graiz

Keep in mind that at the time Microsoft was being sued for being anti-competitive behaviour by Netscape and there were a lot of attention on the company for being a bad actor. Most of Microsofts's patents were defensive (they rarely used them offensively to my knowledge). There was a group of execs that were nervous about Linux but the lack of UI/UX/Usability strength put most of the focus on server capabilities. To this day many people want Linux to be a broader desktop OS but the only strong and focused effort on this are ironically ChromeOS and Android.

Windows Media team made a lot of bone-headed decisions. This was at a time when WinAmp was taking off. WinowsMedia could have leaned into MP3 culture while giving a very basic nod to DRM, instead they created a bunch of confusion. At the time Microsoft was considered the "evil empire" so being considered not one of the good guys would have made the marketing team proud at the time.

I was a former program manager on the Windows XP release so it warms my sprit when I see projects like this. Thanks for keeping the dream alive.

GPT3 is a really impressive auto-complete. It takes inputs and predicts what text should be output. It's super impressive and it looks like it's smart but it is not running code, it's not Turing complete and if you understand how it works it's very easy to cause it to produce significant errors.

Bluetooth is so broken at a deep and fundamental level. We should really have a low-energy wifi protocol. The listed problems are just the tip of the iceberg.

The fact that my headphones can't be simultaneously connected to both my phone and my computer is ridiculous. The connectivity is awful. The latency of connection and disconnection is embarrassing. On top of that Apple has the worst UI to handle all of of the weird things that can happen.

Tracking performance relative to ETH and BTC seems flawed to me. Tokens/Coins can have investment properties that are more or less speculative so the relative price to other cryptos may be irrelevant. As an example FileCoin is a storage token and more of a utility coin rather than a value store.

I think the point around conflicts of interest is sound. A16Z should probably step down from the board.

This wouldn't help Amazon as much because they are obsessive about performance. An out-of-the-box wordpress site loads MB's for anything basic as do most of the top-visited sites on the web.

In the developing world they are using the Internet at dial-up speed, only on mobile, and often charged per KB.

Accessibility is easier because items have clear locations and sizes without computing cascading rules. This also allows for easier navigation (left/right) in addition to traditional logical / tab-stop order. If done right it would be harder to screw-up. Current HTML is very easy to inadvertently hurt accessibility.

In my tests it would be very difficult to create a 1Mb page. You would need 10,000 text/image/nodes of actual content (not tags or containers). This just doesn't happen.

Presentation is strictly separated from code execution so first load is never blocked by code. Code can dynamically change the page but does so by directly changing the DOM using limited server side-commands. In other words it's truly a thin client browser. You can bloat up the server with as much code as you want but the client side stays lean.

SVG is still a markup language, not one that is compiled or streamable. It's also not relational. While you can design a page with SVG it's not parametric in that you can't modify the content or the page and have it re-flow, resize or act responsivly.

I would also describe SVG as a very verbose instruction set rather than a reduced instruction set.

I've thought about this a lot. I'd throw away HTML/CSS and start over with a client centric rendering protocol that is based on presentation first and semantics second. The language would be run-time compiled to describe exactly what needs to be rendered on the page and would be streamable to prevent rendering locks.

Typical page size would go from 1M typical to 64K-128K being typical. Images would stream in after initial page renders but since most pages would fit in 1-3 packets, you'd see pages pop-in very quickly. This would also be very helpful for poor connection, mobile and the developing world.

I'd fund a team to do this if I could figure out who would buy it.

RSS is really based on polling and that's not great. It also doesn't provide a lot/any analytics back to the content site to give attribution of content readership.

I think the idea is super interesting but it's not surprising that Google/Facebook and other ad networks made it less relevant.

One of the best developers I know has a serious visual impairment. Don't say "I can't." Instead ask, "how can I?"

Ironically this was linked to with the .htm extension link but it was nice to discover that the non-page specific URI also worked on this site: https://www.w3.org/Provider/Style/URI

Tesla is a great at many things but managing customer/media expectations isn't among them. I'm sure Elon has the data and will tweet why vision is better, he may even back it up with really good data, but it's reactionary media management rather than preventative.

How hard would it have been to write a press release explaining why they are removing radar, what it means to safety, and how it impacts the future of self-driving?

I think it's awesome that Sam and Co. are doing this. I can imagine variations of GPT predicting bio-tech, medicine, human relationships, environment, politics and more. Many of the companies tapping into GPT3 are just scratching the surface. I love that they are funding some big bets.

Matt - You built an amazing company. So few entrepreneurs really have a long-term vision. Kudos and excited to see the next 10 years brings.

I want a future without passwords, but that future gives me the choice of third parties to host my passwords. I prefer 1Password, some people like iCloud, while others may prefer a Microsoft solution.

Passwords suck and we need a per-site password policy that can act like an API. Kind of like a Robots.txt, to declare, "This site needs 8-20 characters, 1 symbol and the URL's for login, reset and forgot password are these URI's."

If the product is too expensive they should sell it with an optional payment plan with Affirm or a similar company that acts like a credit card with the hardware fully paid upfront. This is what Peleton uses... Tying payment to safety functionality is just waiting for a lawsuit.

I tried this with little success. The c-mount is such a cropped sensor that even with the EOS lens adapter you only get the middle of the shot... also I haven't found a good solution for focus and f-stop that is buried inside the EOS pin protocol. I'm sure it's solvable but haven't found it yet.

As the volume of panels has gone up over the last decade, the he price of panels has also gone down as production scales and efficiencies improve.

You would need enough demand to create a global shortage. That doesn't seem likely and would cause a short-term spike, not a long term price increase.

There are times of day that solar and wind farms produce more energy than the grid can use. This excess energy has no marginal value to the grid. You can also have off-grid solar bitcoin farms that don't need interconnectivity.

Is it possible for people to source or certify sustainable Bitcoin? Since there is a transaction chain, can you choose to buy or boycott “dirty coins”?

Creepy AF.

Agree. My thinking was... Hey it's art, somewhat functional, fun and much cheaper and more practical than many NFT's these days.

That said... I did feel a bit price gouged. There's not enough of a market to drive the price down. It's not really a consumer product and the company reached out and was somewhat surprised that I bought one.

Someone else mentioned a way to push images directly to the display but I didn't find an API to make this work. Also using HTML allows my to inject other dynamic content, not just PDF's such as a dashboard, weather, widgets, etc. It also makes it easier to display error messages if the newspaper server is down without needing to render images for errors.

Not lazy... I googled for quite a bit and couldn't find the link. I wasn't claiming it as original. ¯\_(ツ)_/¯

Agree. It's too expensive. I rationalized it as art and less expensive than most NFT's. As an art piece I do expect it to be good discussion piece and the construction makes me think that it'll last for 10 years without issues.

My initial thought was to turn this into a product but the cost of eInk at this size would need to drop to the $500-800 range to make it reasonably viable.

I thought about this. It would be a throwback to Pointcast (Circa 1990) for anyone who remembers that. They did the same thing... news, weather, stocks on a rotating schedule.

Thanks for posting. I didn't see a way to post directly to the display. I'll have to dig into.

Not sure why http://commoncrawl.org/ wasn't mentioned.