to do ¯\_(ツ)_/¯.

Table balancing in an online poker tournament.

It’s not that simple.

Keep tables balanced, move players from high numbered tables to lowest numbered tables, try to move players to similar positions, Mark players waiting if they drop in small blind and skip the button next hand.

If you repeatedly add 2 and then divide by 2, you approach the number 2.

If you repeatedly flip a coin to either add 2 or divide by 2, what distribution do you approach?

If you have a distribution, what random processes are identities for that distribution? Which are stable? Which will reproduce the distribution from any starting distribution?

It’s related to matrix kernels, but it’s hard to generalize to continuous numbers. I spent a while looking into it years ago and couldn’t find much.

My goal was to eventually take it a step farther and create simple stochastic functions for certain behaviors, such as bistable switches, memory registers, etc.

Working on siuba, a data analysis tool for python. It's a port of the R library dplyr, and can produces SQL queries!

I've programmed in python for much longer than R, and really want to be able to move at the same speed when using python for data analysis :o.

It's a weird problem though because the two languages have basically opposite approaches to DataFrames. pandas has a very fat DataFrame implementation, R an extremely minimal one. (Pros and cons to both approaches).

https://github.com/machow/siuba

Trying to learn enough to build a table-top DNA sequencer with a Raspberry Pi

Procrastination. But I'll probably start working on it tomorrow.

I’m trying to solve the hard problem of finding a new job since COVID-19. It’s not going very well :-D

Upscaling videos.

This is a hard problem. It may not even be perfectly solvable at any time in the future.

When you increase the size of any kind of raster image, you're creating new information from the old.

There's been some pretty good approaches out there, like this [1][2]. It uses a GAN topology for some impressive results, but is incredibly memory intensive and can take a very long time to run.

I've been working on something [0] for a good long while which is a less expensive approach. Instead of attempting to replicate everything 1:1, it intentionally allows some detail loss, whilst attempting to preserve everything important.

It's not ready for the public, and video still needs some significant improvements to remove some of the artifacts. But I've released one TV series upscaled with it [4], thus far.

But as it stands, at x2 and x2.5 scales, it does pretty well, with the average person preferring it to most other resizing methods. It doesn't reach the GAN-approaches quality, but you're looking at an average 12 seconds for upscaling, versus 80 seconds for the GAN approach for the same size upscaling and what people perceive as the same quality.

It already beats most of the traditional resizing algorithms pretty soundly. [3]

[0] https://git.sr.ht/~shakna/upscaler_analysis

[1] https://developer.ibm.com/technologies/artificial-intelligen...

[2] https://arxiv.org/pdf/1609.04802.pdf

[3] https://git.sr.ht/~shakna/upscaler_analysis/blob/master/crus...

[4] https://sixteenmm.org/blog/20200626-Gunsmith%20Hits%20HD

Trying to find the next entry in this list of functional Busy Beaver numbers

https://oeis.org/A333479

Telling where the output of an individual stdout print statement ends and the next one begins, so that I could color code my terminal with alternating colors to more easily tell apart individual log messages from a running process by visual differentiation. Turns out this is impossible! Aside from time passing in between outputs, there's no way to tell. It's just a continuous byte stream with no terminating character or pattern.

Developing better, more humane environments for hanging out and spending time online:

https://www.producthunt.com/posts/cozyroom

And figuring out how to make it easier to write and develop ideas through blogs:

https://jborichevskiy.com/plan/#jborichevskiy-com

Would Fremen Stillsuits like in the Dune novels actually be possible?

These suits allow to survive for weeks out in the deep desert, by catching and recycling all of the body's lost water. Making sure no perspiration can escape is doable. Filtering the sweat to produce clean, salt-free water should be possible as well - membranes for water desalination do already exist today. To have all the required pumping action provided through walking and breathing is a mechanical problem that should be theoretically solvable. The big, unsolved problem I see, is heat.

The book says that the suit's layers closest to the skin allow the sweat to evaporate and thus provide cooling to the body. But the water then has to condensate again somewhere. From my (limited) understanding of the laws of thermodynamics, the amount of extra heat created through condensation, should be exactly equal to the amount of cooling the evaporation provides making the whole cycle a zero-sum affair. But for this cycle to work in the first place, the skin would have to be of higher temperature then the layer where the condensation occurs. If the desert heat is above body temperature, we'd need some sort of heat pump like in a fridge. Using changes in pressure and density (through a compressor) you could cool the suits' inside below body temperature, while heating the outside above ambient temperature - which is necessary to actually give off heat to the outside.

Those compressors are heavy and powerhungry though, and you'd need additional high pressure water lines through the suit - increasing the bulk of the whole thing considerably. Future technology might be more miniaturized and more energy efficient, but still... Maybe piezoelectric/thermoelectric cooling (exploiting the Peltier effect) would be a better choice for a suit like this. Then you'd have a light inner layer that allows for air circulation - so that the sweat can evaporate on the skin, and condensate again at the thermoelectrically cooled middle layer - where it's collected and pumped away to the membrane filters and catchpockets.

The outer layer of the suits would have to be made of flexible solar cells, in order to provide the electricy required for the cooling of the middle layer. Not sure if that could work though. Those solar cells produce a lot of heat on their own, and they sit right on top of the heat-producing side of the thermoelectric cooling. And the sun burning down on it as well - that's a lot of heat right there at the outer layer. I don't think thermoelectric cooling can overcome that high a temperature difference. It works best, when the temperature difference between the cool side and warm side is pretty small.

How to measure pedal (and/or crank) angle on a moving bicycle to 0.05 degree. Anyone?

Making people aware they should test their drugs. MDMA testing is illegal in lots of places, we ship test kits where available.

And I created a reverse image search to look up pill data: https://play.google.com/store/apps/details?id=be.harmreducti...

1: trying to build tools to give people a more tacit understanding of sattelites & space debris: https://space-search.io

2: developing easy-to-use tools for parametric generative design to enable hyper-personalisation: https://hyperobjects.design

Local solutions to stop humans from destroying their environment.

We've implemented recycling maximally and created programs for reuse, repair, toxic waste, general waste reduction, battery disposal, returnables, etc. But it isn't really making a significant impact.

People still spend most of their disposable income buying things they don't need that can't be easily repaired, or that they can't bother repairing, or that they discard due to fashion, or because they are mostly packaging, etc.

We can't get them to stop buying useless stuff that they don't need. We can't stop them from buying new cars every few years.

We can't get them to spend money on quality infrastructure, like insulation, that would reduce their energy needs by about 80%.

We can't get them to stop going to restaurants or getting food delivery or take-out, which expends a multiple of the energy and greenhouse gases of cooking your own locally-sourced food.

We can't get them to understand that we're all going to die in fairly short order unless we bring the environmental disaster under control.

Decomposition of glyph sequences in phonetic transcription alphabets (e.g. IPA representations of phonemes) into phonological feature sets.

Existing attempts to solve this problem are hackish and difficult to customize: they typically treat each glyph as a set of features and handle diacritics and digraphs by naive composition and awkward special-casing. They also aren't written with an eye to customization in either alphabet or featural model: they typically map an ad-hoc extension of IPA to an ad-hoc featural model.

I think a natural improvement would be to develop a specification language in which each individual glyph (base character or diacritic) is a function from a feature set to a feature set, with Haskell-style pattern-matching to allow graceful handling of digraphs and context-sensitive diacritics - although syntactic sugar for digraphs would in practice be required for usability. Ideally it would also be possible to map feature sets to feature sets, in order to preserve a human-readable intermediate form (e.g. "unvoiced dental plosive") which is later mapped to the more customary binary features.

In addition to the utility for phonological databases and the like, this would also enable more rigorous testing of crosslinguistic feature sets: every feature set is implicitly a set of proposed linguistic universals and existence claims. If two segments have the same featuralization, they should never contrast in a given language; if they do, the featuralization is unsound. And if a featuralization proposes the existence of many contrasts that aren't attested anywhere, it could probably stand to be optimized.

But most of my interest in this comes from my work on a phonological database. The database needs some method of handling featuralization to facilitate feature-based search, and I just haven't seen a good way to do that yet.

Trying to figure out what the next version of the internet would look like and start building it :p ps. Structured data and no blockchain :)

I’ve been coming up with a new pedagogy for the fundamentals of interactive and computer art, and trying it out with kids on TikTok. Maybe you saw the software I put out a few weeks ago on HN called No Paint: https://nopaint.art. However, most of the time I’m using pen and paper for this work.

A prediction/promise competition to increase accountability in general.

Politicians/companies spew words and we tend to accept them bc they're "authorities." I believe the best way to increase "skin in the game", accountability, and humble expertise is to predict and have your predictive performance be visible.

There's prediction markets to trade real money with but none that I've seen that ranks your performance against others.

I'm building a platform where it's free to enter/submit predictions in categories of interest. Top ranked players will receive prizes.

I think once players can measure themselves against "authorities" (whose public predictions will be scraped), both will become more accountable.

After predictions, I aim to work on "promises" since they both increase future skin in the game.

