ACCU 2012 was another great conference in the series. Lots of good sessions, occasional not so good one, lots of time for socializing (we'll ignore the euphemism networking, and call it what it is) and a lot learned.

There was more of an explicit emphasis on C++ this year, but this was due to the submissions offered to the conference committee rather than any direction on their part. Almost certainly the history of ACCU being a C and C++ focused organization, and the publishing of the C++11 standard, had much to do with this.

A great deal was made of the "death of C", but everyone was explicit that this is C being a dead language in the way that COBOL is a dead language: you don't use it but there is an awful lot of it about.

My contributions this year was, as ever, to do with concurrency and parallelism. I did a session "Go, D, C++ and the Multicore Revolution" which was asking whether the changes to C++ due to the C++11 standard was just too little too late. C++ now has a memory model, asynchronous function call and futures but this is still an abstraction level below that that applications programmers deserve. Go is trying to replace C and incorporates Communicating Sequential Processes (CSP) as its model of concurrency and parallelism. Go is considerably simpler than C++ and yet achieves most of the tools for abstraction and programming that most programmers need most of the time. D is trying to replace C++ and has far better facilities for generic programming and uses actors as its model of concurrency and parallelism. It could be said that D is already that which C++ is evolving to become.

I introduced the C++ actor system that Anthony Williams' Just::Thread Pro will have – I have a preview version – to show that C++ could catch up with Go and D if there was a recognition in the C++ community and standards committee that C++ needs to make use of the goodwill generated by the C++11 standard. C++ has the potential to be an interesting programming language. In the mean time Go and D are already very interesting languages, they make native code working feasible. Unlike C, but then that is now a dead language.

The slides for the session are here but they probably don't mean that much out of context. The codes shown are in the Bazaar branch to be found here. We didn't do any of the Sleeping Barber material, interactions during the session meant we only looked at the Π by Quadrature material.

I did a second session in collaboration with Schalk Cronjé. The core technical content was to show actors, dataflow, and CSP in action. The subplot was to some some comedy. We had sketched out the plan a few weeks before, but had to leave final planning to the very last moment. So we had a structure and a rough plan, but hadn't done a full script; we had to do a lot of ad libbing. I am not sure it worked completely, but the audience got involved and didn't leave. The problem we tackled was generating signatures for files in a directory hierarchy. This is a real problem from Schalk's work, not a toy problem. If people were enthused that Groovy already has, via GPars, everything that all the other languages desire in the way of actors, dataflow and CSP support, and that the future of applications programming is to use these models, then the session was a win.

The slides for this session are here. The code used is in a Git repository: the bare repository for cloning is http://www.russel.org.uk/Git/ParalysedByParallelism.git, if you just want to browse without cloning then you want to go here.

[18:45] ( Comments)

In his keynote at ACCU 2012 this morning, Uncle Bob made a couple of arguably ill-judged asides. They weren't grossly sexist, but they clearly came from a certain view of sex/gender stereotypes. A woman in the audience, quite reasonably, pulled him up on them at the time, and again after the keynote. Which was good. Uncle Bob, not actually being a sexist, nigh on immediately wrote and published a statement. For me this showed what an excellent person Uncle Bob really is. Serious respect – as the saying goes.

Being from the same era, albeit from the UK rather than USA, and a regular public speaker, I fully appreciate his difficulty. All through growing up we had stereotype sex/gender role models rammed down our throats. In the heat (and stress) of presenting, it is way too easy to let those early year attitudes creep through despite 40+ years of trying to ameliorate them and replace them with sensible attitudes of real equality. Let's forget political correctness here, that is almost as bad as the 1950s and 1960s sexism, let's talk about real equality.

Sadly UK and USA society remains fundamentally sexist: parents are the problem here. Boys are dressed in blue, given monster truck and guns as toys, and expected to be brilliant sportsmen or massive intellectuals. Girls are dressed in pink, and expected to learn to be good mothers, even being given dolls from Day 0. Dress codes and attitudes are fundamentally sexist: girls wear skirts, shirts or trousers, and from puberty onwards, are often expected to wear as little as possible in public showing as much skin as possible. Boys were shorts and then trousers and are expected to always be covered. Think of the rules of evening dress: women are expected to be in fabulous gowns showing as much skin, especially around the breasts as possible, men wear dinner suits. Only Scotsmen have permission to wear skirts and only then if they are 8-yard knife-pleat kilts with the full Victorian extras – but not a dirk, that would be illegal.

OK so there is a massive simplification here but...the core point remains valid. Parents enforce sex/gender stereotypes from Day 0 of a child's life and that is at the core of our society. Is it any wonder those children of more enlightened parents who want to break out of the standard models have a problem? In the case of ACCU 2012 we are talking "Girl Geeks". Sadly there remain far, far too few of them.

Uncle Bob's slip was not that serious but has become a bit of a storm in a teacup. However the underlying problem is very real and is a serious problem: our society has no real equality no matter what the law says. Until we find a way of men and women having equal opportunity generally even though women have to bear the children, sexism will remain systemic. Biology may assign sex/gender roles, but humans are surely intelligent and enlightened enough to be able to create equality despite this. Currently it seems not.

Being an individual male from the same era as Uncle Bob, I feel almost completely unable to do anything that will actually change things for the better. This is really depressing.

[02:28] ( Comments)

Vodafone did me a part-way reasonable deal on a new contract involving me taking possession of a Galaxy Nexus to replace the N900 I have had for two years. Of course when I got home and found it was a 16GB machine instead of a 32GB machine I was not entirely happy; the Galaxy Nexus has no memory expansion capability. Still it is Android, Maemo being well dead, ditto MeeGo, and version 4 at that. (No I am not going to use inane terms like Ice Cream Sandwich, or even ICS.)

I have been reasonably pleased with the phone over all, and the Android Market, sorry Google Play, downloading has worked – even to the extent of updating all the out of date operating systems and applications. I am even looking forward to paying for some apps, not to get rid of adverts but to get some of the extra functionality. I have actually been quite surprised at how immune I have been to the advertising on free apps. If advertising can fund some apps, whilst having no effect on me, then long live free apps.

But there is a downside. Actually there are two:

• Vodafone 3G coverage is still very poor and almost, but not quite, totally unrelated to the coverage map that they publish. For a world rapidly going 4G, the UK is woefully under-resource with 3G and 3.5G. How can the UK compete in the information era with such appalling mobile network systems? Thankfully the 3 UK coverage is not bad. Pricing for data is still a complete rip-off: the whole 3Gb for 1 month business model should surely come under scrutiny. Either it should be unlimited for a month or 3Gb for an unlimited period. I dread to think how little I'll be able to do on 500Mb per month on the Vodafone contract if I can ever get 3G on Vodafone.
• Bluetooth tethering does not work. USB tethering works fine. Wifi tethering works fine. Bluetooth tethering says it has succeeded to the PC, but reports failing on the phone.

So I went to my local Vodafone store where I got the new contract a few weeks back. The very nice chap there listened to my question and with good grace said he didn't know the answer. This I though was good customer service because of the way he dealt with it. He suggested I should contact Technical Support via telephone and set me up using their in shop free phone. Sadly this was the beginning of the problems.

Vodafone answers, and requires you to enter your phone number or account number and PIN number before you can get anywhere. OK, annoying but... You then have to go through four, yes four, layers of listening to a sequence of options before you get to talk to a human being. Well I say human being, but I am not entirely sure it wasn't a trained monkey.

First question: "Please can I have your phone number." For ##### sake Vodafone get your systems working. I already typed in the number to get through, you should present this useful information to the call centre monkeys so that they can give at least a simulacrum of good customer service. So having ranted at the monkey about this, I gave the seemingly duplicate information that he clearly was not being given (not actually his fault though). I then explained the situation and his response was "It's a Samsung problem." I tried to explain that I didn't think it was, but he insisted it was. Basically he was reading the one-line answer printed on the sheet in front of him: "This is not a Vodafone problem, please contact phone vendor." I gave up, asked him to text me the number, which to be fair he did within about 30s. Good monkey. Bad Vodafone.

So, Samsung, or anyone, why is this your fault and not Vodafone's. I switch on Bluetooth on phone and laptop. I switch on Bluetooth tethering on the phone. I select the "connect via Bluetooth" networking option on the laptop. The phone reports tethering is activated, flashes up a message too fast to read and then reports "not tethered" – I have absolutely no idea how to find out what the flashed message was, but perhaps it is crucially important? So the phone believes Bluetooth tethering is active but not happening. Meanwhile on the laptop it is entirely happy that a suitable connection is in place. Even DNS lookup appears to be working. But no packets escape the NAT network. This would indicate that Vodafone is black holing the connection, or am I missing something?

|> ifconfig
bnep0     Link encap:Ethernet  HWaddr 70:f3:95:4e:b3:da  
          inet addr:192.168.44.79  Bcast:192.168.44.255  Mask:255.255.255.0
          inet6 addr: fe80::72f3:95ff:fe4e:b3da/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:24 errors:0 dropped:0 overruns:0 frame:0
          TX packets:38 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:2859 (2.7 KiB)  TX bytes:2702 (2.6 KiB)

eth0      Link encap:Ethernet  HWaddr f0:de:f1:2f:cb:b3  
          UP BROADCAST MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:0 (0.0 B)  TX bytes:0 (0.0 B)
          Interrupt:20 Memory:f2500000-f2520000 

lo        Link encap:Local Loopback  
          inet addr:127.0.0.1  Mask:255.0.0.0
          inet6 addr: ::1/128 Scope:Host
          UP LOOPBACK RUNNING  MTU:16436  Metric:1
          RX packets:126757 errors:0 dropped:0 overruns:0 frame:0
          TX packets:126757 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0 
          RX bytes:11639041 (11.0 MiB)  TX bytes:11639041 (11.0 MiB)

|> route
Kernel IP routing table
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface
default         192.168.44.1    0.0.0.0         UG    0      0        0 bnep0
link-local      *               255.255.0.0     U     1000   0        0 bnep0
192.168.44.0    *               255.255.255.0   U     0      0        0 bnep0

|> dig dimen.winder.org.uk

; <<>> DiG 9.8.1-P1 <<>> dimen.winder.org.uk
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 58035
;; flags: qr rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 0

;; QUESTION SECTION:
;dimen.winder.org.uk.		IN	A

;; ANSWER SECTION:
dimen.winder.org.uk.	86118	IN	A	87.127.116.10

;; Query time: 29 msec
;; SERVER: 192.168.44.1#53(192.168.44.1)
;; WHEN: Wed Mar 28 18:06:45 2012
;; MSG SIZE  rcvd: 53

|> traceroute dimen.winder.org.uk
traceroute to dimen.winder.org.uk (87.127.116.10), 30 hops max, 60 byte packets
 1  192.168.44.1 (192.168.44.1)  6.028 ms  8.318 ms  11.069 ms
 2  * * *
 3  * * *
 4  * * *
 5  * * *
 6  * * *
 7  * * *
 8  * * *
 9  * * *
10  * * *
11  * * *
12  * * *
13  * * *
14  * * *
15  * * *
16  * * *
17  * * *
18  * * *
19  * * *
20  * * *
21  * * *
22  * * *
23  * * *
24  * * *
25  * * *
26  * * *
27  * * *
28  * * *
29  * * *
30  * * *

So Samsung, Vodafone claim this is your problem, that the Galaxy Nexus phone is borked with respect to this functionality. What the #### is wrong. Of course my suspicion is that Vodafone is detecting Bluetooth tethering instead of USB or WiFi tethering and blocking all packets, they are just not going to admit that this is the case.

Actually what I want Samsung to do is to tell me how Vodafone are doing this so that I can undo it. I want Bluetooth tethering. I used to have Bluetooth tethering on the N900. It is a good thing. I want it to work again.

Footnote: I am under no illusions, no-one other than me is ever going to read this mini-essay/rant, but it has been cathartic writing it, so I will publish it anyway.

[18:50] ( Comments)

On 2012-04-02 and 2012-04-03 the JAX Days are being held in London. On the first of these days, 2012-04-02, I will be presenting a one-day “learn in” on the Groovy programming language.

Groovy is a dynamic language designed to be symbiotic with Java: Groovy and Java work together on the JVM, they don't fight each other. When developing systems on the JVM, Groovy makes a very dynamic partner to Java: you write the bits of your system that are best handled by a static language in Java, and those bits best handled by a dynamic language in Groovy.

This one-day session at JAX Days is designed to be for people who have some knowledge of Java to be able to start learning and become proficient at using Groovy, and indeed mixing Groovy and Java in the same system. The intention is for this to be a fun as well as informative day: I am strong believer that programming should be fun.

[12:23] ( Comments)

At the session "Python is a High Performance Language" last evening, I showed some Python code executed using CPython and PyPy. John Hoyland asked what the performance was with Jython or IronPython. An excellent question about which I had no data. I will not be able to check with IronPython on CLR, but I can certainly check Jython. Using the sequential version of my “π by quadrature” using a for loop with xrange and multiplication, the code for which can be found here, I get the following results:

• CPython:

        ==== Python Sequential For/Xrange Multiply pi = 3.14159265359
        ==== Python Sequential For/Xrange Multiply iteration count = 10000000
        ==== Python Sequential For/Xrange Multiply elapse = 5.83972406387
      

• PyPy:

        ==== Python Sequential For/Xrange Multiply pi = 3.14159265359
        ==== Python Sequential For/Xrange Multiply iteration count = 10000000
        ==== Python Sequential For/Xrange Multiply elapse = 0.224266052246
      

• Jython:

        ==== Python Sequential For/Xrange Multiply pi = 3.14159265359
        ==== Python Sequential For/Xrange Multiply iteration count = 10000000
        ==== Python Sequential For/Xrange Multiply elapse = 6.33100008965
      

So I think it fair to say, without fear of contradiction, that PyPy Rocks.

