1. What is in your opinion the most important new feature or update found on Red Hat Linux 9 that could 'push' users to upgrade or purchase it?
Matt Wilson: For technical users, the new Native POSIX Thread Library (NPTL) will be the most interesting addition.
For the desktop or home user, the improved menu layout, refined Bluecurve look and feel, anti-aliased Mozilla font rendering, and alpha blended mouse cursors are all features that could influence users to upgrade.
And, of course, each release of Red Hat Linux supports more hardware than the one before it. Switching our printing system to CUPS helps here as well.
2. How does Red Hat see the competition around its business? Do you mostly compete with other Unix variants or your main focus is competing against Windows/Macs?
Matt Wilson: Red Hat's main focus has been and will continue to be the enterprise. We've had great success in migrating enterprise customers from proprietary UNIX.
3. On the much discussed topic of the absense of mp3 codecs from your products: While it is easy for a user to install the needed codec manually (if he/she knows where to get them) and while it is known the GPL incompatibility with mp3, don't you think that "making it less convienient [for the user] and with less functionality only reduces the prospect of a pulling force for more users" as a Red Hat Linux user said recently? Is there any way around this limitation of Red Hat Linux 9 for future releases?
Matt Wilson: In Red Hat Linux we attempt to provide feature parity for mp3 playback by including Ogg Vorbis. If the codecs were not patent encumbered we would certainly include them. This applies to all areas of Red Hat Linux. We would prefer to enable the patented TrueType hinting algorithms in FreeType to improve our font rendering, but this is not possible.
4. Why there was no RandR GUI tool shipped with Red Hat Linux 9's XFree86 4.3? The current tool still requires you to unload/reload X in order to change a resolution.
Matt Wilson: We're planning to include RandR support in our X configuration tool in a future release. We were not certain that XFree86 4.3.0 would be released in time for us to depend on the RandR functionality. In the meantime, one can modify the screen size on the command line by using the 'xrandr' utility.
5. What features were left out of the 9 release because of time pressure or other reasons?
Matt Wilson: Extended Attributes (EAs) and Access Control Lists (ACLs). During our testing we found deadlock conditions in the code. We have the fix now and EA/ACL support should be in the next release.
6. Red Hat does a lot of work in the desktop area with Gnome and GTK+. What are the plans for the next few years regarding the [corporate] desktop? What changes need to be made to the system to look and feel more integrated to the underlying system in your opinion?
Matt Wilson: The needed changes are a matter of a thousand details, rather than a dozen huge projects. It's hard to predict exactly what will happen as much of it will come from the community of vendors and volunteers outside of Red Hat. For our part, we'll probably continue to focus efforts on integration work (bringing together open source components and our own enhancements into a unified whole), platform infrastructure (such as GTK+, D-BUS, X Window System improvements), and increasingly on applications such as the web browser and office suite. One large area that needs to be addressed is enterprise manageability; there are partial answers in GNOME, KDE, OpenOffice, and Mozilla, but something more unified and comprehensive would be beneficial.
7. Why is Red Hat Linux 9 still uses ext3 while more feature-rich filesystems like ReiserFS and XFS are out and about? While I am aware of the rare cases of instability found on these filesystems, couldn't Red Hat work on resolving these issues in order to provide a more feature-complete fs?
Matt Wilson: I would argue that ext3 isn't feature deficient compared to ReiserFS or XFS, they just have some different filesystem approaches. There is tremendous value in being able to improve the filesystem that all of our users are currently using, allowing them to continue to use their existing filesystems while taking advantage of new developments as they appear.
8. Why isn't Red Hat working together with NVidia to resolve kernel crashes and bugs that happen very often when running the accelerated Nvidia drivers on many PCs (e.g. with some VIA chipsets)? (A strategic alliance of a sort, similar in the way Apple does it, which ensures highest compatibility and testing.)
Matt Wilson: I do communicate with NVIDIA informally to provide them with assistance as they develop their drivers for new releases of our products. We are unable to provide any additional assistance on their code due to the binary-only nature of their driver.
9. Modern desktop/workstation OSes buy the needed licenses (e.g. Apple, QNX, BeIA) and they even create their own DVD applications (closed source). How about including DVD playback support on a future Red Hat Linux? And what about licensing Microsoft's Web Fonts too? Is Red Hat open regarding licensing technologies and services from other sources?
Matt Wilson: We will not include technology that prevents Red Hat Linux from being freely distributed. Including software that places these kinds of restrictions on our community of users does not help drive Open Source software.
10. Currently, no matter how I turn it, downloading RPMs from the web can create many dependancy problems most of the time. Is there any "plan" at Red Hat to find a way around this? For example, "enforcing" developers to statically link needed libraries that don't ship by default with Red Hat Linux (only for small library deps). Or to create a web site, e.g. apps.redhat.com, where people can download tried and tested RPMs from the dev. community that "work out of the box" with only a few clicks away. Any thoughts on something like this?
Matt Wilson: It sounds like an interesting idea. Though if application writers followed the guidelines provided by the LSB, you would not have dependency problems.