What is pyvideo.org
pyvideo.org is an index of Python-related conference and user-group videos on the Internet. Saw a session you liked and want to share it? It's likely you can find it, watch it, and share it with pyvideo.org.
This is the latest status report for all things happening on the site.
It's also an announcement about the end.
Update: See the March 16th blog post covering the current status of pyvideo.org and pyvideo-data.
Back in April of 2015, I added a rough playlist feature. Since then, 80 playlists have been created. Looking at the creation dates, it's a few every week since I did the feature.
Sheila has added a bunch of conferences and we continued to tweak steve to meet our conference-data-adding needs.
That's about it for progress since April 2015.
Where things are at
Right now, things are not great.
I'm pretty tired of working on pyvideo and I haven't had the time or energy to do much on it in a while. The software that runs it (richard and steve and the ansible stuff and all the other bits that make it go) is woefully bitrotted. There are security issues. There's tons of technical debt. We're using libraries that have undergone significant API changes. Persona is going away and I'm very skittish about securing personally identifyable information for what's essentially a hobby site. There's tons of poor architectural decisions to undo. If we move forward, there's a lot of work to do.
pyvideo never grew beyond me and Sheila. I haven't done any work onboarding new people or even getting pyvideo in good shape for other people to help out in a couple of years. The work I did do early on in pyvideo's history never panned out. There is currently no pyvideo community. The IRC channel has been dead for months. The issue tracker is solely people asking us to index new conferences and wondering why we haven't done so, yet.
When I started Python Miro Community and later pyvideo, I wanted to accomplish a few things:
Build an index of video regardless of where the video was hosted on the Internet.
Make it easy to search to find the videos you're looking for.
Build infrastructure to make the video accessible to people who are differently abled and speak different languages.
Build infrastructure letting people download the video so they can watch it on their devices.
I've been working on this for like 5 years now. I think I made a dent in the first, but had a lot of difficulties with the others.
The world has changed a lot since I started this work. Back then, there were multiple video hosts and people were interested in hosting video on their terms. Now it seems like everyone puts it on YouTube. YouTube has subtitles. It supports HTML5. It has playlists. Lots of people have accounts. Should we be building an index of YouTube video?
Search on pyvideo has always been terrible. I did the initial work with whoosh and django-haystack. I've wanted to switch it over to Elasticsearch for a while, but never did. As we added videos to the index and have a wider span of years, I claim search results got worse.
Amara is probably great, but I never wrapped my head around the service enough to really get it working right. I've followed Khan Academy and their use of Amara and they've gone through some interesting transformations over the years. Regardless, I still can't get it working well on pyvideo.
I originally started pyvideo figuring I'd keep it English-only and then people who were interested in curating an index in other languages could spin up their own richard instances to do so. I still think that's the right idea for various reasons. Instead, I caved and added a bunch of non-English langauges to pyvideo which dealt with multiple languages very poorly. We never got strings translated. We never handled multiple languages correctly in search or navigating. It's a big mess.
I don't like doing funding work because that kind of work needs all the skills I'm not good at. Carl suggested I write up a grant proposal with the PSF a while back which they granted (thank you, PSF!) but I only spent half of it because I never finished the work I promised to do. We did the gittip thing and I think that might have made a dent in domain name costs but then it dried up. I've watched other groups that are far more interesting and valuable to the world than pyvideo flounder with raising money for development and infrastructure costs. I never wanted to do ads. Thus I've been stuck in a situation where I should do funding work, but I really don't want to spend the time and energy on it (this is a hobby project) and thus we've never had any funds to do anything which is unsustainable for any kind of non-trivial endeavor. pyvideo is definitely a non-trivial endeavor.
For a while now, I've felt really guilty about all of this. What's wrong with me? Why didn't I spend more time on pyvideo? Why didn't I prioritize pyvideo over other things in my life like jobs and family? Why do I feel so awful about the state of pyvideo all the time? How does the failure of my hobby project reflect on my competency as a software engineer? I hear these questions from other people about their projects, too. I'm not going to talk more about that here because there are much much better treatises on these topics by other people, but this is what's in my heart right now.
Anyhow, so that's where things are at. To summarize:
pyvideo software has a lot of problems and requires a lot of work
pyvideo never achieved my dreams
So, what now?
I'm ending pyvideo. This journey is over.
Sheila and I did some work to export all the data into JSON and put it into a pyvideo-data repository with some workflow infrastructure to let people continue adding to and honing the data. Sheila felt this was important for the various research projects looking at gender and race balance at conferences, linguistics usage and other things. 1
The pyvideo.org website will get shut down at some point this year. I'm not passing it off to someone else. Why? Because I don't think what we've got now is worth continuing. I think if someone really wants to continue pyvideo, they're better off starting something new without all the debt. Maybe a statically generated site?
I think that covers everything.
This is the most important thing to come out of pyvideo. I hope someone shows up to curate this data going forward.
Yes, this is the end. Yes, I'm bummed pyvideo didn't match my hopes and dreams, but that's more my fault and in spite of the efforts of individuals and organizations who helped a lot along the way:
Participatory Culture Foundation and Nick Reville who built Miro Community which is where this all started.
Carl Karsten and Ryan Verner and Next Day Video who footed the bill for pyvideo for the first couple of years and helped guide the creation of the site.
Jesse Noller who reached out with help from Rackspace for hosting. That helped a ton.
Rackspace who have hosted pyvideo for several years now. They occasionally bill Sheila still, but generally have been wonderful.
Steve Holden who always patiently listened to my kvetching and encouraged me.
Deb, Fateh, Asheesh, Chris and Bassam who provided a lot of support and advice over the years.
Sheila Miguez who is an amazingly patient and capable friend who did a ton of work on pyvideo and without whom I would have ended it a couple of years ago for sure.
And lastly all the people who I talked with at PyCons and other places who appreciated the work Sheila and I were doing.