The newsletter of the UK Unix Users Group
Firstly I want to thank everyone who kindly sent in their subscription payments so promptly. We have received a record number of early payments. Those remaining outstanding will be chased next month and any still outstanding after March will not receive the June Newsletter.
The Winter conference and tutorial held on 18th & 19th February was very successful. Tutorial attendance numbers were high and although the conference bookings were slow to come in at first many late bookings brought the numbers up. From the feedback on the questionnaires the event was well received. The papers from the conference will shortly be available on UKUUG web site.
The next event is the Linux 2003 Developers' Conference being held in Edinburgh from 31st July - 3rd August , details, including a web form for potential speakers' submissions is available at: http://www.ukuug.org.uk/events/linux2003
Please note: copy date for the June issue newsletter is: 23rd May 2003
Thursday 31st July (Tutorial)
Friday 1st August - Sunday 3rd August (Conference)
Venue: George Watson's College, Edinburgh
Our 2003 Linux Developers' Conference will be held in Edinburgh at George Watson's College from Friday 1st August to Sunday 3rd August. It will be run along similar lines to last year and we anticipate starting from about 11am on Friday and finishing by around 3pm on Sunday. Preceding the conference, Thursday 31st July will be a Tutorial Day.
The conference website (http://www.ukuug.org/events/linux2003/) will be updated as further information becomes available.
Proposals for conference papers or discussions (BOFs) should be submitted through the website.
New this year, there will be some developer rooms available. Some have internet connectivity and some already contain machines suitable for web browsing and connecting to remote applications. So you may also propose small-scale talks, discussions and/or hands-on tutorials/demos/coding that make use of these facilities and which will be scheduled to run alongside the main conference programme.
The programme and booking form will be available in late April.
The GNUWin II CD that comes with this issue is the result of suggestions made at the end of last year at a time when this project gained considerable publicity on the news sites. The project site is here: http://gnuwin.epfl.ch/en/
It is simply a collection of the best free software applications many of which are familiar to Unix users, but compiled for Windows. So, for instance, you will find here versions of OpenOffice.org, the Gimp, Mozilla, and XEmacs compiled for Windows. Also included are Apache, PHP, MySQL and much more.
So pass this on to your Windows-using friends and family, and any sceptics who ask "Where are the applications?". They will be impressed.
Regarding a previous CD, Mike Smith writes:Attendees of the UKUUG 2002 Summer conference will have enjoyed the Knoppix CD which was distributed there. A revised version of Knoppix 3.1 was released in January if you're interested in an update. There's an English language version available for download.
Suggestions for future newsletter CDs will be gratefully received.
XOM: When SAX is too stringy and DOM is too lumpy
A Presentation by Elliotte Rusty Harold at Room 822, The Institute of Education at 7pm on Wednesday 19 March 2003.
Note: pre-registration is required for this event
This is a joint event hosted by UKUUG in conjuction with XML UK.
It is generally recognised that the standard APIs for processing XML have their drawbacks. SAX, though efficient, is hard to use, and DOM can be 'just plain ugly'.
XOM is a new XML object model developed by Elliotte Rusty Harold. It is an open source (LGPL), tree-based API for processing XML that strives for correctness and simplicity. XOM implementations are available for Java and Python.
In this 90 minute presentation, Elliotte will explain the philosophy behind XOM and give some demonstrations of how its use can boost productivity and lead to cleaner, more understandable code for XML processing.
Who should attend? -- Anybody and everybody involved in the programmatic processsing of XML.
Elliotte Rusty Harold is an adjunct professor of computer science at Polytechnic University in Brooklyn, but is probably better known for his Cafe au Lait http://www.cafeaulait.org and Cafe con Leche http://www.cafeconleche.org web sites dedicated to the practice of Java and XML. He is the author of numerous books, most recently XML in a Nutshell from O'Reilly and Processing XML with Java from Addison-Wesley.
Register by contacting Yvonne Vine (XML UK administrator) email@example.com
Admission is FREE to XML UK and UKUUG members who register in advance.
Places are allocated on a strictly first-come, first-served basis.
A conference on Open Source in Education will be held at Anglia Polytechnic University's Danbury Park Conference Centre on 4th April 2003.
