Week 30 2017 - Container runtimes; Go 2.0; Video DRM

A catchup week, with various container related subjects, the road to Go 2.0, and the W3C’s DRM for videos. My apologies for not publishing for two weeks, I moved house and underestimated how much of my time that would take.

Container runtimes

Just a couple of news items from the container world that caught my eye over the past couple of weeks. First up is Oracle releasing an OCI-Based Container Runtime. The OCI is the Open Container Initiative which basically controls the specifications for both runtimes and images of containers. Oracle’s decision to build their own runtime following this spec means that they’ve developed a compatible alternative to Docker. Their reasoning behind this is to ensure there is an alternative available1, but it also shows that Oracle as a company is taking containers quite serious. This is potentially good news with regards to their purchase of Wercker2, which seems like a likely place for them to use the runtime as well. I doubt it will get a lot of traction outside of Oracle though, but it’s good to see another runtime available. In addition they released a couple of helper tools for CI integration with containers so you might want to check those out in the above linked article.

Speaking of the OCI, about 2 weeks after Oracle announced their new runtime, both of their specifications reached 1.0. As far as I can tell the specs didn’t receive any sudden big changes, but have just been stabilized for future improvements.

It seems to be the time for reaching 1.0, as containerd, the runtime from Docker that was split off and donated to the CNCF is aiming for a 1.0 release. It’s not done yet, but so far 2 alpha releases have been published so I’m hopeful it will show up in time for the next stable Docker release scheduled for September.

Following Play with Docker and Play with Moby, Kubernetes now has a similar playground named Play with k8s. Once again I’ll say that I’m all in favor of these types of playground as they lead to easy experimentation so go forth and play.

Go 2.0

As part of Gophercon, which I can’t really talk about as I was too busy to watch or read much about it, there was a blogpost concerning the plans for moving to version 2.0.

One of the good things about Go is that ever since the 1.0 release they’ve ensured backwards compatibility. This was also their stated reason for not rushing into a 2.0 release as they only wanted to do so if this was needed for breaking changes. The above blogpost has now updated the plans, and if you’re interested in Go I recommend you read it in full. Otherwise, a short summary is as follows.

The plan is to ensure a very smooth transitions to version 2.0. This means that any non-breaking changes will already be introduced in 1.x releases, with any potential breakages delegated to 2.0. But still in such a way that it should be possible to keep running/compiling your old applications. This process is likely going to take years, but they’ve now started it with a call for experience reports. These reports are basically real world examples of where Go in its current state failed to meet your expectations. Again, if this is something you’re interested in any of this, read the whole article and if possible provide your issues in the way mentioned.

Video DRM

In a move to standardize DRM, the W3C has announced it will publish its Encrypted Media Extensions. This isn’t exactly a popular decision, especially as all objections against it were overruled by Sir Tim Berners-Lee, the Director of W3C. Other than those people and companies who believe that DRM is the only way they can publish their content and make money3 not many people like DRM.

As the EFF has stated in their reaction to the news however, there is an understanding that this can’t be avoided. However, there are many downsides to the spec as it currently stands. Not in the least that security researchers who wish to look for vulnerabilities in order to protect legitimate users would be breaking various laws. Or even people who attempt to automate things like creating subtitles for these encrypted videos. And while there were attempts to make alternatives to this work, this didn’t happen.

In a final attempt to stop this, the EFF has now filed for an appeal. This isn’t a great situation and locking things down through these standards doesn’t seem like the best solution, so personally I hope this won’t come to pass.

  1. Technically speaking, the OCI has implemented a reference design called RunC so there was one available already. [return]
  2. Although a lot more will need to happen before I’m happy about that. [return]
  3. A reference about tightening grips and planets would probably be appropriate here. [return]
Don't ignore new content! Get the latest by subscribing through RSS, Twitter, or email!
comments powered by Disqus