3 ms·
I would desperately love to see an HN discussion of this, but despite multiple submissions, the only comment points at an article of the actual paper[0]. Has n
by CarolineW 10y ago
I would desperately love to see an HN discussion of this, but despite multiple submissions, the only comment points at an article of the actual paper[0]. Has no one read this paper? Does no one have anything to say?
https://news.ycombinator.com/item?id=12143845 https://news.ycombinator.com/item?id=12143845
https://news.ycombinator.com/item?id=11853165 https://news.ycombinator.com/item?id=11853165
https://news.ycombinator.com/item?id=11822546 https://news.ycombinator.com/item?id=11822546
https://news.ycombinator.com/item?id=11808331 https://news.ycombinator.com/item?id=11808331
https://news.ycombinator.com/item?id=11805262 https://news.ycombinator.com/item?id=11805262
========
[0] https://blog.acolyer.org/2016/05/31/how-to-build-static-checking-systems-using-orders-of-magnitude-less-code/ https://blog.acolyer.org/2016/05/31/how-to-build-static-chec...
- 0xmohit 10y ago> I would desperately love to see an HN discussion of this This was the reason behind posting it again after 3 days. Worth highlighting: our checkers are built on micro-grammars, incomplete fragments of a language. Tools like vet [0] can also offers insights. [0] https://golang.org/cmd/vet/ https://golang.org/cmd/vet/