[go-nuts] Re: All the Go project history, live in RAM

2018-04-14 Thread Dmitri Shuralyov
Brad, another question. I've read the documentation at https://godoc.org/golang.org/x/build/maintner/godata, but it doesn't make it clear: is it safe/supported to have more than 1 process on one machine call godata.Get and use returned maintner.Corpus at the same time? I know each call uses th

[go-nuts] Re: All the Go project history, live in RAM

2017-05-02 Thread mhhcbon
looks awesome! On Sunday, April 30, 2017 at 11:42:03 PM UTC+2, bradfitz wrote: > > Gophers, > > Want to analyze the Go project's Git, GitHub, and Gerrit history? > > Here a package to make it easy: > > https://godoc.org/golang.org/x/build/maintner/godata > > See the example: > > > https:

[go-nuts] Re: All the Go project history, live in RAM

2017-05-18 Thread Dmitri Shuralyov
> > The data is generally under 1 second behind reality, thanks to the > webhook+incoming-SMTP server we run now at https://pubsubhelper.golang.org/ > to get updates from GitHub & Gerrit. > Just a quick implementation question related to that. I understand you use Webhooks to get data from Git

[go-nuts] Re: All the Go project history, live in RAM

2017-05-18 Thread Dmitri Shuralyov
Another small question. https://godoc.org/golang.org/x/build/maintner#GerritProject.Server says: > Server returns the Gerrit server, such as "go.googlesource.com". I thought that go.googlesource.com was a Gitiles server, and go-review.googlesource.com is where the Gerrit server is (according t

Re: [go-nuts] Re: All the Go project history, live in RAM

2018-04-14 Thread Brad Fitzpatrick
Yeah, I think that should work. IIRC, the cache files aren't appended to, but atomically rewritten & replaced. So running two on the same filesystem will waste a bit of network & disk I/O, but it should be correct. On Sat, Apr 14, 2018 at 5:25 PM, Dmitri Shuralyov wrote: > Brad, another quest