There will be speakers from the DfES, Ultralab, netproject, RedHat and SuSE, as well as seminars on the practical implementation of Open Source solutions in schools by Chris Dawkins (Felsted School), Martin Williams (Powys LEA), Ian Lynch (The Learning Machine), James Wallbank (RTI) and a session by Phil Driscoll on LTSP.
Details of the event can be found here: http://www.mpowernet.anglia.ac.uk/conference/
USENIX and ACM SIGMOIBLE are jointly sponsoring the First International Conference on Mobile Systems, Applications, and Services. MobiSys 2003 is a new forum for presenting the best cutting-edge research on supporting, enabling, and coping with mobility in systems software, applications, and services.
Presentations, tutorials, demo & poster sessions, and BoFs will cover the latest innovations in many important areas, including:security location management application support mobile architectures sensor networks energy management analysis of mobile networks application mobility systems techniques for solving mobility problems
Tutorials on:An Intro to Wearable Computing Mobile Networking 802.11 Wireless Network Security Programming Wireless Sensor/Effector Networks of TinyOS Motes
Keynote: Bob Brodersen of the Berkeley Wireless Research Center and the University of California at Berkeley
MobiSys '03, May 5-8, 2003, San Francisco, California http://www.usenix.org/mobisys03
AUUG Systems Administration Symposium
Theme: New Challenges in System Administration
This is the first call for papers for the first AUUG System Administration symposium to be held on April 9th in Melbourne at the Australian Industry Group's conference facilities and themed to try to canvas new challenges in Systems Admin being brought to us by new technologies or new application of existing technologies.
AUUG welcomes proposals on any subject related to systems administration in general or the theme of future systems administration in particular. Some ideas for potential topics are presented below but these are by no means intended to be exhaustive or limiting. In particular, those with expertise in specific areas or generalists who wish to speak outside these areas should not feel constrained from responding.Open Source: how is it changing the playing field... Database: open source d/bs vs. established commercial players, oracle 9i-rack in the field New PC I/O technologies: PCI-X, Infiniband, HyperTransport (aka LDT) Integrating PCs and UNIX systems: where are we? Storage: how will new technologies like firewire, iSCSI and Infiniband change NAS, SAN, and direct attached storage. Backup: where are tape technologies headed? will they catch disks Networks: zeroconf.org (Apple Rendezvous product) Communications: 10gig ether, 1gig ether, wireless, QoS, VoIP, VPNs Clusters: 6 oxen vs. 1000 chickens? does it work? Spam: new challenges, new solutions? (Bayesian filtering?) LDAP: The promise of an open directory meets vendors and implementors, where is single sign on? (Is it dead?) Security: what's new? what's old? how do you get there from here... secure PC platforms and their implications for us... Capacity planning: how to do it for clusters and distributed solutions Web & e-Everything: new trends, old trends.
Those of you interested in participating are invited to supply an abstract of your presentation of 100-500 words and possibly a short biography if you feel that will help in your successful selection. Papers are encouraged but not mandatory for the symposium proper and will only be published online on the AUUG web site for conference delegates to access there.
Please forward all proposals/abstracts to firstname.lastname@example.org by Monday 24th March, 2003
Change of Date
This year the AUUG Annual Conference will be held in Sydney, Australia, 3-5 September 2003. The Conference will be preceded by three days of tutorials, to be held on 31 August, 1 and 2 September 2003.
You will note that the conference was originally scheduled for 10-12 September. We have since discovered that USENIX is holding the BSDCon in San Mateo in this week. We have decided to bring this date forward by a week.
Information on the conference can be found at: http://www.auug.org.au/events/2003/auug2003/
This includes a copy of the Call for Papers, as well as details about Sponsorship Opportunities for AUUG 2003. Should you wish to discuss either, please contact the AUUG office on 1-800-625 655 or +61-2-8824 9511.
Please check the AUUG site regularly for any updates. We look forward to seeing you at AUUG 2003.Liz Carroll -- AUUG Business Manager
The LISA '03 Program Committee invites you to contribute your ideas, proposals, and abstracts for refereed papers, invited talks, panels, Guru-is-In sessions, and Work-in-Progress reports.
We welcome submissions that address all facets of the practice and theory of system and network administration.
