After nearly 12 years working on Gentoo and hearing blathering about how “Gentoo is about choice” and “Gentoo is a metadistribution,” I’ve come to a conclusion to where we need to go if we want to remain viable as a Linux distribution.
If we want to have any relevance, we need to have focus. Everything for everybody is a guarantee that you’ll be nothing for nobody. So I’ve come up with three specific use cases for Gentoo that I’d like to see us focus on:
People developing software
As Gentoo comes, by default, with a guaranteed-working toolchain, it’s a natural fit for software developers. A few years back, I tried to set up a development environment on Ubuntu. It was unbelievable painful. More recently, I attempted the same on a Mac. Same result — a total nightmare if you aren’t building for Mac or iOS.
Gentoo, on the other hand, provides a proven-working development environment because you build everything from scratch as you install the OS. If you need headers or some library, it’s already there. No problem. Whereas I’ve attempted to get all of the barebones dev packages installed on many other systems and it’s been hugely painful.
Frankly, I’ve never come across as easy of a dev environment as Gentoo, if you’ve managed to set it up as a user in the first place. And that’s the real problem.
People who need extreme flexibility (embedded, etc.)
Nearly 10 years ago, I founded the high-performance clustering project in Gentoo, because it was a fantastic fit for my needs as an end user in a higher-ed setting. As it turns out, it was also a good fit for a number of other folks, primarily in academia but also including the Adelie Linux team.
What we found was that you could get an extra 5% or so of performance out of building everything from scratch. At small scale that sounds absurd, but when that translates into 5-6 digits or more of infrastructure purchases, suddenly it makes a lot more sense.
In related environments, I worked on porting v5 of the Linux Terminal Server Project (LTSP) to Gentoo. This was the first version that was distro-native vs pretending to be a custom distro in its own right, and the lightweight footprint of a diskless terminal was a perfect fit for Gentoo.
In fact, around the same time I fit Gentoo onto a 1.8MB floppy-disk image, including either the dropbear SSH client or the kdrive X server for a graphical environment. This was only possible through the magic of the ROOT and PORTAGE_CONFIGROOT variables, which you couldn’t find in any other distro.
Other distros such as ChromeOS and CoreOS have taken similar advantage of Gentoo’s metadistribution nature to build heavily customized Linux distros.
People who want to learn how Linux works
Finally, another key use case for Gentoo is for people who really want to understand how Linux works. Because the installation handbook actually works you through the entire process of installing a Linux distro by hand, you acquire a unique viewpoint and skillset regarding what it takes to run Linux, well beyond what other distros require. In fact I’d argue that it’s a uniquely portable and low-level skillset that you can apply much more broadly than those you could acquire elsewhere.
I’ve suggested three core use cases that I think Gentoo should focus on. If it doesn’t fit those use cases, I would suggest that we allow but not specifically dedicate effort to enabling those particulars.
We’ve gotten overly deadened to how people want to use Linux, and this is my proposal as to how we could regain it.
Students, this Friday at 1900 UTC is the deadline to apply for this year’s GSoC. It’s an awesome program that pays you to work on open-source projects for a summer (where you == a university/college student).
It’s by no means too late, but start your application today. You can find more information on Gentoo’s projects here (click on the Ideas page to get started; also see our application guidelines) and on the broader GSoC program here.
This year I’m suggesting we do the same, on Wednesday of OSCON week. Let me know in the comments or via email if you’re interested.
I’ll also be hosting a RedMonk beering, beginning Wednesday night around 9:30–10pm. Location TBD, watch the Twitters.
Update (2013/07/19): Beers will begin around 9:30pm Wednesday at Bailey’s Taproom (213 SW Broadway, which is downtown). The place is open till midnight and we’ll likely be there till then.
If you’re a university student, time is running out! You could get paid to hack on Gentoo or other open-source software this summer, but you’ve gotta act now. The deadline to apply for the Google Summer of Code is this Friday.
