Adrián Pérez is a user on mastodon.social. You can follow them or interact with them if you have an account anywhere in the fediverse. If you don't, you can sign up here.

Google has this annyoing internal engineering ethos of "Google knows better" wherein they discard a whole bunch of established industry practices in favor of their own stupid ideas

@jon `#!/usr/bin/env python` rather than using `#!/usr/bin/env python2` (and closing the bugs about it as WONTFIX)

Their bash guide mentioned on HN the other day

AMP

@sir @jon Also, many of their open source projects suffer from acute NIH-syndrome. Instead of contributing to an established one, they start anew. Sometimes even multiple times. To name a few: Skia (instead of helping Cairo), BoringSSL, Gyp, GN, Bazel, gtest, gperftools,... It's not even about the licenses of the existing projects not being permissive enough in most cases. There has to be something dysfunctional in Google that causes this.

@aperezdc @sir It's also possible that google has some unique problem spaces to solve, and by making existing open source projects suit their needs, they make it less useful to others. Their employees contribute a lot open source projects. Granted it's mostly their open source, but it's still open source. My point is there may be perfectly rational reasons for why google does what they do that benefit them AND the community.

Adrián Pérez @aperezdc

@jon @sir I have trouble buying that argument. Google indeed has some unique problem spaces, but precisely the solutions they develop for those tend to be proprietary, like BigTable (en.m.wikipedia.org/wiki/Bigtab) and GoogleFS (en.m.wikipedia.org/wiki/Google). The open source projects they have which duplicate efforts tend to be out of that category.