The Call for Participation, with submission guidelines and sample topics, is now available on the USENIX Web site at http://www.usenix.org/lisa03/
Submissions are due by April 21, 2003.
P.S. The Keynote Address will be given by Paul Kilmartin, Director of Systems Administration of eBay.
This is your conference. You can participate in the planning and contribute to LISA's success by submitting a proposal or making a suggestion. Remember that experts and old-timers do not have all the good ideas.
We look forward to hearing from you!
Sincerely,AEleen Frisch, Exponential Consulting -- LISA '03 Program Chair
SAVE THE DATE
LISA '03, October 26-31, 2003, San Diego, California
The annual LISA conference is the meeting place of choice for system and network administrators. System administrators of all specialties and levels of expertise meet at LISA to exchange ideas, sharpen old skills, learn new techniques, debate current issues, and meet colleagues and friends. People come from over 30 countries to attend LISA. They include a wide range of system and network administrators working in the full spectrum of computing environments--large corporations, small businesses, academic institutions, government agencies, and so on--as well as many full- and part-time students.
A netproject and Computer Weekly event: 'Linux for Business' Conference & Exhibition
10th June 2003, London
'Linux for Business' is a must attend conference and exhibition for IT managers and strategic decision makers looking to move toward open source as a solution, or for those who have already made the move to an open source infrastructure and are now looking for further services in the areas of security, hardware, software, employee education and strategic development.
Conference sessions will be crammed full of case studies presented by those who have actually done it.
Speakers include:Paul Martin, IT Director, Nottingham City Council: Implementing an Email Service using GNU/Linux and OSS Paul Friday, Head of IT, West Yorkshire Police: Building a Secure Open Desktop Architecture for the Police Gareth Lloyd, IT Director, Hill House Hammond: Deploying GNU/Linux in the High Street
Exhibitors include: Hewlett Packard, IBM and NEC.
For further details or to submit an on-line registration please visit: http://www.netproject.com/conference/lfb
Making the move to GNU/Linux and Open Source - What to do next
11th June 2003, London
The principals of netproject and associated consultants have unrivaled experience working in the area of Unix and GNU/Linux since 1980 in a wide range of projects. This wealth of knowledge and experience will be made available to the delegates of this workshop and will answer:Why should organisations go through the change? What are the risks? What are the risks of not changing? What are the first steps to take?
Speaker sessions will include:Establishing the Business Case for change Open Source and Security Software tools to enable the migration of desktop applications Running legacy applications from the Linux desktop
For more information or to book places on any of these events contact: email@example.com +44 (0)20 715 0072
There really could not have been a better choice for the subject of this year's Winter Conference Tutorial than IPv6. It seems to have grown up, settled down, and is starting to tug at our coat tails demanding attention, and I'm sure every attendee appreciated the delivery on a subject which is likely to be pervasive in our field before long.
So now we have a great topic, what is needed is someone who knows IPv6 inside out. Someone who won't give us "I'm not sure of the answer to that", or "I'll get back to you", when you ask them questions but delivers that warm feeling of deep knowledge and real interest - and Dr Tim Chown was just that. He's been involved in more IPv6 deployment projects than I've had hot dinners, and also leads the University of Southampton's pioneering IPv6 activities amongst other things.
It cannot have been an easy tutorial to deliver, bearing in mind the range of backgrounds and skills present in the room. The fact that Dr Chown burned through much of what O'Reilly takes 400 pages to cover, in only three hours, without making our heads spin (too much!), made me feel my employer got value for money on this occasion. Dr Chown began with a whistle-stop tour of the IPv6 packet structure, highlighting differences we should be aware of as well as how the new traffic is to be handled by IPv6-aware routers. Some may have felt this went by a little too quickly, but the tutorial was primarily on IPv6 Deployment so I found the level of coverage was appropriate.
The tutorial's core covered the interaction between IPv6 and network services such as DNS, IPSec, and others. Two big features of the new protocol are its support for host auto-configuration (a leap ahead of DHCP) and mobility (a.k.a. roaming). These I think highlight best why IPv6 is so much more than just an address space extension, and the many designers have done a good job in providing an extensible and versatile protocol. Finally Dr Chown moved onto more real-world scenarios; things we were likely to deal with ourselves if deploying IPv6. This included, of course, OS and router platform support as well as integration and transition issues (say, tunnelling) for the network as a whole. Thankfully this was a 'warts and all' talk so my notes include some of the pitfalls and missing pieces of the IPv6 puzzle.
