Hacker Newsnew | comments | show | ask | jobs | submit | bradfitz's comments login

More specifically, with os.PathListSeparator. (colon on Unix, semicolon on Windows and \x00 on Plan 9)

-----


No, that was one of the reasons for GOMAXPROCS=1 on App Engine, which is a hardened environment.

For your own code, we trust you to fix your races or deal with the consequences.

-----


Does this mean that App Engine will never see it raised above 1?

-----


No.

-----


Glad you liked it! We enjoy making them.

-----


Please keep them coming. Will watch every single one, multiple times! Loved hearing your thought process and confidence with what was being typed. Now when I code I pretend I'm talking to either of you. It's not weird, right?

-----


Not to take away from the awesome videos now, but I'm a bit sad we won't see anything like this from the LiveJournal/memcached/Danga era - that would have been great watching, I suspect, from my memory of yours and whitaker's posts :-D

-----


Could I suggest also recording an episode when you actually gear up to merge this into master?

-----


We committed it straight to master after we finished the video. The code has been worked on quite a bit since.

-----


I was referring to something which Brad mentioned at the start of the video. About making this client reuse an underlying connection for multiple requests.

-----


Please make more of them.

-----


And despite me silencing the master bedroom one, the other four didn't shut off. That's my major complaint.

-----


They can take it. Or they can build better products.

I don't put on kid gloves when a product is this bad.

I'm nicer when it's a beta or unreleased. But I paid $500 for these, for a final product. This is unacceptable.

-----


I now believe that conclusion to be false, after talking to more people.

-----


Can you share what you found?

-----


I'll be sharing more but not yet. Still gathering information.

-----


Okay, fair enough. Thank you.

-----


https://groups.google.com/forum/#!searchin/golang-nuts/ios/g...

Quote:

gophers,

I'm happy to announce that the iOS port of Go recently gained full cgo and external linking support.

I've also got a simple C based application working on a real iPad.

The port lives in ios3 branch of https://bitbucket.org/minux/goios. Read misc/ios/README for some rudimentary documentation.

The port is Go 1.4+, and it's based on upstream dev.cc branch. I intend to propose for inclusion into Go 1.5 after finding a way to test it (either build each test as an App, or use the open source xnu/arm port). Brad mentioned that inclusion of iOS support might happen in Go 1.5 in his talk the state of the gopher (http://talks.golang.org/2014/state-of-the-gopher.slide#40), now I believe it really will happen.

The iOS port is my first Go OS port, but it's the only one that takes almost 3 years to complete. :)

Go build iOS Apps. Let's see when can we see a Go based App in the store.

Minux

-----


I love Rob Pike's contribution to the thread:

Ken and I ported Plan 9 to the SPARC in 6 days over one fun Christmas break.

I wrote the disassembler (for the debugger) and Ken wrote the assembler, so we could cross-check each other's work. The hardest problem other than fighting register windows occurred when we both misread the definition of an instruction the same incorrect way.

-----


You seem to be confusing compilers and build systems. You can use whatever build system you'd like and with whichever compiler.

-----


I'm not sure what you define as a build system, but it's even in the description of the go build tool (https://golang.org/doc/articles/go_command.html):

> An explicit goal for Go from the beginning was to be able to build Go code using only the information found in the source itself, not needing to write a makefile or one of the many modern replacements for makefiles. If Go needed a configuration file to explain how to build your program, then Go would have failed.

And the hope is that go generate would be a hook into the dependency graph of go build. Much like the architecture switches, the naming conventions, etc, it seemed like it really could have solved my project's protobuf problem.

But it's not, and for reasons that rob pike's doc and the mailing list did not make clear. It looks like a mis-step to me, and I don't think it's due to my misunderstanding of what `go build` and 6g/gccgo are

-----


Is it possible to do away with the typical Go project directory structure? I thought all that was enforced by the tools, but it's possible that I didn't dig deep enough to uncover greater flexibility.

-----


No it's not if you ever intend to type `go`.

The go team has decided that the go language doesn't depend on any build system, but pushes a single build system one that requires source code compatibility and provides no benefit over many existing and easy to use build systems (makefiles, tup, etc).

-----


The original motivation for putting it into the standard library is so the 'go' command can use it, to watch the filesystem and know what needs to be rebuilt before you even run "go install" or "go build".

-----


"then deploy a single binary file without worrying about dependencies, dynamic linking, or segfaults"

-----

More

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact

Search: