[read this post in: ar de es fr it ja ko pt ru zh-CN ]
It has a killer refresh rate.
P6 chip. Triple the speed of the Pentium.
Yeah. It’s not just the chip, it has a PCI bus. But you knew that.
Indeed. RISC architecture is gonna change everything.
Yeah. RISC is good.
July 30, 2006
July 27, 2006
As free speech goes it can be used as a tool to bring people together to find and fix problems or as bludgeon for those who care only about being right but not true. If the old saying goes the first casualty of war is the truth then I don’t think there is a day in my life where the world was not embroiled in some kind of war. I grew up innocent, reading a lot about the civil rights movements which swept America and was most prevalent right before my birth. Then I grew up and realized prejudice was still there in all sorts of forms. Worse still as I learned about the world outside of my own domain I realized we were even further from the tolerance and camaraderie that I thought was the natural order of things.
It seems in any group I have observed or been a part of there is always the heckler in the crowd, someone there to incite and play to base emotions. It gets worse when that heckler is a leader. Hate is an easy rallying point, a sure way to motivate people for whatever purpose. It allows the perpetrator to not have to think but still produce results, even if those results are chaotic.
We need a world that thinks, not one that shouts. We need solutions not mere rhetoric from all sides of the isle.[read this post in: ar de es fr it ja ko pt ru zh-CN ]
July 25, 2006
I got my first taste of Sugar today, the software platform being developed for the OLPC project. I was testing out some locking patches in D-Bus and they had an issue with the updates in Fedora so I decided to give it a spin to see if I couldn’t shake some bugs out of D-Bus. I must say the enviornment was simple to setup requiring me to run these few commands:
I ran into a few issues such as gnome-python not being in the module set but it was all easy to fix. Git was really nice for generating a patch which I sent back to the devs. It took all of 10 minutes to compile the 17 or so modules and I was off and running. The devs really put some work into getting a nice envionment to hack in. I simply went around poking in python files to try and fix the issues. In the end it came down to not inializing threads that was causing problems. I did find a bug in D-Bus where a pending call mutex was being initalized to NULL instead of the dummy mutex which I haven’t been able to track down yet. In any case it was easy enough to jump into and was really helpful in identifying bugs in D-Bus along with giving me ideas on where the python bindgings should go next.[read this post in: ar de es fr it ja ko pt ru zh-CN ]
July 24, 2006
There is an issue when building the GLib bindings. As they stand now they require a connection to the system bus in order to retrive introspection data so that it may create convinience functions for the bus methods. We are investigating the best way to handle this, either by distributing the xml introspection file and only generating it on dist or by getting rid of the convinience functions if nobody uses them. For now you can download the dbus-bus-introspect.xml file for 0.9x servers and place it in the tools directory in the dbus-glib package.[read this post in: ar de es fr it ja ko pt ru zh-CN ]
I did some minor releases of D-Bus and bindings. D-Bus got some important bug fixes related to the pending call refactoring. The bindings got mostly build and install fixes.
July 19, 2006
I just turned 29 a few minutes ago. I got a call from my twin sister. She’s at the Yankees game right now – Go Yanks!!! Just in case you didn’t catch it, she turned 29 today too. I have a lot of work to do this week so I won’t be able to enjoy my birthday until this weekend but then I’ve got plans to check out a lake turned into a beach I found last weekend and then go check out the Limelight karaoke bar in the evening. Oh and my talk has been approved for aKadamy so I am looking forward to hanging out in Dublin this fall. I’ll drink to that.[read this post in: ar de es fr it ja ko pt ru zh-CN ]
July 17, 2006
For awhile now whenever a person wrote about DBus it was always up in the air how to spell it. Was it D-BUS, D-Bus, DBus, DBUS, d-bus or dbus? The original correct spelling as conceived by “oh my God it’s” Havoc Pennington, was D-BUS. Here is what he had to say when we caught up with the ever elusive Pennington:
The name D-BUS was essentially a joke; I seem to recall a lot of stupid puns being exchanged on IRC. (e.g., you will notice we still have dbus/bus/driver.c…) Putting it in all caps made it more amusing perhaps.
The devs finally got together to put this burning question to rest. How do you spell d-bus? The overriding majority answered, “I really don’t care” and so after little thought on the subject it is time to announce that dbus will officialy be written as:
So it is, so it shall be written. [read this post in: ar de es fr it ja ko pt ru zh-CN ]
There is always that time when parents have to let go and have their children fend for themselves. Today we release the D-Bus 0.90 API/ABI frozen “Empty Nest” family of packages. For this release D-Bus has pushed its bindings out of the nest so that they can grow and evolve on their own.
This will be the last coordinated bindings release for the foreseeable future. The Qt bindings have been integrated with the latest Qt in their SVN servers so it is not in this release set. We do gain two more future bindings, Perl and Java, which are being considered as additions to the D-Bus Project.
Released packages can be found here:
Core – http://dbus.freedesktop.org/releases/dbus-0.90.tar.gz
Python Bindings – http://dbus.freedesktop.org/releases/dbus-python-0.70.tar.gz
GLib Bindings – http://dbus.freedesktop.org/releases/dbus-glib-0.70.tar.gz
Java 1.5 Bindings – http://www.matthew.ath.cx/projects/libdbus-java-1.8.tar.gz
Java 1.4 Bindings – http://www.matthew.ath.cx/projects/libdbus-java-0.4.tar.gz
July 14, 2006
The patches are in. The bindings are split and we are ready to release ABI/API frozen D-Bus 0.90 on Monday. We will also be releasing D-Bus Python 0.70 and D-Bus GLib 0.70 along side. Some other bindings maintainers have also expressed interest in releasing their bindings at the same time. I will announce them in my Blog.[read this post in: ar de es fr it ja ko pt ru zh-CN ]
July 13, 2006
I’ve been getting some flac for removing the dbus_connection_disconnect symbol. It seems to be breaking a couple of modules in GNOME. That method has been deprecated in D-Bus for over a year. It was replaced by dbus_connection_close. There was a big TODO item which stated simply, “remove all deprecated methods before 1.0″. Luckly there was only one.
Now on to the good news. This all means that 1.0 isn’t that far away. Once 1.0 happens and distros start using it, GNOME releases can just rely on the installed D-Bus instances. The bindings are a different story as the python and GLib bindngs are due for an overhaul though we hope to stay as close to the current public API as possible. How do we make sure D-Bus 1.0 comes quickly and smoothly? By your help in finding bugs and fixing them. I have made this call in the past to various degrees of success. We got Ross for awhile and we still have the Roberts from Collabra who work on D-Bus as much as I do. Thiago from Trolltech while doing the Qt4 bindings has been a big help in identifying and fixing stuff in the core. The point is now that we are winding down we need more help, not less so join us in #dbus on freenode.net and join the mailing list. Your input is welcomed.[read this post in: ar de es fr it ja ko pt ru zh-CN ]