Service(s) affected All VMs running on our production Ganeti cluster will need to be non-live migrated to their secondary nodes (i.e. shutdown and start is required). We expect the outages for each VM to be short (under 5 minutes each). To see a list of VMs that are affected and when please see this page. We will ensure the VMs are pingable after the reboot, but you may want to check that services started properly for any services we don’t already monitor.

As of September 1, the OSL has concluded its first ever crowdfunding campaign. With the support of our donors and the larger open source community, we managed to raise $7,140 to support our students here at the lab in compensating them for their work and enriching their experience through events such as Beaver Barcamp and other conferences. We are so very thankful for the support of our partners and friends. If you would like to donate to the OSL, visit our donation page or our OSU Foundation page to make a direct donation. And follow us on Facebook, Twitter, and Instagram to hear about more fundraising opportunities we hold in the future.

Since its creation in 2002, the Open Source Lab (OSL) has been a beacon of innovation, experiential learning, and hosting at Oregon State University and beyond. As Director, I’ve had the pleasure of watching the lab and its employees grow over the years and I’m so grateful to have been given such an opportunity. I want to express why the OSL is important not only to the open source community, but to our larger academic institution and its students.

In July 2017, the OSL began providing virtual machine and database for hosting the demo service of polr. The project itself simplifies long and complicated urls. It was started by a high school student back in 2014 who has since volunteered on the Fedora Infrastructure team. For more information, visit the project website.

In July 2017, the OSL began providing virtual machine and database for hosting the demo service of polr. The project itself simplifies long and complicated urls. It was started by a high school student back in 2014 who has since volunteered on the Fedora Infrastructure team. For more information, visit the project website.

We’re very excited to announce our very first crowdfunding campaign through the OSU foundation! The OSL relies on donations and revenue generating projects to fund its work and its students–the number of students we can hire and train is dependent on the donations we receive. Help us fund five additional students for the coming year, including wages, professional mentoring, and the opportunity to attend a major open source conference, such as SCALE or OSCON.

Programming languages are a touchy topic in Computer Science. In certain crowds even mentioning a language will elicit groans and eye-rolling. Conversely, there are crowds that will only use certain languages for all projects. These people have lost sight of the fact that programming languages are tools. Languages have certain problem sets that they’re really good at and some not so much. If you were to ask me to do some complex math or signal processing, I would point you to MATLAB. Would I use MATLAB for developing a GUI? Not in a million years. So why do we choose C? Well, C is efficient since it’s practically one step above assembly and with a modern compiler it compiles down to a small executable. Plus modern compilers have extremely good optimization algorithms that can optimize your program better than if you wrote it by hand in Assembly. This makes C a great tool for embedded programming and systems level programming, which is why we have been using it for so long in these fields! However, a downside (and upside!) of C is that it’s like assembly. It will let you do whatever you want, even if that means shooting yourself in the foot. There is no type safety, there is no memory protection, and no thread safety built into the language. You have to do all of that yourself with mutexes, semaphores, and checks. It’s good to know about these concepts and be able to design a system that puts these protections in place, but every project should not be an exercise in memory management and complex concurrency. We should move on to tools that help you rather than give you enough rope to hang yourself with. With computers being as important as they are, security should be our number one priority when writing software. If you’re writing an application that does an unbounded copy from input (ex. heartbleed) in your final release, you just added another vector of attack to someone’s computer.

As of September 8th, my time here at OSU will officially come to an end. As sad as I’ll be to leave my life here in Corvallis and as nerve-racking it is to enter the real world, I realized recently I’ve spent the last seventeen years of my life in school and I’m ready to break free! I’ll be leaving with both a degree in English and in music performance, both of which have taught me so much not just about their respective fields, but about growing up and how to work with what you’ve got.

April 8 started out like any other Saturday in Spring in Corvallis: rainy, then sunny, then windy, windy-rainy sleet, hail, and then of course, sunny again. Despite the crazy weather, people from all walks of life still convened at the Kelley Engineering Center on the Oregon State Campus for the Open Source Lab’s annual Beaver BarCamp. For several years now, the OSL has hosted one of Oregon State’s only unconferences to great success. Just as a refresher, or for any newbies out there, an unconference is an event in which the attendees decide the topics of presentation and discussion the day of, rather than determining these topics ahead of time. This year, we tried a few new things.

The OSL took on a new project and delivers a new mirroring site libpng source for libpng tarball distributions that is 20+ times as fast as the old site. It also now responds to http and https requests as well as ftp requests. For more information and links, see the original announcement.