The PyPy Project needs your help. The project relies on voluteer labour and donations. Currently it is trying to ensure PyPy can work with NumPy and that PyPy conforms to the Python 3 specification. Donate individually on the PyPy home page, get your company to donate. PyPy needs your help.

[19:08] ( Comments)

Yesterday evening (starting 2012-01-31T18:34) I presented a session with the above title at Skills Matter. It was scheduled in their “In the Brain” series, so I assumed licence to be very personal and idiosyncratic. The overall aim of the session was to convince people that although Fortran and C++ are seen as the standard languages of high performance computing (HPC), Python has a place. I opened with promise of a pair of arguments that then came together as a form of “proof of case”.

The first argument was to address “high performance” as a term implying the ability to create correct (and fully tested) solutions to problems quickly and with small amounts of code. The implication being that high performance is about the ability of programmers to use the tools to great effect. I emphasized “ceremony”, in particular “low ceremony” and “high ceremony” as a major factor: “ceremony” here being code that has to be written to satisfy the compiler that contributes nothing to the code that actually executes. Python is very definitely a “low ceremony” language associated with the fact that it is a dynamic language. Traditionally statically typed languages have been “high ceremony” and dynamically typed languages have been “low ceremony”. The question is whether the drive to type inference in all the major statically typed languages (D, C++, Java, Scala, Haskell) reduces the level of ceremony to be equal to that of the dynamically typed languages.

This led to the issue of meta-object protocols (MOPs). All the major dynamically typed languages (Python, Ruby, Groovy, Lisp, Clojure) have run time MOPs in one guise or another, and this gives them great capabilities in creating domain specific languages (DSLs) as internal languages. C++ and D even though they are statically typed languages have MOPs; it is just they are compile time MOPs rather than run time ones. So there can be no dynamism to the DSL, but they are still very capable of creating DSLs – just static ones. Although not stated explicitly the issue of internal DSLs leads directly to the idea of coordination languages, and in the context of HPC to parallelism coordination languages.

I then switched tack to address computational performance as the focus of “high performance” ; arguably the more traditional interpretation of the term. Python performs badly compared to C++ and Fortran, generally about 50 to 200 time slower, at least using the CPython interpreter. PyPy performs somewhat better being 10 to 20 times faster on the small microbenchmarks I showed that CPython. As ever for me this was calculating a quadrature approximation to the value of π. The problem is small, embarrassingly parallel and never ceases to be able to be used for showing issues associated with parallelism in any number of languages. (All the code for this problem is available in a Bazaar branch here: feel free to browse or branch, if you branch please let mw know of any amendments or additions.) Both CPython and PyPy have a global interpreter lock (GIL) which means no parallel execution of threads at all. This can be got round in three fundamental ways:

1. Remove the GIL. This is unlikely to happen in CPython, but the PyPy folk are experimenting with software transactional memory (STM) as a way of being able to remove the GIL.
2. Use the multiprocessing package (or Parallel Python) to make use of multiple Python virtual machines (PVMs) each of which runs a single threaded process.
3. Use native code PVM extensions for the parallel computations.

Using Python for the bulk of the code, and C for the core loop meant that execution was about the same as using C for all the application. Thus Python is a high performance programming language. OK, Python is not going to be used for the computationally intensive part, but it can be the coordination language.

Skills Matter videoed the session, see here.

[18:05] ( Comments)

I forgot to upload the slides from my Groovy and Grails eXchange 2011 session and point at them. They are now here. Remember though not many of them were used. All used code in this Bazaar branch

[11:12] ( Comments)

Groovy and Grails eXchange 2011 is as good, if not better, than previous years. The evolution of Groovy towards 2.0.0, and the imminent release of Grails 2.0.0 has resulted in an increased energy in the Groovy and Grails community. I suspect this is founded on the more positive energy in the Java community since Java 7 came out and the timetable for future Java releases published.

GPars is an integral part of the Groovy and Grails, and increasingly Java, communities, as it allows folk to harness the increasing parallelism of computers using Groovy and Java. I ran a session “Just keep passing the messages” yesterday 2011-12-08T16:30. The slides were a minor variant on the ones I used for JAX London, but the use of them was very different and in fact I only used about 25% of them.

Most of the audience were not familiar with actors, dataflows, Communicating Sequential Processes (CSP), and data parallelism. Everyone did appreciate though that the question “Is Java an object oriented language?” was actually a deep and non-trivial one. So I skipped a lot of the ranting and sermonizing and skipped straight to giving a summary view of the various models and showed some GPars code – though I did retain my ranting against shared-memory multi-threading as an application level technique. I even managed to run some it to show how GPars really does lead to parallelism, and how Groovy is slow, but this doesn't matter as long as the computationally intensive tight loop is in Java. Groovy/GPars for coordination of concurrency, Java/GPars for grunt computation. Yes GPars does have a Java API!

The example I used my standard example of using quadrature to approximate the value of π. Even though I have used it for a while people still report that it remains a good example since it is small relatively easy to comprehend and allows examples to be comprehensible. So it works! All the code is available in a Bazaar branch at http://www.russel.org.uk/Bazaar/Pi_Quadrature. Feel free to branch, play, and report data back.

Skills Matter videoed the session, you can find it here.

[10:18] ( Comments)

End of this week (2011-12-08 and 2011-12-09) is the Groovy & Grails eXchange 2011 being held at Skills Matter. Definitely a case of “Be there, or be square” – or perhaps that should be “if you are not there you cannot be groovy”, or some such weakly humorous play on 1960s lingo. Kick off is 08:30-ish, so that'll be 08:30:01 then.

Sadly the conference is two track and I want to go to all of the sessions. I therefore need a volunteer to be my proxy at the sessions I am not actually at. This is especially important at 2011-12-08T16:15 when I present my session “Just Keep Passing the Messages ”.

[11:55] ( Comments)

120 people turned up yesterday (yes Saturday) for the London Java Community (LJC) annual unconference. I had intended to be a non-presenting delegate. However when it came time to propose sessions, I could not resist proposing a lightning talk on GPars and a group discussion on "Parallelism on the JVM".

I quickly put some slides together for the lightning talk, they are to be found here. I thought the talk went well, certainly people came up during the lunch break to discuss various things about parallelism ion the JVM generally and GPars in particular.