After something like a decade in development, it is clearly now possible to use IPv6 (almost) without limitation. Due to the site's network restrictions Dr Chown wasn't able to give us a live demonstration although he did invite us to inspect his laptop's configuration after the session was over. It's not something I particularly missed, though, as the notes were thorough and it's far more thrilling to enter the ping command yourself and see it work! As a systems administrator, the tutorial really made me think about my own organisation's deployment of IPv6, and provided inspiration but also a warning: that this is sure to take time; preparation and experience are equally important, but the site and world-wide benefits will be well worth the pain.Oliver works for the Network Software Group at Oxford University Computing Services.
18th and 19th February
This year the winter conference was held in the Institute of Physics in Central London; commencing bright and early on the Tuesday morning with an IPv6 tutorial given by Tim Chown which was enthusiastically taken on board by those who had managed to get tutorial places -- or taken a chance of coming along anyway.[See the description of the tutorial by Oliver Gorwits.]
The conference proper opened after lunch with a pair of talks on the subject of OS X, the operating system from Apple which has gained increasing prominence in the UNIX world since its release nearly two years ago. Lindsay Marshall gave an enthusiastic commentary on OS X from the user's perspective before the audience got its teeth into a technical run-down of the technology behind the operating system given by Paul Burford from Apple, which although buzzword-compliant did provide an interesting insight into the design of OS X. Questions included some heckling on Apple's policy towards Open Source software and open protocols/APIs.
Following on from this was Stuart McRobert from Imperial College, who gave an overview talk on very high speed host networking, focussed around "TOE"s -- TCP/IP Offload Engines. These specialized chips have the capability to significantly improve the performance of file servers by putting the gruntwork of TCP/IP processing into hardware, with some cards having Linux binary drivers now, and more on their way. The aim with this approach is to be prepared for 10 Gbit/s file serving.
After a short break for tea and coffee, where we had another opportunity to browse O'Reilly's books and Dave Green (of NTK fame)'s tee-shirts, the conference resumed with a talk on a topic sure to generate controversy amongst a bunch of seasoned UNIX hackers: The Grid. Steven Newhouse, from the London E-Science centre, gave an overview of the current state of Grid technology, and sparked an impromptu debate on the real value of Grid systems.
The last talk of the day, on MTA performance comparison, gave no clearcut conclusions but brought home the point that software benchmarking is hard! Brad Knowles took an Apple Powerbook and a Compaq Armada running OS X and FreeBSD respectively, and showed us the sort of differences different filesystems, memory and MTA options had on the overall performance of a busy MTA, using Postfix and Sendmail as examples.
After the day's proceedings closed, some of us went straight to a pub, and then onto an casual but enjoyable meal at the Spaghetti House on Googe Street where much lively conversation ensued.
Wednesday opened with David Holdsworth from Leeds University, presenting a paper on Leeds' homegrown authentication system based around X.509 certificates. As well as a brief introduction into the nature of X.509 certificates, David went through the practical aspects of setting up a large scale system with automatic distribution of certificates, in a talk entitled "Authentication beyond ATHENS". The ultimate aim of the system would be to inter-operate between Universities, sharing authentication information.
Next up was a talk scheduled as "UK Law and the System Administrator" but actually titled "Self governance and the Internet's naming and address allocation systems" which was a discussion of ICANN, its mode of operation and the impending changes. Despite its deeply political nature, Richard Francis, from Internet Governance Consultants, gave a very relevant talk that was appreciated by the audience.
The next talk was something of a novelty. Andrew Nicolson from Bristol spoke about the "Open Development for a Fuzzy Client" which discussed the use of Open Source development in order to create a system for LETs groups in the UK. It was an entertaining talk, describing a process a far cry from any commercial or even academic development, without any such demands as speed, security, or consistency, and with no budget or deadlines.
The penultimate talk of the conference was an update, for those of us who were at the UKUUG Linux 2002 conference, of Josh Howlett's nomadic network service at Bristol University. He gave again an overview of the setup of this interesting network access setup, and was able to give some information on real-world experience now that the system was in regular use.
