7 ms·

Also he's had a look at the Imperial COVID19 model that has attracted so much bile in the last week: https://twitter.com/id_aa_carmack/status/125487236876327731

• renewiltord 6y ago
  I’ve always liked Carmack’s walking the walk. It’s common on Internet forums (like this one) to talk about how some code is shit or whatever. Carmack submits patches and sees them through. An ideal to aspire to. His Twitter is a great replacement for his planfiles. One of the few I follow.
• hedora 6y ago
  > I can’t vouch for the actual algorithms, but the software engineering seems fine. I also looked at the code a bit, and it seemed like typical bad natural sciences / electrical engineering grad student code that could have as easily been done the same way in matlab or python (or, as Carmak suggested, fortran). I guess I’m in good company in thinking it was ok-ish, so now I’ll ask. Why are people so upset with this code base / the choice of C++?
  • toyg 6y ago
    > why are people so upset Politics. “Anything that supports a position I don’t like, should be attacked with full force. Anything less than perfect must disqualify the entire position.”
  • sgt101 6y ago
    I think that there are two pathologies. The first is that alot of programmers have very specific and utterly unshakable views on what constitutes good or acceptable code. These are rooted in their experiences, but often have no foundation in terms of evidence or science (because there is very little good research on this). The second issue is politics; there are a lot of people who hate the idea that data science or science can provide inputs to policy decisions (in business and government). I see modern political groups as : - traditionalists : protecting what they have, policy aimed at preserving status quo - capitalists : untrammeled use of capital power to maximize capital return - idealists : interested in a higher causes/morality (as they see it) and unconcerned about other impacts or concerns - technocrats : concerned about taking the highest utility actions (measuring utility in various often unconnected ways). Capitalists have held the whip hand for 50 years, idealists (trade unionists, safety campaigners, ect) have been in full on retreat, traditionalists have managed to align with capitalists (by co-opting them with nice things like country estates). Technocrats have swayed in and out of influence; the crisis of 2009 really smashed up peoples faith in economists for example, the first gulf war really brought logistics and management science to the fore... but the war on terror has smashed this up now. Data scientists are mostly technocrats but often are compromised/partially capitalists (because... money). In every case that I have seen where capitalists are thwarted by data science they react with fury - because at the moment they are very unused to being thwarted. Recently though idealists have been grabbing more of the mic, and have a large contingent of data scientists of their own - ie. university scientists. This is why climate change modellers are so hated!