If you've read this far, you should consider joining :) - https://oraclerank.com

Keeping a working Wireguard configuration with 4 peers, for both VPN-through-house, and remote LAN access.

I'm trying to run jobs on a timer when the jobs are on a cluster and I don't know which server the job is assigned to.

It's essentially a problem of distributed timers and distributed transactions.

(If anyone has any resources on how similar problems have been solved in the past, I'd appreciate it.)

My short term thing is this mathematical programming puzzel https://projecteuler.net/problem=160 from Project Euler where you need to calculate a very large factorial.

My live goal is to bring back more programming and design techniques that we as a society used to make the best ps1 and gameboy games and old school animations to the way we are currently developing games. (Any tips and advice how i could help improve the game development industry is welcome. Next year I will be doing my masters so i'm also still looking for a subject for that. The last couple of months I have been experimenting with water color effects in openGL, so I'm looking for something like that)

I'm trying to tackle the problem of information overload by providing a service that extracts, summarises and curates the key insights from books, articles and research: https://www.sivv.io/

1: How to download the internet. 2: How to parse html in a way that gives it consistent structure and meaningful interpretations. 3: Comvine 1 and 2 and make it searchable.

I’m about 12 years into it and expect I have a couple/few more to go. Hopefully people still use HTML by then.

One-click API data ingestion.

Just about every company these days has their data spread out all over the cloud: marketing data on Facebook and Google, social media data on Twitter and Snapchat, customer data on Salesforce, sales data on Shopify and Amazon, and so on. Most companies will either (a) hire a team of data engineers to collect and exploit this data or; (b) hire an expensive consulting firm to build an ETL pipeline, or; (c) let this data rot in the cloud. For the past 6 years, I've worked as a data engineer (where I became intimately familiar with Facebook and Salesforce APIs), and I'm confident that I can automate around 80% of my job.

It's clear that the value prop is astronomical: just one data engineer will run you at least 150k/yr and most of the work will involve maintaining API data pipelines. Having a "one-click" solution where one simply provides an API key and what data they'd like to warehouse (e.g. marketing data, social media data, customer data) and where (FTP, S3, Redshift, DynamoDB) would be invaluable to companies that want to make sure they exploit this treasure trove.

Some hard/interesting problems:

- API specs constantly change (Facebook, for example, has a quarterly update schedule)

- Inferring JSON schemas is hard

- Data integrity is hard (data types sometimes change willy-nilly)

- API rate limiting is tricky

- Resilience is hard

- Recovering old data (especially for certain services) might be impossible

Everyone is starting to become keenly aware that letting the data rot is starting to have a higher and higher opportunity cost. Not warehousing your own data is simply not a tenable option any more: the world’s most valuable resource is no longer oil, but data[1].

[1] https://www.economist.com/leaders/2017/05/06/the-worlds-most...

I'm working on a website that helps me to keep track of the podcasts that I'm listening to across platforms (Think Last.fm / Trakt.tv but for podcasts) by automatically importing the listening data from various podcast apps.

detecting the tempo of a rhythm a drummer is playing in real time (including tracking variations/drift in tempo) based on the time stamps of each drum hit (and starting at an assumed tempo). I've found in depth resource on problems that are similar, but not the same (wave form data rather than time stamps, all-at-once rather than real time, etc). I'm trying to keep myself open to the fact that the solution might be incredibly simple, and just unrelated to any path I've gone down, but it's led me down some interesting paths that I'm enjoying, so also just taking that for what it's worth

Working on an algorithm for recovering the text you type just by analyzing the keyboard sound captured through the computer's mic (i.e. acoustic eavesdropping). The hard part is doing it without having train data for the keyboard.

I'm trying to build a storefront experience for building customized home improvement services. These services can be anything between a new set of curtains, new tiling for your kitchen or a false-ceiling throughout your home. This is difficult because these services are dynamic in nature, and as such we cannot sell fixed SKUs like normal e-commerce services do.

We are trying to develop a unified service-building experience wherein the user will be able to punch in their requirements and get a product tailor-made for them, along with the estimated price (which comes with ~10% tolerance). It's tough, but we're getting there.

Verified location using GPS.

Problem: Currently, users (or attackers) can easily manipulate the location provided to an app on a phone.

Solution: Use raw measurements from positioning sattellites to check if the location reported by a user actually lines up with the measurements of their phone.

Why is it hard? - Lack of documentation, standardization and support on collecting raw measurements on phones. - Processing raw measurements is tricky - Finding anomalies in this raw data is even harder

Some of it is working - yay! - and there's also a public API, such that others can use it too: https://claimr.tools

Figuring out how to fill the gap between code search and static analysis (code checks.)