The final talk was given by Eddie Bleasdale, from Netproject Ltd, on "Extraordinary Popular Delusions and the Madness of Crowds". He described some of the problems with the IT industry today as he sees them, and some possible solutions. He spoke about Netproject's project to put Linux on the desktops of the West Yorkshire Police, and enthusiastically called upon the audience to help resist the commercial lock-down resulting from high licence fees and proprietary software, and soon, the so-called Trusted Computing Platform Alliance.
Gregory Neil Shapiro
The BSDCon 2003 Program Committee invites you to contribute original and innovative papers on topics related to BSD-derived systems and the Open Source world. Topics of interest include but are not limited to:Embedded BSD application development and deployment Real world experiences using BSD systems Using BSD in a mixed OS environment Comparison with non-BSD operating systems; technical, practical, licensing (GPL vs. BSD) Tracking open source development on non-BSD systems BSD on the desktop I/O subsystem and device driver development SMP and kernel threads Kernel enhancements Internet and networking services Security Performance analysis and tuning System administration Future of BSD
For more information about the BSDCon 2003 Call for Papers, visit: http://www.usenix.org/events/bsdcon03/cfp/
Submissions in the form of extended abstracts are due by April 1, 2003. Be sure to review the extended abstract expectations before submitting. Selection will be based on the quality of the written submission and whether the work is of interest to the community. For detailed author guidelines, including sample extended abstracts and final papers visit: http://www.usenix.org/events/bsdcon03/cfp/guidelines.html
We look forward to receiving your submissions!
Sincerely,Gregory Neil Shapiro BSDCon 2003 Program Chair
Bradley M. Kuhn
Professor Lawrence Lessig Awarded the 2002 FSF Award for the Advancement of Free Software
Brussels, Belgium - Saturday, February 8, 2003 - The Free Software Foundation (FSF) bestowed today its fifth annual FSF Award for the Advancement of Free Software. FSF President and founder, Richard Stallman, presented the award to Professor Lawrence Lessig for promoting understanding of the political dimension of free software, including the idea that "code is law". Lessig has also promoted ideas similar to free software in other related fields.
The award ceremony was hosted at the Free and Open Source Software Developers' Meeting (FOSDEM) in collaboration with the Free Software Foundation Europe.
A committee of Free Software leaders selected the winner and two other finalists from the nominations received by the public among the thousands of mostly volunteer programmers worldwide who dedicate their time to advancing Free Software. The selection committee included: Enrique A. Chaparro, Frederic Couchet, Hong Feng, Miguel de Icaza, Raju Mathur, Frederick Noronha, Jonas Oberg, Eric Raymond, Guido van Rossum, Peter Salus, Suresh Ramasubramanian, and Larry Wall.
Lessig was chosen from three finalists for the award. The other finalists were Bruno Haible and Theo de Raadt.
This was the fifth award of this kind. The prior winners were Larry Wall, Miguel de Icaza, Brian Paul, and Guido van Rossum.
About Free Software Foundation Europe:
The Free Software Foundation Europe (FSF Europe) is a charitable non-governmental organization dedicated to all aspects of Free Software in Europe. Access to software determines who may participate in a digital society. Therefore the freedoms to use, copy, modify and redistribute software -- as described in the Free Software definition - allow equal participation in the information age. Creating awareness for these issues, securing Free Software politically and legally, and giving people freedom by supporting development of Free Software are central issues of the FSF Europe, which was founded in 2001 as the European sister organization of the Free Software Foundation.
More information about the FSF Europe can be found at http://www.fsfeurope.org/
About Free Software Foundation:
The Free Software Foundation, founded in 1985, is dedicated to promoting computer users' right to use, study, copy, modify, and redistribute computer programs. The FSF promotes the development and use of Free (as in freedom) Software - particularly the GNU operating system and its GNU/Linux variants - and Free Documentation for Free Software. The FSF also helps to spread awareness of the ethical and political issues of freedom in the use of software. Their web site, located at http://www.fsf.org, is an important source of information about GNU/Linux. They are headquartered in Boston, MA, USA.