In the end my proposal for a discussion meeting and a proposal for a Q&A panel session got merged, so I found myself on the panel. The natural direction given the members of the panel was parallelism so that seemed fine. I thought the discussion worked well.

The conference was excellent. All the session I went to were great. Well done the team for putting on such a good event.

[12:06] ( Comments)

Tomorrow being the 90th anniversary of the British Legion Poppy Appeal, the UK having troops in Afghanistan, Afghanistan growing most of the world's poppies for the drug trade, one of which is opium, and having been exposed to the appalling sight of parliament in action yesterday, it struck me that the political classes really are putting one over on the electorate. Probably everywhere but certainly in the UK.

The pathetic posturing, ranting, and petty point scoring on both sides of this verbal sparring, were awful enough. I am guessing this is what 19th century school and university debating was like – preparing the boys of the rich and famous to take on the role of leadership and governance. What was really saddening though was that the two principles in yesterday's events were women. Instead of evolving politics so as to be more sane, these women have descended into the worst pit of puerile behaviour aimed at creating good television ratings that passes itself of as politics in this country.

So rather than bringing a new form of politics, the push to have more women in parliament (in and of itself a good thing, there should be 50% female and 50% male MPs) has reduced these female MPs to the same stereotype of a 19th century male MP as the men continue to adopt. An opportunity seriously missed.

I have no idea what can be tried next to bring a more constructive politics, our combative system is about destruction not about construction – going to war appears to be more important to our politicians than having an efficacious health system for all the people.

It has been said that democracy is a bad system, but it is better than all the others that have been tried. If what we in the UK have today is democracy, please offer me something new that hasn't been tried yet.

[18:40] ( Comments)

Lots of people appear to be getting very excited about tomorrow being 11/11/11 as being something special. Well as 11/11/11, rather than the more correct 2011-11-11, it is a special pattern. But what is really special is that this is the 90th anniversary of Remembrance Day. 11th November is Armistice Day every year, and every year we should wear a poppy as symbol of remembering the carnage and horror of war, and of the mutilation and death suffered by innocent people and combatants. Even FIFA have relented in recognition of what it means to remember on Remembrance Day. Do it. You know when – 2011-11-11T11:00+00:00.

[18:22] ( Comments)

I have just signed up to do an evening session on 2012-01-31 at Skills Matter entitled “Python is a High Performance Programming Language”. See here for the Skills Matter advert and sign up page. Don't let the Skills Matter short bio let you believe I only work with Groovy and GPars, Python rocks.

Also don't be concerned that Skills Matter categorize Python sometimes under .NET and sometimes with Scala and F#. This is because they are just putting together a Python training activity and so there isn't yet a proper silo for it in their portfolio. If people take up their newly established Python training activity this will change. cf. this advert

[18:56] ( Comments)

I only went to 15 sessions of the total of 40, but given that Day 1 had five tracks and Day 2 had four tracks, I think it was good going to be in a session at every slot. Sadly though there were at least 10 other sessions I would like to have been at.

The four keynotes were reasonable, with a nice mix of motivational aspects and informative aspects. Nothing really stood out though, except perhaps that in Simon Ritter's summary of the road map for Java, JavaFX on Linux is not due till half way through 2013, which shows how much the JavaFX team and Oracle care about being truly multi-platform.

Given my main interest is concurrency and parallelism, I took in all the session related to that topic, along with the Java 8, 9, 10, . . . summaries. I didn't hit the Android track.

Of course I have to mention my session: “Just Keep Passing the Messages”, slides here. This was really a rant against unthinking use of shared-memory multi-threading as an applications programming technique, and a plea for applications development folk to learn about and use higher level models such as actors, dataflow, communicating sequential processes (CSP). I plugged GPars as a framework that Java and Groovy programmers can use now to do all this, so they don't have to wait.

All the other 10 sessions were good, worth attending, and I enjoyed them. The three “stand out” sessions for me though (other than mine :-) were “An Introduction to Neo4j (and Doctor Who)” by Ian Robinson, “JavaFX 2.0: Great Interface in Java” by Simon Ritter, and “Understanding the Disruptor, a Beginner's Guide to Hardcore Concurrency” by Trisha Gee and Mike Barker.

Ian Robinson managed to educate everyone attending in the subtleties of studying Doctor Who, including retcons, whilst really showing that Neo4j is really a unique form of database management system. He focused mainly on daleks, which was fine as they are clearly a common enemy! Ian mentioned that the daleks first appeared in 1963-4, which came as a bit of a shock, as I remember watching those episodes at a friend's house the first time they were shown.

Simon Ritter's presentation of JavaFX 2.0 showed for me that it was the right decision by Oracle to drop JavaFX as a purely script based system and to transform it into a Java API. This is not because I want to code UIs in Java, but because it makes it an open system that works well as an integral part of the JRE. Simon's Java examples were fun, and worked well, but showed that Java is not the right language to actually write UIs with. I will be using Groovy most likely. The point is that there is not one fixed language to write JavaFX 2.0, and soon JavaFX 3.0, UIs with. JavaFX 2.0 seems a splendid successor to AWT and Swing. But of course AWT/Swing UIs can incorporate JavaFX bits, so people can use evolutionary rather than revolutionary development. Personally I will just dive straight into JavaFX 2.0, but only when the Linux version is released so that UIs can be platform independent: Windows, Mac OS X, Solaris, and Linux.

Trisha Gee and Mike Barker's session stood out since above all it showed how to do some real science when engineering a solution to a problem. Their problem is to process vast numbers of messages with various transformations of data needed. They showed, via experimental evidence, that standard approaches to concurrency and parallelism we insufficient to the task. Indeed employing parallelism made things slower for their problem compared to a simple single threaded solution due to all the overheads associated with the concurrency and parallelism tools. Here we are not just talking of Java and JVM overheads we are talking of JIT generated assembly language and processor cache line management. I am used to C++ folk worrying about cache lines, it was inspiring to have Java folk seriously worrying about these things entirely appropriately. They then showed how they had analysed various data structures, reasoning with the Java memory model, and experiments probing the JIT code generation, to write a lock free ring buffer that is the data structure at the heart of the Disruptor. Opinion not necessary, the experimental data shows they have been massively successful.

I will be submitting a session proposal to JAXLondon 2012.

[12:28] ( Comments)

Having just been to Ian Robinson's talk on Neo4j, where he presented significant quantities of Dr Who data, I am sensitized to time travel. I am therefore loading the slides for my talk before I give it rather than after. If I change the slides between now and then, there will be a retcon.

"Just Keep Passing the Messages" 16:20--17:10

[12:30] ( Comments)

OK so here I am sitting in a restaurant in Salford Quays, aka MediaCityUK, aka BBC North's now home, on my tod, and bizarrely finding myself reflecting on personal history. In particular, at the moment, wondering how the heck I am in the position of presenting Python training courses (*). For some obscure reason I am transported back to a lecture room at UCL, sometime around 1985, where I was presenting one of a 30 lecture course on Software Engineering. I recollect in particular two faces, one I shall label Winnie Pun, the other I shall label Martin Fowler. The question in my mind is: is Martin Fowler Martin Fowler? Winnie Pun I know to be Winnie Pun as she subsequently undertook, successfully, a PhD with me as supervisor. But is Martin Fowler Martin Fowler?

I remember a number of conversations in the corridor between the aforementioned lecture room and John Washbrook's office involving John Washbrook, myself and the individual I have labelled Martin Fowler – who may or may not be Martin Fowler.

I then skip across the Paradigm Wars (object-oriented vs. functional), various OOPSLA conferences involving drinking sessions with Bruce Anderson trying to kick start the pattern movement (cf. OOPSLA 1990 and 1991), and workshops with Kent Beck on "what is an object?" (cf. OOPSLA 1990), but I come back to: is Martin Fowler Martin Fowler.

What is really at the heart of this angst is: have I done enough real world work with Java, Python, Groovy to be able to teach Java, Python, Groovy? In 1985 I had had so little experience of actually doing big systems development, that my lectures were basically entirely theoretical based on published work by people who had even less practical experience than I had – is this a synonym for bullshit. I had at least been a successful UNIX systems programmer before becoming an academic. I continue to wonder if any of the people writing about software engineering in the 1980s had any actual experience of doing it at all: far too many of the 1980s software engineering community clearly had absolutely zero real experience of developing systems. How can anyone who has not done something for a period of time even pretend to be able to teach that something?

OK, a necessarily rhetorical question in reality. Or perhaps not...

If Martin Fowler is indeed Martin Fowler, did he realize at the time I was lecturing that it was all theoretical waffle or did that realization only come later?

There is a flip side here: being an effective teacher/trainer is a skill in itself. A good teacher/trainer needs some real experience, but do they need more than is required to have empathy with the practitioners and what it is they do? Or is this just self-involved double-think?

I guess the real question here is what gives a person the right to be considered as, and consider themselves as, a guru on a given subject. Does being a guru allow one to pontificate? Or even teach/train? What balance of practitioner and teacher/trainer is the right one? Oh and is Martin Fowler Martin Fowler?

(*) Now don't get me wrong, I really like Python and presenting training courses on Python. Not only do I like doing this sort of thing, I like the income it generates :-)

[23:26] ( Comments)

The Scala user mailing list regularly and seemingly increasingly frequently has debates about attitude. This is not just about the attitude of the theorists to the pragmatists, it is about newbies to oldbies, trainers to practitioners, etc. Ultimately though, at the core, is the functional programming versus object oriented programming debate. The difficulty is that this is not being debated in a programming paradigms context alone, all the alignmnets of people on all the dimensions are conflated. A posting by Tony Morris:

I don't really buy into the "two worlds coming together" thing. There aren't two worlds -- it's an illusion -- a destructive one in my opinion. The issues at hand are so easily resolvable and indeed, they are regularly resolved for certain individuals. It's comes down to a matter of what is important to whom.

In the UK at least, there is very much a "two worlds", at least historically. In the 1980s, the functional programming (FP) folk set themselves up as people involved in the only form of computing that mattered. They got into positions of research power and hence research funding. But they became insular, and indeed dismissive of the non-declarative forms, e.g. object-oriented (OO). Also they did not worry about being relevant to folk making money from software. As OO rose in the late 1980s and early 1990s, the OO folk became an equally arrogant and insular crowd, who were as dismissive of non-believers as the FP folk were. The FP crowd began to demean the approach of being relevant to industry and commerce, a line being trodden by the OO folk. Both sides made jokes about any integrative work, cf. FOOPlog.

There was a real Paradigm War. You were on one side or the other, the two could never merge, cf. making FOOPlog the butt of jokes.

End result: C++ and Java are the only programming languages that matter in an industrial and commercial setting, with a smattering of C for the embedded systems folk. OO won the battle and the war. At least for the hearts and minds of CEOs, CTOs, CIOs and entrepreneurs. Not to mention the VCs.

C++ though led the way for the comeback of FP. The STL opened the door to a more FP approach to data structures in an OO world. Moreover template meta programming is FP.

Simon Peyton Jones (with Simon Marlowe) has been putting Haskell "out there" not just for its FP purity but because it can be made to be industrially and commercially relevant in an increasingly parallel world. Old views of concurrency are (finally after 40 years) almost irrelevant in the post Multicore Revolution world of ubiquitous parallel processors, involving many different hardware architectures.

Nonetheless there is a underlying and pervasive "two worlds" mentality. People are being engaged by the ideas of FP, are even looking to Haskell as an alternative to OCaml (which has a global interpreter lock (GIL) and so is problematic). Yet unless the there is a connection to C++ or Java, languages do not get actually used and hence absorbed into "real world" use.

It is not clear where the future is for native code, is it Go, is it C++, is it a merge of Miranda and C++ totally unlike the Objective-C mashup of C and Smalltalk?

On the JVM there is a much more obvious route given Groovy, Jython, JRuby, Clojure and Scala. Integration is available for free on the JVM, unlike the attempt via p-code 40 years ago.

Scala is in a position to bring the "two worlds" together in association with the Groovy, Jython, JRuby and Clojure communities. By not acknowledging that there really are, from a social perspective, "two worlds" out there, then the opportunity to create a new computing fit for the ubiquitously parallel world will be missed.

[07:26] ( Comments)

Last Wednesday evening I did my talk about GPars to the BCS Edinburgh Group. I am not sure whether the audience was bigger or smaller than usual, but most of the chairs set out in the room were taken, so about the number the organizers expected turned up. The biggest problem was not the projector, which seemed to “just work”, it was that the lectern was facing the wrong way, and so not usable, and the video cable wasn't long enough to reach a good place to talk from. Nothing the meeting organizers could do about it, but something the venue folk really ought to look to. In the end though a minor inconvenience.

For some reason or other, almost certainly a figment of my imagination, I had the belief this was to be a 90 min session and had planned accordingly. In reality the organizers were expecting 50 to 60 mins of presentation, with a hard deadline of being out of the building 89 mins after the session was due to start! Not a disaster, or even really a problem, but it meant rearranging the code presentation part of the session on the fly to just pick out the core highlights rather than the lengthier flow originally planned. Or to put it another way: sorry to the audience for a less than smooth presentation of the code.

Having said that, the audience reacted in the expected ways at the expected points in the session, and the questions and points raised were all on point and perceptive. Also the conversation at the adjournment to a local hostelry was good. I enjoyed it all, I hope the audience did.

Oh, and I finally got to meet one person I have been having email exchanges with for four or five years and another person I have been emailing with for two or three years without previously having met them. It was great to finally meet them both. It must be said though that the photos on the Web do not make it entirely easy to recognize them!

A PDF file of the slides I used for the session can be found here.

Thanks to Seb Rose for the invitation, and to whoever sacrificed whatever to ensure a dry and sunny two days to ensure The Royal Mile etc. was at it's best for September.

[09:21] ( Comments)

OK so I am referring to a report of a report of a report, but this article on The Register makes it absolutely clear that racketeering is not only legal but a whole business strategy when you have software patents.

Microsoft claim, but have never put to the test, that they have software patents that are infringed by Linux. Without facts, we have just FUD. FUD is though that which you can threaten people with. Microsoft are now bullying and cajoling manufacturers of Android-based equipment into signing protection deals, sorry IP licences. Since Microsoft have no input at all into the development of the software that is the focus of the deal, there can be no maintenance contract involved. It must be purely to do with software patents. Yet there is no proof that the patents exist, let alone apply to Linux. And even then this would only apply in those jurisdictions in which the patents have been taken out and approved.

Microsoft are clearly executing a protection racket, there can be no other word for it. Yet this protection racket is actually legal since it involves state approved monopolies, i.e. patents. So what we have here is state sponsored crime. Should the US government be put on trial for sponsoring crime? Won't happen of course – patents are tools of business, not of crime, and it is these businesses that keep the politicians in their positions of power.

Perhaps the Mafia should invest in a portfolio of software patents and thereby make their protection racket operations entirely legal?

Endnote. This article on Groklaw is interesting. Further reasoning as to why software is mathematics. Mathematics cannot be patented – well not yet, I am sure some big business will start making inroads on this soon. Till mathematics can be patented it would seem software cannot be patented. On the other hand try telling that to the big business who are making shed-loads of money from either extortion relating to, or licencing of, software patents.

[14:16] ( Comments)

Repeat announcement since it is now under a week away. Which means I probably need to think about writing the talk :-)

I am doing a session with the above title for the BCS Edinburgh branch on 2011-09-14 at 18:30, unsurprisingly in Edinburgh. Details are to be found on their website here.

[15:05] ( Comments)

In our “quite surprising” section: T-Mobile 3G mobile Internet now recording 1.38Mb/1.1Mb. That upload speed is really quite competitive.

[08:43] ( Comments)

As I have ranted about previously, Vodafone claim they have excellent 3G signal in B21 8EQ, and yet there is provably (by experiment) absolutely no 3G signal at all. Vodafone really do need to reconsider what their coverage map is saying, it clearly looks like an assumed expectation, rather than an experimentally determined fact.

I decided therefore on this trip to B21 8EQ to try other network operators. I purchased a T-Mobile dongle from the still open (this is a significant observation a week after the Clapham Junction “looting with serious violence” incident) and busy T-Mobile store. I was also going to get an Orange and a 3 dongle so as to compare. The 3 shop was trashed by the “looting with serious violence” of last week and was closed, whilst the Orange shop was pristine but no customers. The 02 shop was also trashed and closed. I took these observations as a sign and went to Carphone Warehouse for a 3 dongle. There are two Carphone Warehouse shops in Clapham Junction, both had been trashed in the “looting with serious violence” but both were in good state, open and busy.

After an uneventful, but extremely wet, drive to the B21 8EQ destination, I was eager to try things, so I did. Knowing that 3 block port 25, whereas Vodafone and T-Mobile do not, I tried Vodafone first. Why did I bother? Very little 2G signal and definitely no 2.5G, 3G or 3.5G signal. T-Mobile next. Very weak 3G signal but connection made. speedtest.net reported 0.6Mb/0.1Mb. Fairly poor. Then it dropped out to 2G. Very poor. But at least they allow HTTP, IMAPS, SSH, and SMTP. Next up 3. Poor signal but definitely 3G, speedtest.net reported 1.2Mb/0.3Mb. But no port 25 and I need SMTP. Bed time.

Next morning I try things again and put the dongles on the other side of the laptop. 3 gives 3.2Mb/0.6Mb. Healthy. T-Mobile 1.4Mb/0.6Mb. Passable. T-Mobile allows SMTP so wins – hands down.

Summary: Vodafone, get your act together and stop lying to people, and actually get 3G to where you say you have it. O2, care more about your business. 3, care more about you business, and get your act together and open port 25. T-Mobile get your act together and stop over-profiling TCP/IP connections.

PS Oh for a globally accessible WiFi, at a non-extortionate price.

PPS Why is mobile Internet so poor in the UK?

[08:31] ( Comments)

There has been an increasing number of articles recently coming out against software patents. The latest that I have seen are Martin Fowler and Brad Feld. I guess my problem is that the whole situation is simply depressing, not only of me but also of innovation in software. The actions that Lodsys are bringing (see for example Groklaw's archive) against developers of smartphone apps is clear indication that patent trolling is going to kill small company and individual software developer activity, leaving the field clear for the huge corporations to be the only organizations able to develop new software. This will definitely lead to a significant lowering of innovation. This is not because large corporates don't innovate, they do, but the pace of their innovation is driven by commercial pressures not by the ability of technology to support new and novel ideas.

What makes it all worse is that the situation is a self perpetuating one. Given that legal costs generally exclude individual software developers and small and medium size enterprises (SMEs) from any form of legal activity – be it offensive or defensive – the only players are the big corporates. Big corporate either have large patent war chests, or they have to buy them. Notice this is exactly what Google is having to do as a protection from Apple, Microsoft, and many others. If you don't have patent war chests, then when an organization with a big patent war chest decides to attack, you have no form of defence – there are only three forms of defence against a (assumed) valid patent attack, and that is either to have patents for a counter attack, leading to a cross-licencing deal, to have enough money to win in court against all comers, or to get the patents found invalid. Sadly this last relies on the judges ensuring cases are only heard after challenges are made to patents, and so is very unreliable – as Google and others are discovering.

The upshot of all this is that big corporates have to have patent war chests to stay in the game. Google is proving this. It means there is no way of stopping patents being the tool of business.

In some areas patents can actually be sensible: Dyson's cyclone technology for vacuum cleaners springs to mind. There is an application of physics leading to a new and novel technology for which the inventor deserves some protection so as to recoup development costs and make a profit. Less clear cut are drug formulas. New drugs cost a lot of money to research and develop, there needs to be some guarantee on a return on investment. So having patents on the formulas seems like a good idea. Of course, patent holders then charge extortionate amounts of money for their drugs especially to captive markets, e.g. the National Health Service. There needs to be compromise between using patented drugs, generics, encouragement of research and innovation, and profit margins.

Then there are user interfaces. Should these be patentable? My instant answer is no, and despite some issues, I'll stick with this. The problem is the Apple problem: Apple has spent many years building up a brand image based on certain user interface styles, architectures and techniques. Their brand deserves some form of protection from “knock offs”. It is just not clear that patents are the right tool. The whole of business processes really come under this heading as well. Why should processes be patentable? Are they ideas or inventions? Well they are both, but patents relate to physical artefacts not to ideas. So user interfaces should not be patentable, but products employing them should be. There must be a way for the lawyers to enshrine this in law?

Finally for this piece, software. I have no problem with software embedded in a physical artefact having the protection of patents. I have a huge problem with patents on software itself and software techniques. The obvious one I have used before in the patent of linked lists. Why should anyone have a patent on this idea. True these ideas have to be invented, they are inventions, but they are actually just new ideas explained in source code. By all means patent products using such ideas, but not the ideas themselves.

Of course the problem is that now software patents are an integral part of the USA way of business, they are immovable. The efforts of the USTR to force all signatories to ACTA to have software patents means all ACTA signatories will have to have USA style software patents. This will kill innovation dead. How can it be stopped? Only by direct political intervention. Big corporates cannot now get out of the game, indeed they have to get into the game – cf. Google. Politicians can put a stop to this, but will they? They seem to have been caught up in the web of power being pushed by the agents behind ACTA, a tool for big corporates to maintain control of brands. (Not a bad thing per se, but some of the side effects inserted into the tool are.) Moreover politicians have to be funded, this is generally by large corporates, and the politicians then basically have to support the business interests of their funders. We have Catch-22. The upshot seems to be that software patents cannot go away, even though they kill innovation. Time perhaps for all software developers to give up interesting things and get jobs with the big corporates? As I mentioned earlier perhaps depression is an appropriate state of mind?

Footnote: Only one person I know is a supporter of software patents. I think he is wrong, but he is entitle to his opinion – albeit wrong :-) He is though very clear that he thinks that the current system is broken. Patent trolls need to be removed from the system, and the granters of patents need to stop granting clearly stupid software patents. You know, if the system actually did this, there would be an awful lot less objection. Still wrong, but less so.

[08:17] ( Comments)

UPS batteries have a short life, 3–6 years on average. The usual indicator of a failed battery is a failed UPS self test. Till last evening, I had purchased a new battery immediately on a self-test fail, had it arrive next working day, installed the new one, and taken the old one to the recycling centre. No dramas, just a period of tension between old battery fail and new battery fitting, in case of power problems.

Last evening was special. Self test fail, new battery ordered, enter tension state as always. Then something new, horrendously nasty smell, and unusual irregular beeps from the UPS, not to mention the batter charge lights going off and coming on. So I turned off all the equipment, and found that the whole UPS casing around the battery was too hot to touch. I let it cool overnight, to discover the battery casing was severely mis-shapen. The previous evening the battery had got so hot it had started melting its casing. It is a good job the human olfactory sensors detected the problem, it seems the UPS internal sensors didn't have this as a use case.

[12:55] ( Comments)

I forgot to add a link to a PDF of the slides I used for my session “You Need to Know Python” at Skills Matter 2011-08-02T18:34+01:00 in my previous posting. The first link in the previous sentence rectifies this omission.

[10:38] ( Comments)

I presented an “In the Brain” session at Skill Matter last evening. It was initially intended to be focused on Python and why it is a real “must learn” language – hence the title You Need to Know Python. However it became more a Python in context sort of session.

I had initially assumed most people would have a little familiarity with Python, but it became clear fairly early on that there was a very wide ranging experience of Python, from the very knowledgeable to the “haven't tried it yet”. I therefore changed the content of various parts of the session on the fly. This is why there might have appeared to be a few discontinuities, well more than usual anyway.

As ever Skills Matter videoed the session and have already uploaded it, you can find it at: http://skillsmatter.com/podcast/scala/need-python

I enjoyed the session, and various people came up afterwards and said how much they enjoyed it. So overall a success. Thanks to everyone who came, especially those who chipped in and made the session interactive. Thanks to Skills Matter for hosting the event.

[10:27] ( Comments)

Given the tone of the writing in this article about Apple winning a patent judgement against HTC, the author gives all the evidence of being a shill for an interested anti-Android party (probably either Apple or Microsoft). Putting aside all the emotive stuff (which is almost certainly just FUD), there is the point that two patents have been cited explicitly. Both of these are on Google Patents:

1. US Patent 5,946,647
2. US Patent 6,343,263

Read them and fear.

Admittedly on a relatively cursory glance, and remember I am not a lawyer, just a person who does expert witness work, it would seem that Apple have US patents covering all software that manipulates data structures (5,946,647) and all devices that do real-time signal processing (6,343,263). Ok that seems to mean all software and all signal processing hardware are covered by these patents and so royalties are due to Apple for every sale of such a thing in the USA.

One has to hope that HTC files for re-evaluation of these clearly over-broad and obvious patents. I would suggest little need to find prior art to invalidate the patents (though I bet there is a real mass of that – every bit of software written and every digital electronic device created before 1990 really), they fail on these other grounds. Why were the US Patent Office (USPTO) issuing such obviously over-broad and obvious patents? I assume because that is part of the way of business in the USA. The USPTO issues a patent if the paperwork is in order, and allows the courts to decide everything else – thereby making the lawyers rich and everyone else not.

If USPTO had sent these patent filings to any sort of expert in the domain, the patent would not have been granted. So whatever you believe about the rights and wrongs of software patents – personally I am anti – it must be agreed that the USA software patent system is out of control, indeed broken. Will it get fixed? Clearly not if Apple, Microsoft and IBM have anything to do with it. These are the tools of protectionism for USA big organizations. They won't now give them up without a huge fight, one the USA government are unlikely to have an appetite for. Sadly.

Don't forget though the USA, via the USTR (USA Trade Representative) are trying to impose the USA system of patents on the rest of the world via ACTA (Anti-Counterfeiting Trade Agreement) . We really must fight this patent imperialism.

[08:02] ( Comments)

As some people will already have seen from the announcements on Twitter, I have teamed up with Skills Matter to provide a Python training activity. The initial public courses are workshops for people who have some programming experience, but little or no Python experience – see this page for details. I am sure, over time, there will be a demand for my intermediate, and possibly even advanced workshops.

Although this activity is new, the courses themselves are not. I have been presenting Python training since 2006 – and a Python user for many years before that. The courses are not of 5 year old material though. After each presentation, the courses are reviewed and updated in the light of the feedback from the presentation, and the state of the Python world at the time. So the material undergoes continuous evolution to stay in tune with the Python language, its philosophy and its practical use.

If you fancy coming on one of these courses contact the Skills Matter sales folk by phone on +44 20 7183 9040. Or even drop me an email.

If you spot anything you think is wrong or missing in the advertised course, please send me an email.

