Sunday, October 23, 2011

The Delicate Art of Falling on your Sword or Find Your Way Through Contrition

Ubuntu Studio 11.10 "Oneiric Ocelot" is released and I offer these simple, direct reflections about it and the future.  As usual, it seems my tenor is positive and upbeat, despite the somewhat dismal context.

Past Tense
Epitomizing simple and direct statements: this is kinda a 'meh' release and it's my fault.

Fundamentally I believe this to be a stable and functional release.  But it has an unfinished or unpolished look due to the XFCE transition remaining incomplete.

The team suffered some difficulties this cycle, but it is still my responsibility and, therefore, my fault.

However, reflections have given way to some insight on deficiencies and yielded some plans to action that should improve our state.

Among these deficiencies we experienced:
  • our goals and scope were too large
  • our efforts were unfocused and not efficient
  • started late in development cycle
  • most new contributors failed to have an impact
  • code changes lacked follow through

Present Tense
So, how will we improve these concerns?

Goals/scope to large - We were trying to fix everything and ended up almost completing nothing.  We are narrowing our focus for the upcoming release, focusing on what we feel are the minimal priorities in order to yield a quality release.

Unfocused, inefficient efforts - It is hard to focus on what to fix and how to fix it without a plan.  Using blueprints, creating specifications, developing and assigning a sequential plan of action should help focus our attentions where they are needed and give us the efficiency these goals deserve.  This also provides the structure and support for others to assist.

Late Start - This is easy...start earlier! And having a well thought out and documented plan helps.  The small scope and blueprints with specifications and steps are inspiring and development has already begun!

New contributor's lacking impact - This may have been the most frustrating deficiency of all.  Obviously the public solicitation of contributors is working, but the tools are lacking to engage and support them.  Much like the blueprints, a wiki page for future contributors is being created that includes a listing of topics requiring contribution, background on those topics, and steps necessary to complete them.

Code changes uncompleted - This could be summarized as the lack of getting our bzr branch updates pushed to the repository.  The fix is a double pronged attack; working more directly with sponsors and developing repository permissions for specific team members.  The former will yield results quicker and the later will allow more control but take longer to acquire.

Future Tense
So what are our goals?  There are few and they are prioritized roughly in the following descending order:
  • updating website [1]
  • -lowlatency kernel shipped as default [2]
  • completing the XFCE transition (including icons and theme)
  • live dvd [3]
  • updating the LightDM theme
Updating the website may be a surprise as the first item, but the current website makes it seem like the project is abandoned.  The update is already in progress and should take a nominal amount of time to accomplish.

Shipping a tuned kernel for audio work is befitting a multimedia distribution, hence the second positioning.  This one probably has the most difficulty to it and will almost certainly take a lot of navigating to complete.

Completing the XFCE transition should not require much explanation.  I imagine we will lean heavily on what the Xubuntu team has already done so we can complete this task as efficiently as possible.

A live dvd has been a desire for a while.  This would ideally include dropping the alternate installation image but hopefully also include Edubuntu's ability to allow the user to select specific components for installation via a GUI.  Think of the later as the ability to select specific work flows [4] during installation, e.g. recording a band, creating music with synths/sequencers, podcasting, DJ'ing, film making and creating videos.

Updating the LightDM them should also not require much explanation.  The LightDM used by Ubuntu Desktop looks nice so I am once again going advocating we steal as much as we can to make our efforts efficient.

This may seem like a straightforward, obvious, and diminutive list.  It is!  That's the point.  These are the goals that we think we can accomplish that will make Ubuntu Studio 12.04 LTS release something of which to be proud.

Lastly, I alluded to some team difficulties and attrition.  We have suffered that.  But we have also gained some very talented assistance as well.  I hope to discuss them and their contributions next post.


Thursday, October 13, 2011

Ubuntu Developer Summit

I'm going to Ubuntu Developer Summit this November! I'm pretty excited about this and slightly nervous.

I was surprised about a month ago when I received an email saying that my sponsorship was approved. Funny thing is that I didn't apply for sponsorship, perhaps someone else did.

Thank you "someone else". I would like to know who you are.

I've been overloaded with work for the past three or so months so I haven't thought too much about this other than making the initial arrangements. Therefore, I have given very little consideration to what I hope happens, however, offhand I would say that I want to:
  • learn, learn, learn (those who know me also know that I always want to learn stuff)
  • get the -lowlatency kernel into the repos so it can ship by default in Ubuntu Studio
  • talk to someone (Mark maybe) about getting a developer support for Ubuntu Studio
Am I crazy to expect these things? I view it as I would be crazy to expect these things to happen without doing something about it!

Oh, I've been growing my hair for the past two months or so in preparation for UDS.  This is me now:
Hirsute Scott (aka Bandersnatch)
Ever since I was in the Marine Corps I have cut my hair every two weeks (sometimes less) with clippers and no guard.  It's taken a little bit to get used to after not having hair for so long.
Scary, No Hair Scott