Wednesday, May 04, 2011

An interview with Brion

Now that Brion has had some time to get settled back in, it seemed like a good idea to ask him some questions. I did and he returned them with some really nice answers. Enjoy..
Thanks,
       GerardM

What did you like best while you worked at Status.net?
Working with something *different* was very nice -- I was able to learn new techniques, get exposed to new tools, etc.

StatusNet's emphasis on federated networks and interoperability gave me an appreciation for issues of service discovery and data sharing, which has inspired me to think about better ways we can share media, pages, and code within and between projects.

What was the biggest change you noticed when you were back at the WMF?
Air conditioning! San Francisco only has a few hot days a year, but they were miserable in the old office space. ;)

Generally I'm struck with how well we've curated our pool of open-source developers -- a lot of the contract & full-time hires in tech have come from our community, and it seems we're now better able to keep several projects going at once *and still* deal with the usual meltdowns. There are still some growing pains of course -- getting code review & release process back on track has been a bit bumpy after the long 1.16 & 1.17 cycles, but we've got both smart programmery people and smart organizery people so I think we'll work it out soon.

You are working on a revamp of the parser and WYSIWYG did this change in the last year?
Several things:
  • Wikimedia Foundation has recognized the importance of human factors to our problems of editor acquisition & retention, and is prioritizing projects to make the new-user experience and the ongoing user-to-user interaction experience less painful. Breaking down more of the technical barriers to editing at the same time should combine nicely.
  • Browsers and browser-based tools have continued to advance in power and stability; richer text editing has become standard for blogging tools, and things like Etherpad and Google Docs are really showing that you *can* do the sorts of things we want to be able to do, reliably and awesomely.
  • Most of the world is finally migrating off of IE 6, making it easier to target new tools at more modern browsers without losing our audience.
  • MediaWiki has integrated a powerful JavaScript module loader and the wonderful jQuery framework, making it much easier to integrate and deploy such tools.
The world has clearly wanted a good rich-editor solution for MediaWiki and several alternates are available that partially work; it's just finally time to solve and integrate it officially.

What is more urgent, the parser or the user interface?
The interface is more urgent, but in the long term the parser is more important. Wikimedia's projects are meant to provide reusable content, but our poorly specified ad-hoc data formats make that much more difficult than it needs to be.

You have always been happily involved in helping new developers, do you still have room for this?
I do try... helping others keeps us honest -- if we can't find the answer to your question, we know you wouldn't be able to on your own. ;) 

It proves hard for the WMF to get the MediaWiki 1.17 releases out, will this get better and why?
It's been a struggle to get MediaWiki back onto a regular release schedule after the major refactors of the last couple years and shifts in our review processes. I hope this will be worked out soon to everyone's satisfaction! 

Do you have any magic plans for growing the traffic of the WMF projects?
Driving traffic by itself is uninteresting to me; what's important is to make sure that Wikimedia is providing useful resources and that they get used by people who need them.

A lot of folks are looking specifically at growth in Wikipedia reading & editing activity in parts of the world that have had limited internet access, or limited materials in peoples' native languages; projects are under-way to improve script support for Indic languages, bring better reading & editing tools to mobile, etc.

But it's also important to move beyond the textual reference work, and put more effort into active and interactive education. Projects like Wikibooks and Wikiversity need a good strong overhaul and a focus on human interaction. 

If you had one day a week that would be yours to work on something MediaWiki related, what would it be ?
Interactive media authoring. The modern web has caught up with the goodies you used to only be able to do with Flash in terms of infographics, interactive timelines, maps & diagrams, and games and simulations; good collaborative tools for creating things that are both information-rich *and* move and respond to the user are still sorely lacking.

We've done a good job at the traditional textual encyclopedia and have dabbled in photos and video, but there's so much more we can do to support different learning styles and more explicitly educational tools. 

translatewiki.net is becoming ever more sophisticated, does this positively influence the development of MediaWiki?
The active UI translation community that built up around MediaWiki has always been a huge positive influence! Internationalization issues were what got me into working on MediaWiki's predecessors back in 2001-2002, so seeing such a strong i18n-oriented community that's actually branching out from supporting Wikipedia & MediaWiki themselves to supporting UI translations for other open-source projects makes me very happy.

The work we put into supporting localization also supports a lot of customization, which has certainly been good for both Wikimedia & third-party users of MediaWiki.

Cxu vi parolas esperanton aux nur lin legas?
Mi fakte malofte parolas esperanton pervocxe, sed jen video de mia prelego pri la rolo de Esperanto je la frua internaciegado je la softvaro de Vikipedio! (Esperanto kun germana tradukado, anglaj subtitoloj


-- brion

1 comment:

James Salsman said...

Can we please try some IBM blade servers with 24 MB cache for OLAP tasks and not buy from vendors who support regressive politicians?