The encoding/gob package has been rewritten to eliminate the use of unsafe operations, allowing it to be used in environments that do not permit use of the unsafe package. For typical uses it will be 10-30% slower
This seems like a highly questionable decision to me. I've invested a lot of time into optimizing my code that uses encoding/gob, so to now get hit by a wholesale 30% slower stamp from the Go team is a bit weird. These environments that can't use encoding/gob with unsafe, can't be that common, however it seems that everyone will suffer from their limitations.
Trivial solution: "fork" the old, fast version of std library code into its own repo and the only change you need is to import that repo instead of "encoding/gob"
I did that few times (mostly in the other direction, where I wanted features from unreleased version of Go)
> These environments that can't use encoding/gob with unsafe, can't be that common
They aren't, but I will provide one example that benefits from this change. As of 1.4, because of this change, it will be possible [1] to use encoding/gob (and other packages that import it) in the frontend via the GopherJS project.
It's true that you can't directly import unsafe in your own Go code on App Engine. However importing standard library packages like encoding/gob, which then proceed to use the unsafe package for themselves, works just fine on App Engine.
App Engine is actually the biggest reason why I don't immediately like this change. For other environments I can just fork the Go 1.3 encoding/gob implementation and use that. However on App Engine I can't import unsafe myself, so the only way to run really fast unsafe code, is to use the standard library. This latest change removes that option from me.
- `atomic.Value` is more or less a type-safe pointer with atomic updates. https://codereview.appspot.com/147720043 uses it to speed up reads on a concurrently accessed, rarely updated map, for instance.
- look forward to the 1.5 GC changes, can see them being huge for apps that are sensitive to peak latency (where there's a cluster of Go services you hit with fan-out, etc.).
- interested in seeing how folks use `go generate`, both what they do with it and how it is to use.
"Go 1.4 can build binaries for ARM processors running the Android operating system. It can also build a .so library that can be loaded by an Android application using the supporting packages in the go.mobile repository."
I just saw David Crawshaw do a demo of a “full Go” program on Android, a 3D gopher rendered and rotatable by touch. I don’t claim to understand the details but, if I understand correctly, Go was handling the whole thing – touch events and OpenGL – without depending on interop with Java APIs.
NDK code can only produce shared objects, which are loaded into the Java application processes.
The so called Native Activity is just a Java wrapper with a pre-defined set of methods, like touch events that map into a set of pre-defined export symbols.
The so called native activities run on their own thread and get the Java events, like touch, over an UNIX domain socket.
3D graphics programming is one of the few blessed NDK APIs. Even 2D is not accessible to the NDK in a developer friendly way, except for a bare-bones framebuffer, even though Android UI makes use of SKIA.
Outside 3D graphics, sound and partial POSIX support there isn't that much available in the NDK world without interop to Java APIs (sensors are actually JNI wrappers).
It will be nice to have Go available, but most likely the Android team will keep the same cage for Go code.
You're correct; these aren't APKs. I saw the lead dev present the project at a meetup and it appears that they're targeting games and apps developed using the Android NDK rather than standard Android Java apps.
How much of the Android API can you access from the NDK? My understanding is that it's for OpenGL and building regular apps isn't something you want to do.
Worth noting that Go has been able to generate Android-compatible binaries for some time -- Linux ARM binaries -- however they are executables, not libraries. Still possibly usable, but more awkwardly.
So now they added the ability to generate libraries. Which is interesting because, to my knowledge, they still don't support building libraries on any other platform. It would be useful.
You were thinking maybe a gigantic new language feature would show up in the delta from 1.3 to 1.4.
If you follow Golang development, one of the uncanny things about it is how they really seem to nail the release schedule. They say what they're going to do. They say when they're going to be done. Then they do it, and announce their completion.
I think you can safely assume you'll know a ways ahead of time if Golang is suddenly going to become acceptable to people who require parameterized types to adopt a new language.
I'm sure you're right; I'm just chasing the little dopamine hit from predicting someone would make a comment about generics and knowing how I would respond.
I think the devs have been clear on no need for generics anytime soon. If they do anything like generics its going to be structures and patterns for concurrency related development.
