The purpose of the “Planet” sites is to have the blogs of people
involved in the projects, rather than to blog about the
projects. So John Fleck should continue to post away in my opinion.
Nick Petreley
predicted the success of VA Linux and Caldera.
Congrats to the Evolution hackers on today’s announcement.
(This post was originally found at http://log.ometer.com/2004–05.html#11.2)
Olsen
twins endorse the Velvet Underground.
(This post was originally found at http://log.ometer.com/2004–05.html#9)
Large group of Red Hat hackers went to see Van Helsing last night;
afterward, argument over whether Dan Williams or I had first suggested
it, and which of us should have our ass kicked. Though it was bad, I
was impressed by how over the top elaborate the whole thing was.
Scratching the surface: multiple kinds of monster, slimy alien pods,
VampireVision and WerewolfVision, Rube-Goldberg-meets-Tarzan
action scenes, machine-gun crossbow, Batman grappling hook, exploding
baby vampire gremlins, evil Oompa-Loompas. Every scene is like the
opening stuntfest from a Bond movie, only more so. Nonstop in this
vein for over two and a half loud hours of THX, and then even the
credits were overkilled. It was a certain kind of awe-inspiring.
(This post was originally found at http://log.ometer.com/2004–05.html#8)
Red Hat
Desktop is here.
The press is pretty good so far. A
few people continue to be confused and think that retail/consumer
is the same as desktop, or that Fedora Project vs. Red
Hat Enterprise Linux is a desktop vs. server split. Hopefully that
will be cured in time. Red Hat Desktop is a member of the Enterprise
Linux family and only available for purchase “in bulk” to emphasize
these points.
(This post was originally found at http://log.ometer.com/2004–05.html#4)
Apparently this
person would be happier to pay for Red Hat Enterprise Linux if we
removed open source code from the Fedora Project and included it only
in Red Hat Enterprise Linux under a proprietary license. He says we
don’t understand “value add,” but maybe this is one of those RMS is
right moments (see the joke about “freedom subtracted”).
The technical points in the article are often off-base too — hint, the
Red Hat Enterprise Linux 3 feature list was driven by the most common
enterprise customer requests, and the things on the list aren’t in
this article… but I can live with that. I’m just blown away that
he’s complaining that he could use Fedora Core 1 for free.
(This post was originally found at http://log.ometer.com/2004–04.html#30)
Glad to hear the good news.
(This post was originally found at http://log.ometer.com/2004–04.html#30.2)
Testing timestamp support.
(This post was originally found at http://log.ometer.com/2004–04.html#30.3)
I’m kind of amused by this discussion
of kernel backports.
As I pointed out
earlier, we should hand the kernel hackers a copy of the GNOME
release process. Asking vendors not to backport is conceivable if you
have regular releases. If you don’t have regular releases, then
there’s no way vendors can ever work on HEAD without backports -
because nobody can invest money to develop a feature that may not be
usable for an unknown number of years.
The “fragmentation” issue is totally bogus anyhow — backports aren’t
the same as dead-end patches that the mainline project rejected.
(This post was originally found at http://log.ometer.com/2004–04.html#20.2)
I realized today that all the talk about Longhorn is “look how cool
the technology is” — Avalon, XAML, WinFS, blah blah. This includes the
noise coming from Microsoft. There is some small amount of Aero
user experience info, but most of it describes building blocks — a
sidebar that can hold tiles, a notification feature, WinFS. The user
functionality shown for WinFS is not that exciting; search, sort, and
filter. Yay.
People seem to be saying “here’s a ton of code — imagine the
user experience possibilities!” but not really talking about any
specific concrete enhancements the OS will include. How do real user
tasks get better: calendars, email, document collaboration,
whiteboarding, etc.? How does this thing help me get my work done?
If I’m not a developer, why should I care about Longhorn?
I don’t doubt that Microsoft has some ideas, but I think it’s
interesting that they aren’t talking about the OS in these terms.
Am I just missing the relevant information?
The Longhorn
home page links to a bunch of developer stuff.
More importantly, why is GNOME talking about Longhorn on Microsoft’s
terms? We’ve been talking about “competing with Longhorn” in terms of
matching all the technical gizmos. But shouldn’t we instead be talking
about the most expedient way to get similar user experience
enhancements? Our desktop offers similar functionality to Windows XP
today, but is implemented completely differently. But for Longhorn,
we’re talking about competing with the implementation before we even
know the functionality.
I’m sure everyone will agree with the principle that we should plan
the Linux desktop by starting with what we want users to be able to
do, and ending with how to implement it. But in thinking about
Longhorn, at least I haven’t been doing that.
(This post was originally found at http://log.ometer.com/2004–04.html#20)
Glynn,
unfortunately fascinating isn’t the only metric. 😉
I should clarify exactly what I think James Gosling gets wrong: he
says the Gnome world has “formless dreads,” I think the concerns are
in fact very detailed, rational, and well-understood, though there’s
some noise and some inability to post legal advice that may make this
hard to see.
One thing I skipped in my post is the discussion of the pragmatic
reasons why open source has to be defined as it is, and why its
definition enables a particular development model and set of customer
advantages. This model and these advantages are genuinely lost if you
use an “almost but not quite open” license. It’s not a matter of
religion.
Dr. Gosling also compares the Java license to the GPL, but it’s worth
noting that none of the GNOME or common Linux libraries are under the
GPL; they’re under the LGPL or less-restrictive licenses. In other
words, if we compare apples to apples, when he says:
Unlike GPLd software, the Java sources don’t come with a viral
infection clause that requires you to apply the GPL to your own
code. But the sources for the JDK do come with a license that has a
different catch: redistribution requires compatibility testing.
In fact the GNOME and Linux platforms don’t come with this
“viral infection clause” — only the applications have that.
And you aren’t going to use an application in your own code. So
comparing the JDK license to the GPL is wrong. You have to compare it
to the less-restrictive licenses used in the platform.
I just remembered that I wrote an article about the SCSL 5 years ago,
before I worked at Red Hat: here
it is, still on my web site. I no longer agree with some of what I
said and how I said it, in fact the article sort of sucks, but it’s
interesting how little has changed in some respects.
(This post was originally found at http://log.ometer.com/2004–04.html#14)
