Me: 2014 retrospective

, Will Kahn-Greene | share this (mastodon)
Note: This is an old post in a blog with a lot of posts over a long span of time. The world has changed, technologies have changed, and I've changed. It's likely this is out of date, the code doesn't work, the ideas haven't aged well, or the ideas were terrible to begin with. Let me know if you think this is something that needs updating.

This is belated because I've been busy. This is my 2014 retrospective.

2014 was chaotic. It had a lot of epically lousy parts, but also had some good parts. Like 2013, I don't know what I did with my 2014 goals.

2013 involved a lot of projects. 2014 involved fewer.

  • dennis: Lots of development on Dennis. It's really come along as a tool, though I wish I had more users since that would help flesh out issues.

  • denise: A web front-end for dennis. I did some work on it, but it needs some more to remove some of the sharp edges.

  • ernest: Everyone has a Bugizlla front-end--this is mine. I work on this with Mike, Ricky and Rehan. We made some solid improvements in 2014. Having said that, it's still pretty hacked-together and it's not generally useful to anyone else.

  • denton: A bunch of us keep a log of what we've worked on day-to-day in Standup. Denton pulls the weekly data and throws it into an email. It doesn't sound like much, but it runs on a RHEL 5 system with an aged Python 2.6 and it's a pain in the ass to use. Given that, Denton has virtually no dependencies and sports its own (mildly terrible) template parser. It continues to be a mildly fun hobby project to tinker with periodically.

  • elasticutils: I did a ton of work on ElasticUtils so that people could use it and get over the Elasticsearch 0.90 -> 1.x hump. After talking with Honza, Jannis, Erik and Rob at PyCon US 2014, we decided it wasn't worth keeping ElasticUtils going. Honza built elasticsearch-dsl does some of the things ElasticUtils does, but better. Several Django-focused shim libraries that sit on elasticsearch-dsl have appeared since then. In January 2015, I deprecated the project and no longer work on it.

  • richard, steve and pyvideo: I did a bunch of work on these three right after PyCon US 2014 and then I totally burned out. Sheila took over admin of pyvideo. I tried to reduce my role to fixing architectural problems with richard and steve. I haven't spent much time on either, though. I'd like to spend more time on them, but now I'm having difficulties finding free time to spend. These projects are currently struggling.

I also spent time on a bunch of other projects:

  • peep: I helped a bunch on peep adding support for github tarball urls which we use a lot for Input and SUMO. I'm now a maintainer and help out with project maintenance.

  • django-ratelimit: James did an overhaul that I reviewed and helped with.

  • Sandstone: Schalk built this project as a Bootstrap version of Mozilla Sandstone look and feel. I was thinking I'd use it, but it seriously lacked documentation so I helped work on that for a while.

  • elasticsearch-dsl-py: I spent a portion of PyCon US 2014 sprints talking with Honza, Rob and Jannis about what I liked about ElasticUtils and we worked out parts of the elasticsearch-dsl-py API. Since then, I very occasionally comment on issues. I'd like to spend more time on this project, but Input and SUMO are still stuck on Elasticsearch 0.90, so I don't have anywhere to use it, yet.

At work, I spend most of my time on fjord which powers Mozilla Input.

2014 was interesting:

  • The sec-champs group fizzled out. I continue to help manage Django security updates across the various Mozilla web-sites. (Security)

  • I continue to work on l10n issues by improving the Dennis PO/POT linter. I also wrote Denise which anyone (particularly translators) can use without having to install Dennis. (L10N)

  • I got around to writing a site-health dashboard for Input. This makes it much easier to spot issues after deployments and code changes. (IT)

  • I wrote a smoketests test suite using Selenium for Input. This reduced goofs with feedback that we had periodically. (QA)

  • We set up Fjord to work with Vagrant. The vagrant provisioning script is a bash shell script and isn't idempotent which is a bit of a drag occasionally. However, this system has made it a lot easier for new people to contribute to Fjord/Input development. L. Guruprasad helped a ton here. (Community building)

  • I overhauled the feedback forms on Input making them more responsive, accessible and generally better. (UX/UI)

  • I wrote a translation system for Input which does automated machine and human translation. It's currently using Gengo, but the system is modular so we could write modules for other translation companies. This translates incoming feedback in non-English languages allowing our analysis tools to operate on everything in English. (Software engineering)

  • I started a Dashboards for Everyone project which coupled an API for querying feedback on Input with some examples using d3 for converting that code into dashboards. I thought this project would do a lot to alleviate peoples' needs for data, but as far as I can tell, it never caught on. (Software engineering)

  • Early in 2014, I started working on a data retention policy for Input data. Over the course of 2014, this policy gelled and was implemented. (Privacy)

  • In 2014, I committed to writing up project plans for the bigger Input projects and maintaining them in public. The thinking was that this lets other people follow along and correct missteps. Also we have a record of why we did certain things the way we did them. In reality, I don't think anyone really cared and/or looks at them. Still, it occasionally helps me, so that's probably good enough. (Project management)

Challenges for the year:

  • I think I'm easier to work with than I was in previous years. I spent some time fixing how I did things. Over the course of 2014, the indications that I was hard to work with have mostly disappeared. It's entirely possible that's because I switched projects and work with different people now. As a side note, User Advocacy group at Mozilla is amazing to work with.

  • I'm still working on too many things. I continued working on fixing this. I retired ElasticUtils which was taking a lot of time. I retired from working on playdoh, funfactory and friends. Still, there are projects I want to work on that I haven't/can't, so that bums me out.

  • Two roommates is still chaotic. It'll be like that for a while, I think.

  • Wearing lots of hats on the Fjord/Input project really took a toll: project manager, architect, lead/solo developer, QA, deployer, ... I do it all. I learned I can do this, but that it's exhausting and I spent most of 2014 feeling overwhelmed and like I can't move fast enough. I really appreciate Ricky, Rehan and Mike helping with reviews and technical questions I have, else I would be sunk.

In 2015, I want to:

  • Spend time working on pyvideo, richard, dennis and ernest and get them to better places.

  • Overhaul my blog. Douglas is great, but the blog theme has issues and has since I last overhauled it in like the mid-2000s.

  • Find more time to tinker. I had like zero tinker time in 2014--it was just too chaotic. No tinker means I'm falling behind (or at least feeling like I'm falling behind).

  • Blog more. I did a few impressive things. I wish I had blogged about them so that there's a record of me doing those things and also so that others could learn from my experiences. It's hard to find the time, though.

That's 2014 in a nutshell!

Want to comment? Send an email to willkg at bluesock dot org. Include the url for the blog entry in your comment so I have some context as to what you're talking about.