If this sounds like your dream come true, you can find some Gentoo project ideas here and Gentoo’s GSoC homepage here. For non-Gentoo projects, you can scan through the GSoC website to find the details.
When I’ve wanted to play in some new areas lately, it’s been a real frustration because Gentoo hasn’t had a complete set of packages ready in any of them. I feel like these are some opportunities for Gentoo to be awesome and gain access to new sets of users (or at least avoid chasing away existing users who want better tools):
- Data science. Package Hadoop. Package streaming options like Storm. How about related tools like Flume? RabbitMQ is in Gentoo, though. I’ve heard anecdotally that a well-optimized Hadoop-on-Gentoo installation showed double-digit performance increases over the usual Hadoop distributions (i.e., not Linux distributions, but companies specializing in providing Hadoop solutions). Just heard from Tim Harder (radhermit) than he’s got some packages in progress for a lot of this, which is great news.
- DevOps. This is an area where Gentoo historically did pretty well, in part because our own infrastructure team and the group at the Open Source Lab have run tools like CFEngine and Puppet. But we’re lagging behind the times. We don’t have Jenkins or Travis. Seriously? Although we’ve got Vagrant packaged, for example, we don’t have Veewee. We could be integrating the creation of Vagrant boxes into our release-engineering process.
- Relatedly: Monitoring. Look for some of the increasingly popular open-source tools today, things like Graphite, StatsD, Logstash, Lumberjack, ElasticSearch, Kibana, Sensu, Tasseo, Descartes, Riemann. None of those are there.
- Cloud. Public cloud and on-premise IaaS/PaaS. How about IaaS: OpenStack, CloudStack, Eucalyptus, or OpenNebula? Not there, although some work is happening for OpenStack according to Matthew Thode (prometheanfire). How about a PaaS like Cloud Foundry or OpenShift? Nope. None of the Netflix open-source tools are there. On the public side, things are a bit better — we’ve got lots of AWS tools packaged, even stretching to things like Boto. We could be integrating the creation of AWS images into our release engineering to ensure AWS users always have a recent, official Gentoo image.
- NoSQL. We’ve got a pretty decent set here with some holes. We’ve got Redis, Mongo, and CouchDB not to mention Memcached, but how about graph databases like Neo4j, or other key-value stores like Riak, Cassandra, or Voldemort?
- Android development. Gentoo is perfect as a development environment. We should be pushing it hard for mobile development, especially Android given its Linux base. There’s a couple of halfhearted wiki pages but that does not an effort make. If the SDKs and related packages are there, the docs need to be there too.
Where does Gentoo shine? As a platform for developers, as a platform for flexibility, as a platform to eke every last drop of performance out of a system. All of the above use cases are relevant to at least one of those areas.
I’m writing this post because I would love it if anyone else who wants to help Gentoo be more awesome would chip in with packaging in these specific areas. Let me know!
Update: Michael Stahnke suggested I point to some resources on Gentoo packaging, for anyone interested, so take a look at the Gentoo Development Guide. The Developer Handbook contains some further details on policy as well as info on how to get commit access by becoming a Gentoo developer.
As one of my four talks at FOSDEM, I gave one on Gentoo titled “Package management and creation in Gentoo Linux.” The basic idea was, what could packagers and developers of other, non-Gentoo distros learn from Gentoo’s packaging format and how it’s iterated on that format multiple times over the years. It’s got some slides but the interesting part is where we run through actual ebuilds to see how they’ve changed as we’ve advanced through EAPIs (Ebuild APIs), starting at 16:39.
If you click through to YouTube, the larger (but not fullscreen) version seems to be the easiest to read.
It was scaled from 720×576 to a 480p video, so if you find it too hard to read the code, you can view the original WebM here.
App developers and end users both like bundled software, because it’s easy to support and easy for users to get up and running while minimizing breakage. How could we come up with an approach that also allows distributions and package-management frameworks to integrate well and deal with issues like security? I muse upon this over at my RedMonk blog.