(Back from nose to the grindstone, Spring at home is busy and end-of-quarter targets I’ve been working on for Red Hat are nearing completion. Appears that blogging and tracking email lists has fallen a bit to the wayside.)
From Wednesday 08 April to Friday 10 April, I was in fabulous and mildly-rainy San Francisco to attend the Linux Foundation Collaboration Summit. I hadn’t spent that much time in Japan Town, it was cool to walk through the streets and catch bits of a Japanese-in-San Francisco history walking tour signs. The hotels and conference site were all within a few blocks of each other.
The summit was a meshing of multiple mini-conferences, including an embedded Linux summit. This brought a wide range of Linux-related developers to one location and gave them reasons to intermingle. The first day I was there (Wednesday) was a middle-of-the-summit day for some, and a chance for all to pause and take part in general interest topics in the main hall. These included a final three-way discussion with Jim Zemlin of the Linux Foundation, Ian Murdock of Sun, and Sam Ramji of Microsoft, titled, “Why Can’t We All Just Get Along?”
As usual, I ran in to old friends (e.g. Elena Zannoni, still Queen of GCC tools, and Pete Graner, still one of the best engineering managers you’ll ever know) and made new ones (e.g. Jeff Osier-Mixon, who was there pimping Montavista’s new embedded linux community, Meld.) Jeff and I talked for a long while about community affairs, and ‘best community practices’ was a theme for my time spent there. I got the down-low invite to the Community Leadership Summit, since announced.
One role the Linux Foundation has been filling better in the last few years is being a catalyst between Linux developers, especially kernel, embedded, etc. and the companies that employ the contributors or deeply rely upon the contributions. I overheard a few embedded Linux developers saying that the Ottawa Linux Symposium (OLS) was down their list this year and got cut with travel budget considerations. That doesn’t mean OLS is not relevant, but that it has some competition in meaningful Linux developer conferences. This is more a recognition that there is a gap to be filled here, one that a foundation traditionally fills as a neutral entity between otherwise competing interests, and the Linux Foundation is filling that gap. This creates a gravity that can draw in or away from independent efforts.
Since simply networking with other free software folken isn’t enough, I was also there to talk on a panel about measuring community contributions, which I covered in another article. I also signed up to do what I could to help Tim Waugh prevent tragedies at the OpenPrinting Summit, a task which I and Tim largely failed at.
The OpenPrinting group met to work on improving printing in Linux with the typically laudable goal of “just work.” They have come to a solution they feel is a good interim step toward an undefined future. They are going ahead with plans to provide an automagic way of downloading and using proprietary printer drivers from within the printing dialogs. This situation is analaguous to providing a way to access proprietary multimedia codecs, which Fedora tried to solve with CodecBuddy, one of the greyer parts of the project’s history. I felt that mistake was made and lesson well (l)earned.
This automagic tool gives the printer manufacturers a way out instead of doing the right thing, that is, working on opening printer drivers or at least opening their standards so equivalent free drivers can be implemented. I was just lurking in person while Tim was on the conference call, but I had to interrupt the meeting to make sure all there understood. Their plan to write an application with the sole purpose to download and use/install proprietary code was not going to have an easy time passing Fedora package review. In the end, the only thing it can deliver in Fedora are free drivers, and those are or should be packaged and already in the distro.
Even if the tool does make its way to Fedora, it still may not make it in to the next version of Red Hat Enterprise Linux, and even if there, the printing companies are still likely to negotiate individually with Red Hat to get proprietary drivers available in RHEL. Nobody there seemed to have a problem with this.
I did not walk away feeling the state of printing in Linux was improved with this situation, although I’m sure millions of users will disagree with me. Don’t those users realize this is a bandage over the situation that is made this way because of the printing manufacturers? A bandage does not heal a wound or remove an infection. Instead of a holistic solution that treats the whole body, we have a festering wound that is going to require surgery one day.