Lindsay Marshall, evangelising Mac OS X at the recent UKUUG Winter Conference, remarked "It's great - it even runs Microsoft Word!". He then launched into an anecdote about one of his students who had submitted some coursework - in the form of a Microsoft Word document that consistently crashed the Microsoft Word application running on Lindsay's PC. It wasn't clear whether or not the version of Word running on his Mac fared any better, but that is not the point. What on earth is he doing encouraging students to submit coursework in this format? Why aren't those of us in educational establishments taking the opportunity to teach our students a better way of doing things?
I have long railed against the use of Word documents as a format for information interchange, and Lindsay's example proves the point. In his article "We Can Put an End to Word Attachments", Richard Stallman writes:Distributing documents in Word format is bad for you and for others. You can't be sure what they will look like if someone views them with a different version of Word; they may not work at all. Receiving Word attachments is bad for you because they can carry viruses: see http://www.symantec.com/avcenter/venc/data/acro.html Sending Word attachments is bad for you, because a Word document normally includes hidden information about the author, enabling those in the know to pry into the author's activities (maybe yours). Text that you think you deleted may still be embarrassingly present. See http://www.microsystems.com/Shares_Well.htm for more info. But above all, sending people Word documents puts pressure on them to use Microsoft software and helps to deny them any other choice. In effect, you become a buttress of the Microsoft monopoly. This pressure is a major obstacle to the broader adoption of free software. Would you please reconsider the use of Word format for communication with other people?
Although many of the Free and Open Source word processors and office packages lack the wealth of features of the Microsoft Office suite, the main contenders have a very different approach to document formats. OpenOffice.org, for example, has a mission "...to create an open and ubiquitous XML-based file format for office documents and to provide an open reference implementation for this format".
In November of last year, the Organisation for the Advancement of Structured Information Standards (OASIS) announced the creation of a new technical committee to create an open, XML-based file format specification for office applications. The proposed file format is to be suitable for office documents containing text, spreadsheets, charts, and graphical documents. Perhaps most important in terms of interoperability, it will be friendly to transformations using XSLT or similar XML-based languages or tools.
The charter for the technical committee is available online. Sun Microsystems has already contributed the OpenOffice.org XML Format to the technical committee under Reciprocal Royalty Free terms, and this will be used as the basis for the committee's work. Arbortext, makers of the XML-based publishing software, Epic, are represented on the OASIS technical committee, as are Corel, makers of the Word Perfect word processor.
Cover Pages, hosted by OASIS, provides an excellent collection of references on XML file formats used in office applications. Free and Open Source office suites such as GNOME Office and KOffice already use XML for their document formats. Apple use XML widely throughout their Mac OS X operating system, and recently published the schema describing the XML file format used by their presentation software, Keynote.
Although members of OASIS and aware of the technical committee, Microsoft will not initially be taking part. But have they finally seen the light? In a press release last November, Microsoft announced that the next version of Microsoft Office "...broadly supports XML (eXtensible Markup Language) and enables the exchange of data across diverse systems, platforms, and applications". The new version, codenamed "office 11" is due to ship in mid-2003. If they follow Apple's example and release their XML schema, perhaps later this year I will be able to stop my railing.
Published by O'Reilly and Associates
reviewed by Bob Vickers
This book has the virtue of a clearly defined target audience, plus a pleasing slimness and a lack of those irritating screenshots which clutter up so many books nowadays. But I was disappointed; I did not sense the authors really understood what a Unix geek would wish to see in this book.
My background includes over a decade of Unix experience but as a Mac user I am a complete novice. I volunteered to review the book because someone at work has just acquired a Mac and we were having great difficulty persuading it to import NFS file systems. Eventually we managed to find the solution in the help system, but I thought maybe the book would make things easier next time that we hit a problem.
I would have really appreciated an introductory chapter giving an overview of the Mac OS architecture and the concepts that might be unfamiliar to me. For example: what is a resource fork? These beasts are mentioned 2 or 3 times in the book without any explanation, but I suspect you could go badly astray if you tried to back up data without knowing what they were.