Right now the tools we have for programmatically reading through code are: 1. Code search, which is fast, but inaccurate/heuristic. 2. Static analysis, which is slow to run and difficult to write, but very accurate.

I'm building a tool that is as fast and easy to use as code search, and is as accurate and expressive as static analysis.

Still just a landing page [0]. Looking to get a public playground people can mess around with this week.

[0] https://sourcescape.io/

ECS based (https://en.wikipedia.org/wiki/Entity_component_system) GPU-first GUI framework.

A piece of music playing software that can react to the lead of an instrument. I'm picturing the software will be able to play a concerto (or simply a duet) with a real instrumentalist and react to dynamic and tempo changes in real time, like an orchestra under a conductor.

The same principle can also be used to create a real-time software harmonizer [1] for live performances, but this problem already has a reliable solution through hardware.

[1] - https://www.youtube.com/watch?v=DnpVAyPjxDA

I am trying to build a humanoid with bunch of servos. Back in 2017, I bought 205 servos from a closeout sale.

Currently, I am trying to hook them together and come up with basic APIs to control them. Next is printing a 3D model.

I'm trying to build a program to optimally schedule time to work on your tasks based on your schedule.

I have a bunch of tasks I have to complete by a certain deadline — these include things like engineering sprint tasks, drafting a design document, completing an assignment, etc. I have to get these tasks done in between my regular schedule of meetings, lunch breaks, and rests. I want to get a program to tell me when I should work on what depending on a task's due date and priority. If something comes up, I want my schedule readjust to accommodate the interruption.

Maybe not weird but certainly hard. I’m working on a way to build better habits. I know there’s a million apps out there that present the typical log a habit behaviour however I see this as mostly punishing.

A positive streak can easily turn into a negative streak.

At the moment I’m brainstorming with a google sheet and a few manual tweaks. I believe there’s a way to help people/myself build a better life by removing the nasty things (like smoking) and grow the nicer things like healthy eating or exercise. But our desire for instant gratification and our lofty goals get in the way.

An open source search index shared and updated over the bittorrent protocol.

Related to the future time travel. You somehow get thrown into the past, say 100-150 years. How do you put out a message proving you're from the future and requesting time travelers to come back and rescue you?

My proposal is to take out classified ads in newspapers, couching it as a code or puzzle to be solved, and put in dates of major future events you know about. D-day, Kennedy assassination, Challenger explosion, 9/11. Top it off with Murder Hornets and they'll know about when you came from.

I believe that computation, mathematics, information and semantics all share the same set of simple foundations. And that the means to understanding these foundation is to look at how physical computational processes use information from their environment, and produce information that is used to make "real world" outcomes happen.

I am working on this at the moment, and no, I don't expect that what I write here will convince anyone. And I don't have any summaries of the work at the moment.

I'm developing a web-based system which supports every browser in existence.

So far, I have full baseline feature support for everything back to Mosaic, including Lynx, IE3, Netscape3, Opera3, and many others.

At the same time, still including advanced features like client-side PGP for browsers which will support it.

Every browser presents its own challenges, and it is not always the oldest ones which have the dumbest behaviors.

My intent is to promote interoperability and offer something as an alternative to today's near-monoculture.

Trying to find a backend (php) freelance project since March after having worked 5 years in a startup/scaleup. Its the hardest problem I faced in ages. ¯\_(ツ)_/¯

I built out a sequence where you take some large integer n, and a smaller positive integer x, the first number in the sequence, such that the next number in the sequence, x_1, is the remainder plus the dividend of n/x, x_2 is the remainder plus the dividend of n/x_1, and so on until you reach a cycle. The problem I'm trying to solve is given some n and some x can you give at least two solutions for x_-1 (one number previous in the sequence)?

I'm doing the hardware part of a CB packet radio infrastructure that can be deployed quickly after a disaster, and allows basic BBS functionality in addition to passing messages, and works with existing cell phones.

It's inspired by the CellSol network in the "Left Behind" novels.

Schems and code are at https://www.aaronswartzday.org/lora/ and we could use help.

The hardest problems to solve are deliberate problems. Where solving it results is large financial loss for the people that profit from the problem. For example: https://www.youtube.com/watch?v=eDSDdwN2Xcg

Getting discussions about these issues to include financial incentives is a weird and hard problem. People are too sure the players (on their team) are alturistic.

Trying to create Discourse like E-commerce app that will make it easy for people to sell anything

It will have all features that WooCommerce have but much stable and easily customisable.