Linux Telephony: Not so short Overview
Google engEDU
1 hr – Jul 25, 2007
Google Tech Talks
July 19, 2007
I will talk about voice services in packet-switched (VoIP) and circuit-switched (traditional telephony) networks, using Linux and commodity hardware. This will include short introduction into telephony in general, VoIP and PSTN type of telephony, plus overview of VoIP services (and popular opensource software packages used to implement those services) and kernel interfaces that used in that packages. I will cover the topic of connection to the Public Switched Telephony Network using digital interfaces (E1/T1), kernel support for it in framework called "zaptel", and userspace integration issues. The conclusion is about current trends in Linux-based and open-source telephony.
This is a second run of my presentation which I gave on the Ottawa Linux Symposium’2007. Read the rest of this entry »
Sphere: Related ContentOSS Speaker Series: The State of the Linux Kernel
Google engEDU
1 hr 22 min – May 1, 2007
Google Tech Talks
May 1, 2007
Andrew will present a broad look at the trends in recent changes in the Linux kernel: what areas of kernel functionality are people working on, and what changes can we expect to see over the next year ?
This feature-by-feature walk-through will be tied to an examination of the motivations of the kernel contributors: why do particular individuals and organizations choose particular things to work upon ?
He will also examine areas of the kernel which are arguably suffering from some neglect, the reasons for this and some possible corrective actions which might be taken.
Andrew will finish with a discussion of the importance of individual testers to the kernel development effort, as well as a look at the steps which a kernel tester should take to maximize the effectiveness of his or her contribution to the kernel. Read the rest of this entry »
Sphere: Related ContentHardware/Software Hacking: Joining the Real and the Virtual
Google engEDU
1 hr 6 min – Jul 31, 2007
Google Tech Talks
July 31, 2007
Software developers usually confine themselves to working entirely within the runtime environment of a computer just pushing around bits and pixels. Even virtual worlds such as Second Life exist only in the confines of our CPUs.
On the other hand, hardware hacking has really taken off in recent years and there are now magazines such as MAKE devoted to modifying everyday objects. It’s a lot easier than software jockeys may expect, and this talk will begin with an entertaining exploration of simple ways to get started with linking a computer to real-world objects.
But what happens when you knock down the boundaries between the real world and a virtual world? The talk goes on to show specific techniques and examples for linking real-world objects into the Second Life environment so that changes in the real world can be reflected in SL and vice versa.
Jonathan Oxer is founder of Internet Vision Technologies, author of "How To Build A Website And Stay Sane" ( www.stay-sane.com) and "Ubuntu Hacks" (www.ubuntuhacks.com), is currently President of Linux Australia, convened the last 5 Debian Miniconferences, and sits on various boards and advisory panels for groups including Swinburne University and the federal e-Research Coordinating Committee. Read the rest of this entry »
The Reiser4 Filesystem
Google engEDU
1 hr 3 min – Feb 6, 2006
Google TechTalks
February 6, 2006
Han Reiser
Hans Reiser was concerned that hierarchical and relational naming systems add structure not inherent in the information, and that boolean algebra fails to represent structure, and so he develeoped a set of semantics which attempt to match rather than mold structure.
He then founded Namesys in 1993 when he went to Russia and hired a small team of programmers to implement the storage layer for these semantics, known as the Reiserfs filesystem for Linux. He spent 5 years arguing over algorithms on evenings and weekends, and working day jobs in Silicon Valley, and then finally the code started to work well enough that he could quit the day jobs.
Then he took all the code, chucked it, and rewrote it from scratch to reflect what he had learned the first time doing it. The new code is called Reiser4.
ABSTRACT
The ReiserFS project aims to add support for semi-structured data querying to the filesystem namespace. Reiser4 is the storage layer for this. It stores all files in a dancing (not balanced)tree, and is currently the overall fastest filesystem for traditional filesystem usage patterns. Read the rest of this entry »