Although the book leaves a lot unexplained it does devote a lot of space to quite unnecessary material. For example, the first chapter includes 10 pages on tcsh; surely it can be taken for granted that a Unix geek will already be familiar with one of the 4 shells available on the system? In the same chapter 10 pages are devoted to an incomplete list of Mac OS commands; this list would be both shorter and more useful if it omitted all the standard Unix commands; we really don't have to be told that there are commands called 'echo', 'man', 'cp', etc.
After this poor start the book improves. There is a chapter on Startup
explaining how to configure which services get started, followed
by a vital one on Directory Services which tells you how the familiar
They are followed by three chapters on building and installing applications. The first two tell you about compiling Mac-specific applications as well as telling you how to port existing Unix applications; the third describes the package management systems used by the Fink and GNU-Darwin projects. I felt there was too much detail here and not enough overview. I do not buy a book to replicate instructions I can easily find on the web or in man pages; I want the authors to give me insight and the benefit of their experience so that I if I need to do something I can go straight to the right tool and not waste time ploughing through documentation irrelevant to my needs. In this instance they could have explained when you would choose to look a package at Fink rather than GNU-Darwin or vice versa.
In part III of the book there are chapters on Building the Darwin Kernel, System Management Tools and the X Window System. This last chapter goes into quite a lot of detail about installing X11, which is the first step a Unix geek would need to take before downloading a batch of favourite applications. There is also information about using ssh to connect between machines...yes, you can login to a Mac remotely just like any other Unix box.
Appendix A gives us a tour of the Mac file system and appendix B is a list of missing man pages. We must hope that the man pages find their way into their rightful place in the distribution.
So in summary: there are some good things in the book but also some irrelevant parts and a lot of omissions.Bob manages the Computer Science computers at Royal Holloway, University of London.
Published by O'Reilly and Associates
reviewed by John Collins
This book sets out to describe HTTP - Hypertext Transfer Protocol, on which so much of the Web revolves. In my view it goes a great deal further than just defining HTTP, which is done in the first section. The book covers just about everything you could possibly think of, and quite a few you would probably take quite a while to think of, to do with the web apart from web page content, HTML and so forth.
It is in six sections of four or five chapters each. In the first section, HTTP itself is described in some detail. We learn about requests, URIs, HTTP Messages, and Connection Management in successive chapters.
The second section describes what are described as "HTTP Architecture". This covers the mechanics of servers, although to be fair, Apache is the only server covered in much detail, various kinds of proxies, caching, gateways, tunnels and relays, robots and how to program for them, before concluding with future directions for HTTP.
Next there is a section on identification authorisation and security, covering HTTPS and digital certificates and explaining how they operate.
The fourth section describes entities, encodings, and internationalisation including details of content negotiation.
The fifth section talks about content publishing and distribution. This covers web hosting with multiple severs, redirection and load balancing and has a whole chapter on logging and log file formats
Finally there is a comprehensive set of appendices listing URI Schemes, HTTP Status Codes, HTTP Headers, MIME Types, Base-64 encoding, Digest Authentication, Language Tags and country codes and MIME Charsets. There are various C and Perl programs provided to illustrate some of these concepts.
Every part bristles with clear diagrams explaining the logic or mechanism for what is being described.
I think this book is an extremely useful, very comprehensive and clearly-written reference to all aspects of the internals of the Web going well beyond just the bare mechanics of HTTP. Even where its huge detail does stop on a topic, there are extensive and useful references for further reading on each topic covered given at the end of nearly every chapter.John Collins -- Xi Software Ltd www.xisl.com
Published by O'Reilly and Associates
reviewed by John Collins
This book introduces the C# programming language, the object-oriented language devised by Microsoft for their .NET environment. It does not assume a great deal of knowledge about programming, object-orientation or .NET.
I confess before I start that I do not like C# very much myself (nor do I like .NET or Microsoft but that's another story). I think that it's too like C and C++ whilst holding out a huge number of surprises in various places, such as "char" being 16 bits and what look like ordinary declarations automatically being references in some cases. It doesn't have multiple inheritance or pointers to members (in fact it scarcely has pointers to anything) but it does have a whole new idea called interfaces. I get the feeling that it was designed by someone who has strong view on some aspects of C and C++, like forbidding "falling through" in switches and banning the passing of parameters by references which have not been initialised (though there is a new keyword "out" to deal with that). C++ users will probably miss templates, although some of the commonly-used containers, notably strings, are built in to C#.
