Hacker Newsnew | past | comments | ask | show | jobs | submit | Etherlord87's commentslogin

This seems like such an obvious answer to the problem, your program isn't truly modularized if logging is global. If an error is unexpected it should bubble all the way up, but if it's expected and dealt with, the error message should be suppressed or its type changed to a warning.

I’ve worked on systems with “modularized” logging. It’s never been pleasant because investigations involve stitching together a bunch of different log sources to understand erase actually happened. A global log dump with attribution (module/component/file/line) is far easier to work with.

No, it's not much easier to please. You will compete with Chrome and you will lose. It is the core audience that is much easier to please, because those are loyal users that trust you and all you have to do is respect this trust by providing a stable service without nasty surprises.

If most people move from Firefox to Waterfox, then Waterfox can acquire Firefox devs, no? Obviously it comes to money, but the first step to gain funding is to gain popularity...

It's largely about pattern detection in our brains. If the pixels always look the same, it's easier to spot them. For many people graphics in a video game are a secondary addition but the decision making is uncontested priority; modern 3D graphics get in the way by making everything less readable.


If the volume was defined by a manifold mesh, it was already trivial to scatter points, then remove points that when raycasted down on that manifold mesh "volume" hit a normal.z < 0, hardly a node spaghetti as the article says.


Plus if you could use WASM modules as opposed to node systems, you would have a more powerful programming environment - for example, the mentioned "repeat" zone doesn't have "break" functionality, so you need to use a huge number of iterations and do an equivalent of "continue" once you're done, but even if you perfectly estimate the number of iterations you will need, the implementation is very slow. There are other problems like limited scope access, or in general slow node evaluation which make sequential algorithms (not reliant on parallelized work done by loops internal to many nodes) problematic.


I came to the comment section hoping there would be a discussion about that, but sadly overwhelming majority of posts is about malfunctioning CSS...


Yes, a 24 acre soybean field uses a lot of water.


And an average US soybean farm has 312 acres (13x larger than 24 acres): http://www.ers.usda.gov/topics/crops/soybeans-and-oil-crops/...

Which means that in 2025 Microsoft's Querétaro sites used 1/13th of a typical US soybean farm's annual amount of water.


If only millions of people suffering from lack of water knew this.


Would we be sending that water to those millions of people instead?


If you redistributed this water to a million people suffering from lack of water, they'ed get about 2 shot glasses worth per day.


Yea but it's not like those people have never seen water. And yet it's not so simple, that you can use water but the water eventually comes back to you. There is a hell lot more nuance to this.


According to this logic the ideal situation is when there are no farms anymore because then each (out of zero) farm gets maximum water.


Eventually people stop building more data centers as food becomes scarce and expensive, and farms become the hot new thing for the stock market, cereal entrepreneurs become the new celebrities and so on. Elon Husk, cereal magnate.


I was just pointing out the ridiculousness of the argument.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: