malthejorgensen

Don’t you have to manually “refresh” Postgres materialized views, essentially making it an easier to implement cache (the Redis example in the blog post) rather than the type always-auto-updating materialized view the blog post author is actually touting?

In the ancient times there was a startup called The Grid. Very hyped. They implemented the Cassowary layout algorithm for the web and called it GSS.

I loved the idea of GSS but it never caught on: https://gss.github.io/

sourcehut isn’t weird at all.

It’s made by Drew Devault who is mostly well-respected in the hacker community, and it’s made exactly to be an alternative to BigCo-owned source hosts like GitHub, Gitlab and Bitbucket.

Feels similar to `sd` (https://github.com/chmln/sd)

which in my mind was the first “replace” version of ripgrep

grep -> ripgrep

sed -> sd

The argument made here is like “RIP C” now come over to my new project: “C++”.

Nope. Nope, nope, nope. Neither C nor Jekyll are dead.

I’d imagine it’s more like an adjacency list structure with various indexes (similar to a regular relational dbs) to allows lookups based on node properties

My impression is that most editors already use this or similar optimized data structures for representing the text internally.

Headline is a bit rich, since duckduckgo exists (and Bing/Microsoft if you really stretch it).

I get the point though -- duckduckgo doesn't provide a browser. But I'm guessing Brave doesn't provide hosted email service (Gmail), file and document hosting (Google Drive and Docs) so the analogy breaks down either way.

That snark being said, I do want more privacy on the web -- so yay Brave!

Totally agree – KaTeX is good enough. Browser implementors should spend their time on other things.

The first part rings dangerously close to solipsism in my ears, and once you’re there nothing much matters. Just trash the planet, use people and move on. Cause once you’re dead it doesn’t matter? I might be wrong but you can’t build a civilization in that way.

Second thing is that immortality creates conservatism. Old age does too, but it seems to me immortality over-indexes self-preservation over progress of civilization, where the former is just slightly less of a concern for the individual when lifetimes are as short as they are now.

Colonialism still exists in the form of lopsided trade economics.

Most of Africa is not able to trade freely with EU, creating a favorable economic situation for the EU. I'm not sure the milk example is a good one, but historically (post-colonialism) most value gained from mining in Africa has been gained by companies in the EU -- not much value gained for the African countries and citizens. This is due to the lopsided trade agreements but also corporate exploitation of the African countries.

The African countries needs our help in the form of proper free-trade agreements, and in the form of us not meddling with their natural resources.

Source: https://en.wikipedia.org/wiki/List_of_bilateral_free-trade_a...

As an early user -- the speedups are very real! Feels great to just skip build step after build step, due to Docker-like layers.

We got hit by this as well. Very similar story to this and others shared in this thread: Use an S3 bucket for user uploads - and Google then marks the bucket as unsafe. In our case a user had clicked “Save link as...” on a Google Drive file. This saves an HTML file with the Google login page in some cases (since downloading the file requires you to be logged in). The user then proceeded to upload that HTML file. Then it was automatically marked since it looked like we were phishing the Google login page.

It should be noted that Firefox uses the Google banlist as well so switching browsers does not work!

Given the opinion expressed in this blog post I wonder how the author intends to process data that they don’t know the type of :P

Yes, Haskell is born out of academia. That’s where its strengths come from but also its weaknesses. Many new languages and language features are directly inspired by Haskell’s strong typing and type safety and the world of programming languages is better for it.

I do question whether Haskell will ever be ready for mainstream - probably not. But hopefully some other more practical ML derivative will gain enough momentum to truly become mainstream. Arguably Scala is that language, but Scala feels the C++ of ML-derivatives - the language is too big and the syntax irks me.

How do you manage the bare metal cluster? (E.g. apt/yum updates but also networking and such)

Thank you so much for this reply!

Cloudflare Workers as a way of bypassing whitelisted Cloudflare IPs makes a lot of sense, and of course is an attack vector that needs to be mitigated. Now I'm really dreaming – but it would be really neat to have a mechanism similar to LetsEncrypt where you could prove ownership of a domain and allow arbitrary requests to that domain. Given the potential throughput of Cloudflare Workers, it could be valuable as a safeguard even in the current setup (I'm guessing it's possible to amplify one request quite a lot which could annoy a victim server even with if the requests have the wrong `Host`-header?).

You're partially right about the CNAME. The reason I ended up needing workers is to preserve the cookie check that I currently have (implemented in NGINX/Lua) which redirects to app.peergrade.io if a session cookie is set (the cookie is scoped to the top-level domain). Not the neatest, but it used be quite common -- Heroku still does something similar.

I didn't think about Node not being as set-in-stone-standard as The Web Platform and not being sandboxed. An example of where the current setup is awkward would be using `atob` and `btoa` with actual binary data, I had this case when unpacking base64-data in the cookie.

We're building Eduflow (https://www.eduflow.com) – a light-weight LMS born out of some of the same frustration aired in this thread: LMSs are to enterprisey and seem to be designed with somebody else than the actual end-user in mind.

(Plus, we were in YC batch S17 with the predecessor to Eduflow – Peergrade)

This.

IMO he's still a "rockstar". It's just the Node and JS community that hypes everything disproportionally (they used to at least). The fact that he single-handedly built Apex (https://github.com/apex/apex) show that he's still prolific, and a programmer of note.

There's a similar story for Sindre Sorhus, who moved on from the JS community to Swift.

Thank you!

First customers were through personal network, we then quickly transitioned into cold emails to universities in our region.

We are already making money, primarily through sales to institutions in higher-ed. The longer term plan is continuing down that path, with institution sales making the majority of revenue, but with individual teacher plans available as well. The freemium model is to get teachers hooked, and then have their school, department, or university purchase as they see increased use and a need for integration with their existing school system (LMS). This is a path we have seen for a bunch of our current customers. Especially integrations seems to drive institutions to buy.

K-12 vs Higher-ed is a complex beast, but an analogy can be drawn to Consumer vs B2B. K-12 is high-volume, low price, low touch deals and have some similar characteristics to the consumer market (virality + simple products). Higher-ed is lower-volume, larger deals sizes, but more hands-on sales. Currently our sales and marketing efforts are focused on higher-ed because it does seem to have the best ROI.

Hi Jakob – thank you – we're thrilled to hear that!

We, at Peergrade, strongly believe that there's a lot to be gained if students share more of their experience (knowledge gained, questions, struggles) when engaging with the class material. Through sharing those learning experiences students will learn faster and our aspiration is that Peergrade will be the facilitator of exactly that sharing and improved learning.