Came to share this too. I used Joost at this time. Got me into watching the world poker tour on the platform and into playing poker and watching other live poker.
Good point. I see his current domain is "devblogs.microsoft.com". A few times I have seen @dang (or a teammate?) upgrade the HN software to show more URL details on the homepage. It would be nice to do the same for this one. A good example is fortune.com. Their blogs are utter rubbish, and their URLs now appear on the front page as something like "fortune.com/blogs" instead of just "fortune.com" (reserved for actual articles from the magazine).
Definitely one of the top blogs on that website. I wish win32 function articles would link references in blog posts, his explanations and examples helped me out a lot. Larry Osterman was another good blogger when he was still active.
MattKC recently did an amazing documentary on the development of Lego Island for PC. Quite a bit of the story was about the company behind it convincing LEGO to let them use their IP etc. Was a great watch:
https://youtu.be/bG55COe_f8I
Of course, this is from Tesla, and they famously puff up everything.. but I would expect even if not as good as claimed here, for the downhill part to certainly help a lot still.
What may not help is if they fully charge at work and have no battery capacity to soak up the regen on the way home. Regen ability gets limited the closer you get to 100%.
Neither are Teslas, but our cars range from getting 10%-33% back when going back down a mountain (curves, no traffic, avoiding using the brake pads or switching between charge/discharge on the throttle).
The results they report are incomparable for two reasons:
- straight line steep mountain roads don’t exist. On curvy roads (like mine) they’d need to repeatedly round trip energy to and from the battery, and that’s going to eat about 5% each time.
- the article claims going up and down the mountain uses as much energy as driving one mile. During the ascend/descend they drive 500 miles. The wind resistance for one mile flat and one mile on the mountain should be comparable, so their 95% efficiency number must be subtracting that (and rolling resistance) out.
What it does is uses a PCIe pass-through GPU to render the output, but instead of sending the actual video result to the actual GPU output (and thus, needing to switch input on your monitor), it renders from the GPU into a shared memory buffer which the host can display (in a window, or full screen). So it sortof works like remote desktop, but lower latency and full resolution due to using a shared memory buffer instead of TCP.
