751 karma · joined May 2, 2013
Cert pinning is basically free and is sort of a "you must be this tall to ride the ride" thing--not secure, but keeps the riff raff out.
Of course, Apple used binary patching and compression and all that jazz... this was for the "full" version that delivered entire files.
IIRC, topic branches were used because:
- Mercurial named branches were half baked at the time and I was advised against them by folks who knew hg better than me.
- We were already using separate clones for security update versions (3.5.x and 3.6.x) so we just kept doing the same. Some tooling could be reused.
- Rapid release was a big change and we didn't want to disrupt mozilla-central dev. Having rapid release in different clones got us out of the way and reduced the risk of negatively affecting development. m-c was the domain of developers, the rr repos were the domain of the release team.
There are projects and examples to do all three but we don't want to assume one solution is right for all domains.
We take this to the extreme...Juniper doesn't even require a web server and isn't tied to a particular serialization format! Of course, we provide optional integrations with popular web frameworks using json but the key is they are optional.
Not to be rude, but you are literally talking to someone who was in charge of this, looked at the graphs daily, and made the call when we went over the OTA limit on iOS. When I say multiple-apks are a large reason for the discrepancy, you should give me the benefit of the doubt. It's not the only reason mind you.
Also note that both platforms do delta updates but generally show the total size in the UI. So every update you aren't actually downloading that much over the wire.
Source: I used to manage the Android and iOS release engineering teams at Facebook
Facebook has done it twice at this scale: PHP -> Hack, plain JS -> Flow JS.
* It also writes build output to an output directory
* It detects a tty and does the right thing, so on CI you get a linear log
* In the output directory it also includes timeline graphs that can be loaded in Chrome's devtools traceview (https://github.com/catapult-project/catapult/blob/master/tra... and in Chrome proper).
* The tracing can be viewed live https://buckbuild.com/command/server.html
(disclosure: my team created Buck when I was at FB)
I would argue Buck actually helps to solve what you are concerned about. At Facebook, everything builds with Buck (and Google with Bazel I hear)...which means you learn it once and you know it for your Objective-C library, your Android app, your Cxx service, your python scripts, etc. It really helps to standardize on a build system, and at many companies you can only do that if it supports windows/mac/linux, it supports many languages and platforms, is fast, and is easy to pick up. Buck is all of those things.
https://buckbuild.com/rule/genrule.html
Some other benefits that may or may not be important:
- Use the same build system for iOS/Android/Cxx/Python/D, etc.
- Integrates with editing in IDEs.
- Encourages modularization...more modules = more cache hits = faster builds
https://buckbuild.com/command/buckd.html
Note the daemon is automatically started when you `buck build`.
(disclosure: my team created Buck when I was at FB)
- Don't use modification times and instead do intelligent hashing (see https://buckbuild.com/concept/rule_keys.html)
- Include the compiler and a bunch of other stuff in the hash. This means you can share cached artifacts safely via a http cache (https://buckbuild.com/concept/http_cache_api.html)
- Can refer to remote files (see https://buckbuild.com/rule/remote_file.html)
- Can use any script to do anything as long as it has a single consistent output (see https://buckbuild.com/rule/genrule.html)
Because their model is so much better, you get crazy fast local caching and remote shared caching of artifacts.
Please don't use make, the newer build tools are better in virtually every way.
(disclosure: my team created Buck when I was at FB)