3 ms·
To preface this I have not looked into the internal details of vgo, so it could be the world's most interesting solution to this problem. I feel like this is a
by i2cpsi 9y ago
To preface this I have not looked into the internal details of vgo, so it could be the world's most interesting solution to this problem. I feel like this is a punch in the gut for a lot of people that spent tons of time looking at ways to solve dependencies in Go. From his post, it looks like Russ never really looked into dep closely as he gave it his "blessing" during gophercon. It seems like whenever there is an issue that gets brought in a proposal (not always but it seems like a lot) it gets shot down by the "higher ups". Not until it gets their attention, or it becomes a big deal to someone (Cloudfare) then is it important. Look don't get me wrong, vgo could be a great technical solution, but disregarding the work of those who actually cared about this issue before you did is a big mistake. It will alienate people in the community who care on improving the ecosystem for those outside of Google. I want to thank all of those involved in dep, it's a great tool that solved my dependency problems! Without you, we would not be having the discussion about vgo and a better way to dependency management in Go.
- dilap 9y agoIt's tough, but sometimes you need to throw away work. It's better ultimately in the long run. I think Russ has been doing a pretty good job of giving recognition and credit to the dep folks for their exploratory work.
- derekperkins 9y agoSaying that dep and prior art has been ignored or not "looked into" is unfair. Russ is very clear about where dep and similar tools fall short of a perfect solution. Not only that, but Google has contracted Sam to help with vgo, since he presumably has the most Go experience in the matter. I'm sure it's disheartening for people who worked on dep. I've had a great experience with it too, especially after the .4.1 release. While not everyone agrees, one of my favorite things about Go is that every decision is well thought out in terms of its effects on the entire ecosystem. Just because some other language or toolset does something isn't a good enough reason to force it into Go. I trust the overlords, they've been good to me so far. :)