The book tackles the language in the time-honoured way in which programming languages are presented, talking about simple variables and constants and control structures before moving on to classes, inheritance, operators, structs (which are much more different from classes in C# than in C++), then onto arrays, strings and regular expressions before concluding with exceptions.
The book concludes by suggesting further reading for serious use of the language.
I think that the book introduces the language well although I am sure that the reader will have to go for one of the fuller manuals to use the language seriously. One or two features did not seem to me to be fully explained, for example the use of "new" and "override" with and without "virtual" for derived class methods wasn't completely clear to me, and it would be nice to have had an outline of one or two of the operators, such as those for multithreading, in the end.
A chapter or set of appendices introducing the language to C and C++ and Java programmers and highlighting the differences would not have gone amiss in my opinion, but apart from that I am sure this book will get those of us who have to use C# off the ground at reasonable speed.John Collins -- Xi Software Ltd www.xisl.com
Piracy Is Progressive Taxation, and Other Thoughts on the Evolution of Online Distribution
Tim O'Reilly draws upon his own publishing experience to comment on the current controversy over digital file sharing that's being waged by the movie and music industries. http://www.openp2p.com/pub/a/p2p/2002/12/11/piracy.html
How Perl Powers Christmas
We've had some fantastic Perl Success Stories in the past, but this one tops them all: How Perl is used in the distribution of millions of Christmas presents every year. http://www.perl.com/pub/a/2002/12/18/hohoho.html
An Interview with Jim Kent
In this O'Reilly interview, Jim Kent talks about the future of genomics, modeling biological processes, and open source bioinformatics. Jim will accept the annual Bioinformatics.Org Award in a plenary session at O'Reilly's upcoming Bioinformatics Technology Conference. http://www.oreillynet.com/pub/a/network/2002/12/10/kent.html
Stephen Wolfram on Bioinformatics
Tim O'Reilly interviewed Stephen Wolfram, creator of the Mathematica program and author of "A New Kind of Science," about his research, his new book, and its connection to bioinformatics. Wolfram will give a keynote at O'Reilly's 2003 Bioinformatics Technology Conference. http://www.oreillynet.com/pub/a/oreilly/bio/news/wolfram_0120.html
O'Reilly Bioinformatics Technology Conference
This conference takes place at the Westin Horton Plaza in San Diego on February 3-6, 2003. http://conferences.oreilly.com/biocon/
Remote Application Development with Mozilla
This article will give you a firm grasp of the potential uses for and limitations of remote XUL. Learn how to start using remote XUL to enhance your web sites and applications, from Myk Melez and O'Reilly author Brian King, coauthor of "Creating Applications with Mozilla." http://www.oreillynet.com/pub/a/mozilla/2002/12/17/app_dev.html
What is RSS?
In Mark Pilgrim's inaugural Dive Into XML column, he reviews the history and technical details of the varieties of RSS on the Web. He also describes a method for parsing most active RSS feeds. http://www.xml.com/pub/a/2002/12/18/dive-into-xml.html
Screen-scraping with WWW::Mechanize
Screen-scraping is the job of programmatically navigating through a usually visual task--like a web site--and then dealing with the result; and WWW::Mechanize is the best screen scraper out there for Perl! Chris Ball puts the two things together, to ensure that he never misses his favourite TV shows again. http://www.perl.com/pub/a/2003/01/22/mechanize.html
Opening Microsoft File Formats to Java
Microsoft's file formats were once black arts to developers on non-sanctioned platforms. Thanks to Jakarta POI, however, it's possible to read and write them from Java. Apache's Andrew C. Oliver and Avik Sengupta explain the basics of the project in the first of a series of three articles. http://www.onjava.com/pub/a/onjava/2003/01/22/poi.html
2003 O'Reilly Emerging Technology Conference
Join the leaders from the many worlds of pervasive computing to explore and invent the new internet infrastructure and culture. Speakers include: Howard Rheingold, Alan Kay, Clay Shirky, Eric Bonabeau, Tim O'Reilly, David S. Isenberg, and many others. http://conferences.oreilly.com/etcon/
Council Chairman; Events; Newsletter
07973 231 870
PO Box 37
01763 273 475
01763 273 255