• Archives

  • Categories:

SV Code Camp, Datathon at UC Berkeley and GitHub Universe

“Conference season” in the autumn for me is roughly the end of October through early November. I didn’t want to travel quite so much this year, and there were plenty of events local to me, so I ended up with an unexpectedly busy October and November, either traveling or doing local events several weeks in a row.

I kicked off my conference season with an internal event at the IBM Silicon Valley Lab, where I have an office. My team focuses on external advocacy, but we have finite resources and career development is frequently aided by giving talks, so we’re always seeking to encourage colleagues do external-facing work and advocacy. To help this, my internal talk was on open source software development, with a tie-in to the work that IBM Z has been doing in this realm.

My first external event this season was the Silicon Valley Code Camp in San Jose. I had never been to this event before, and it was over a weekend so I could only carve out time for the first day, but I’m glad I went. We had an IBM Z booth there, positioned next to the IBM Developer booth.


IBM crew. From the event photos collection, source.

My colleague Matt Cousens was out from New York, so along he, my boss Jeanne Brooks and I staffed the booth for most of the day, giving away mainframe stickers, books, and other goodies.

I also gave a talk on Developing for the Modern Mainframe, slides. The room was laid out in a way that made it easy to shift into a conversational-type talk, so with a couple dozen attendees it was easy to slip into a casual back and forth as I made my way through my slides. It was fascinating to see the range of experience in such a small audience, some folks having strong experience with Z and others just vaguely curious about what exactly mainframes are and what needs to be done specifically to develop on them. This diversity also allowed for a nice crew of allies in the audience, not just IBMers, but experienced folks who could chime in on some of the z/OS topics I’m not strong with yet. It was probably the most satisfying talk of the year enjoy-ability-wise, and I hope to participate in this event next year too.

Then there was a Datathon at UC Berkeley, which my colleague Sudharsana wrote about it here: Datathon for Social Good: IBM Z & UC Berkeley. Our role was to sponsor the event and provide key content, then eventually judge the outcomes from participants. In preparation, an environment was created on IBM Z that the students could log in to and use Jupiter Notebooks to interact with Z and run data analysis on their chosen data sets. The event began on Friday evening with an introduction and then a panel talking about what IBM Z was. I was one of the panelists, and it was interesting being on it with some of my colleagues from IBM and a representative from the State of California. My perspective really is quite different from those who have been working in this space for a long time. It was also at this event that I learned that most traditional students in college today aren’t very familiar with IBM as a company. Upon reflection, the reason I was so familiar with them was probably because I had IBM PCs in my youth. My first and second computers were both IBMs. But IBM no longer has a PC division, and much of their work is with enterprises, how would regular people have experience IBM today? I’m not sure.

The event continued on Saturday, during which we sat with the students to offer mentorship in case they needed it, and to wrap up the day with presentations and picking the winners. One of the strengths of the mainframe is the ability to quickly crunch data, so it was fascinating to see the students whose Datathon projects actually used that. As a result, we really honed in on projects that were (or planned on) using large data sets and/or processing a lot of real time data. The winners focused on student financial help, wildfire modeling, and homelessness support spending.

Later that week, we invited the winners to the IBM Watson office in San Francisco for a tour and some talks on AI and Machine Learning, along with a quick talk on Open Source on Z that I gave. That’s where I found the best IBM logo ever!

More photos from the Datathon and follow-up in San Francisco here: https://www.flickr.com/photos/pleia2/albums/72157711951959732

Finally, I attended GitHub Universe in San Francisco in mid-November. It was a last minute thing for a very exciting reason: Travis CI now supports builds on Z! Alongside Arm and newly announced Power support, this is momentous. With this initial implementation, open source projects get builds for free, and it paves the way for paid support for other customers. Making it easy to build your software for architectures other than x86 is something I’m passionate about, so it was nice to be able to attend GitHub Universe at the Travis booth just after the announcement to talk about it. There I met with some folks from IBM Power too, who have done a lot of great work with open source projects that I hope to learn from and replicate. Our role there was offering support when people had questions, and a demo of the pipeline was given in the afternoon.

A few more photos from the event are available here: https://www.flickr.com/photos/pleia2/albums/72157711953504618

And then I was off to KubeCon in San Diego! My last event of the year is Open FinTech in NYC next weekend. Phew.