X.Org 7.5 Released. Wait, Nope!
Written by Michael Larabel in Software on 1 April 2009. Page 1 of 1. 52 Comments

Today X.Org 7.5 with X Server 1.7 is scheduled to be released, per the release schedule that Daniel Stone proposed last year. X Server 1.7 includes X Input 2 (a.k.a. Input Hotness) and Multi-Pointer X is now enabled by default (it has been in the master branch for about a year, but it has been disabled due to X Input 2 missing). This key piece to the open-source Linux desktop also features E-EDID support, the X Server no longer needing to symlink to Mesa sources, and a horde of bug-fixes. Aside from an updated X Server, X.Org 7.5 will include various updates to different input and graphics drivers.

While the 7.5 release is not yet uploaded to the X.org web-site, the latest X.Org 7.5 release schedule says today's the date. With X.Org being such a critical part of the free software ecosystem, it would be unthinkable that there could be any major delays with the release or to delay the release without any notice or revised schedule, right? Like every day, we will be monitoring the RSS feeds of the different Git repositories, the different IRC channels, etc as we await the X Server 1.7 / X.Org 7.5 release or any important announcements.

Okay, well, X.Org 7.5 is not actually coming out today... It is indeed April Fools' Day, but originally the X.Org 7.5 / X Server 1.7 release was set to come out today, and this article is indeed truthful. Daniel mentioned in the release schedule, "1st April 2009: Release 7.5, makes good April Fool's story. If not, Phoronix gets to run '7.5 released ... nah, X.Org didn't release on time again! April Fool's! Ha!'. Either way, everyone's a winner." This article is not -- by any means -- to laugh in the face of free software developers due to yet another X.Org delay, but to illustrate a problem.

If you head on over to the X Server Git repository viewer, do you see an X Server 1.7 branch? Nope, there is no server-1.7-branch to be found. While X Server 1.6 did not end up getting released by the end of 2008 as Intel had proposed (it actually got released at the end of February), the original X.Org 7.5 release schedule called for it to be branched back in February. Two months later, we have yet to see any 1.7 branch! The feature freeze for X Server 1.7 was supposed to be at the end of January, but we have yet to see that announced either. On top of that, there has not been much (if any!) communication publicly in regards to reviving the X.Org 7.5 / X Server 1.7 release schedule. Realistically we would predict that X.Org 7.5 / X Server 1.6 is at least another six months out.

This is not the first or second time there has been a problem releasing a new X.Org or X Server on time, but it has become a trend. In fact, last year when just delivering a bug-fix release it came 212 days late. Yes, over seven months late for a point release. The X.Org 7.4 / X Server 1.5 release was also several months late. There is a problem with releasing on time. Whether it can be blamed on a lack of release management, development manpower, or other factors, that is a different matter.

Regardless of whether or not development manpower is the true cause for these development delays, they sure could use more developers. There are many active projects going on within the X.Org ecosystem from input with Multi-Pointer X to graphics drivers with kernel mode-setting, in-kernel memory management, etc. Beyond the topic of release management, there also appears to be a lack of clear focus within the development community and quite a bit of fragmentation. Just look at the ongoing TTM vs. GEM situation last year and other flip-flopping and the quick abandoning of different projects.

To help in this situation, there are many companies that could step up and contribute more to the development of the X Server and related packages. Red Hat and Intel in particular contribute a lot, but not a lot of upstream work can be attributed to Canonical, for example. It is just not the distribution vendors, but more hardware vendors could be involved in the X.Org process too. Dell, for instance, ships many Linux desktops, netbooks, and notebooks, yet they have not gotten that involved with X.Org at all. IBM is a very big contributor the Linux kernel, but since Ian Romanick joined Intel, not a lot (if any) ongoing X.Org work can be attributed to this company. Just recently, Novell even let go of one of their main X.Org developers that was also responsible for quite a bit of the work on the RadeonHD driver. For some statistics, you can look at our Contributors to the X Server article from last year.

Beyond just the companies that directly or indirectly capitalize upon the work of X.Org, there is more that individuals and other interested parties can do. Heck, at Phoronix Media we would be happy to setup an automated test farm with our flagship testing software and help more in the way of tracking the performance of the X Server and the different drivers. Bounties could also be considered if that is what would really help in returning to scheduled, quality releases. If you happen to be a student, a good way to get involved with X.Org development is through the X.Org with Google Summer of Code. There are also numerous other ways to get involved whether it be documentation, helping with their Wiki, working to close bugs, etc.

While this situation has gotten worse over time, it is not exactly a new problem. Back in September of 2007, Sun's Alan Coopersmith expressed concern over the degrading quality of X.Org releases. Beyond the releases coming late, blocker bug lists are not being cleared, no one is ensuring that the entire package set is buildable on at least one platform, no one is making sure XTS runs and successfully passes on one platform, and that the documentation is being updated and released. The status quo is clearly not sufficient.

Without a reliable X Server, the rest of the Linux desktop can falter, so let's turn this problem around.


About The Author
Author picture

Michael Larabel is the principal author of Phoronix.com and founded the site in 2004 with a focus on enriching the Linux hardware experience. Michael has written more than 10,000 articles covering the state of Linux hardware support, Linux performance, graphics drivers, and other topics. Michael is also the lead developer of the Phoronix Test Suite, Phoromatic, and OpenBenchmarking.org automated benchmarking software. He can be followed via Twitter or contacted via MichaelLarabel.com.


Related Articles
Trending Linux News