[05:10] ( Comments)

I'll be doing a session with the above title for the BCS Edinburgh branch on 2011-09-14 at 18:30, unsurprisingly in Edinburgh. Details are to be found on their website here. I repeat the marketing blurb here:

[13:53] ( Comments)

In this world of a plethora of programming languages, where new ones seem to appear on a weekly basis, C, C++ and Python have been, and remain, the main non-Java Platform languages. Even on the Java Platform there is Python, in the form of Jython. Gnome emphasizes C, C++ and Python, as does Canonical and many other organizations. The Python Software Foundation is definitely pro Python. Why is Python so "front and centre" and high profile? What makes Python so strong, used and loved as a programming language?

Even though Python is over 20 years old, it has evolved to stay modern and relevant. This is especially poignant given the Multicore Revolution that has caused so much change in C++ and so much distress for C. How does Python, a fundamentally sequential system in is CPython form, deal with the parallelism of modern hardware?

These questions, and I have no doubt many others, will be investigated in a free event I am running at Skills Matter at 2011-08-02T18:30+01:00. To sign up go to this URL.

If you don't sign up, we'll send Monty the Python round to put the squeeze on you.

[16:44] ( Comments)

This paper by James Bessen has stirred up the world somewhat regarding software patents – and a good thing too. As well as this Groklaw article (it's also worth looking at this one about the proposed changes in the US patent system), it even got to be a Slashdot article

The headline from the Bessen article is that small companies do not generally obtain patents, whereas large companies get them by the truckload, and that this is bad for innovation.

In the UK, software is not patentable per se, neither are algorithms. Bessen states that the case In re Alappat was the one that changed the nature of things in the US system, and allowed software and algorithms to be patented. Let us hope that the UK and EU do not allow the US to bully them into allowing software patents, and especially not via the tool of ACTA. What would be even worse of course is the US bullying the UK and EU to accept US patents as valid in UK and EU. This would likely kill small business software development stone dead. And UK software innovation would go out with it. Remember the US strategy is to support US development not to foster UK development.

What is the problem? Well a patent is enforceable even if the person violating the patent didn't know about it. Say you write some proprietary software and guard it as a commercial secret so there is no publication. Say a patent troll wants to target you. They find one of their seriously general patents which is bound to have been violated, cf. the patent on linked lists as an obvious example, they get an injunction to have your source code analysed to see if it is in violation. Along the way they will of course find others of their patents that might apply, and now they have access to your software,they can do lots of research on it. They then threaten to take you to court unless you pay licencing fees. At this point the issue is one of extortion (normally considered a criminal offence, and yet completely legal in the world of patents), you either suffer the indignity of paying licencing fees or you fight the case in court. Either way you lose financially. And most small companies cannot afford to go to court, it is simply too costly. So they either pay the fees or go into liquidation. Or sometimes, get bought by the patent holder for a trivial fraction of the value of the company.

So the situation where potential innovators do not play the patent trolling game but big companies do, means either potential innovators avoid any area that might be seen as competition to the big companies or they develop simply to sell to the big companies (not to create their own business). The big companies thus drive the agenda, usually for their own ends, and genuine innovation is sacrificed at the altar of big company profits and stock price.

The conclusion is inescapable: software patents destroy innovation in software development. Anyone professing to support innovation, must therefore be against software patents. Yet the UK government is towing the US party line and supporting ACTA and hence software patents. Clearly the UK government wishes to terminate all software innovation in the UK.

Sadly though software patents are now an integral part of the US business model. Witness the way Microsoft is using its unnamed patents as ammunition to force companies making Android phones pay licence fees for what it claims is its intellectual property. What intellectual property can Microsoft possibly have embedded in the Android system? Microsoft has made no contribution to Android and Android is not based on any Microsoft technology. Microsoft of course will not say and the contracts of licencing require silence from the licencee. This is definitely a “shakedown” operation by Microsoft. Although in any other area than licencing patents this sort of extortion would be illegal, with patents it is a valid business operation. So you can't blame Microsoft for employing valid business tools even if it is evil and objectionable.

Clearly the “conspiracy theory” goes along the lines of: Microsoft are trying to validate claims they will eventually take up against Google and Oracle, by having a lot of small companies already signed up as licencees. The eventual goal is clear, Microsoft are failing to keep their position in the market by creating new and possibly innovative products, so they are turning to software patent trolling as a way of generating an income stream from other people creating new and possibly innovative products – well at least products that people want to buy, in increasingly large numbers.

It would be good for the US government to put a stop to this software patent trolling business so that everyone can get on with innovating and letting the market decide who makes money and who doesn't. Software patents are a tool of protectionism, indeed a protection racket. It is a mechanism for squashing innovation. Let us hope the UK and EU governments recognise this and do the right thing for innovators. Make it clear that software and algorithms are not patentable, and let the innovation start.

[10:19] ( Comments)

A few days ago, Debian Testing changed its Linux kernel from 2.6.38 to 2.6.39. My Wifi connection between Lenovo X201 and Lenovo T500 laptops and the Netgear DGN3500 suddenly stopped working. Oh it would start fine and the laptops would connect, then a short while later, or sooner if there was a steady heavy load, the Netgear would crash requiring a power cycle – even the HTTP server console froze. This never happened with 2.6.38.

I reported this problem to the Debian bugtracker only for the bug to be immediately closed as "So there's a bug in the router". I beg to differ that it is so certain. Anyway the triager has agreed that should the evidence become more than a single data point, then the bug would be reopened and a message sent upstream that there is a problem. Personally I think the iwlagn folk in the Linux development community really ought to be very interested in the problem that has suddenly started occurring.

I also reported the issue to the Netgear forum – I had forgotten how much I hate fora, email for me any day. Someone has also spotted someone having a similar problem with a different Netgear router, and a different distribution, but the same Linux 2.6.39.

I tried some experimentation varying the WiFi connection speed on the Netgear and varying the kernel version. I also tried Ubuntu Natty as well as Debian Testing, both the 2.6.38 kernel and 2.6.39 kernel – Natty only has a 2.6.38 kernel as far as I can tell.

• The 2.6.38 kernel works fine with the router at “upto 270”, “upto 130” , and “g only”.
• The 2.6.39 kernel fails to work after a period or with higher loads at “upto 270” and “upto 130”, but works fine at “g only”.

• has diverged from the standard as implemented in the router; or
• has corrected a divergence from the standard, that is incompatible with the router.

I would have thought the Netgear folk would be investigating this now that two separate people have reported it independently on different Netgear devices. Also the Linux/iwlagn people really ought to be interested as to what change they made that is causing this problem to exhibit. We shall see.

[15:28] ( Comments)

A small postscript to the previous post: I get the following stats:

    +----------------------+-------+-------+
    | Name                 | Files |  LOC  |
    +----------------------+-------+-------+
    | Controllers          |     1 |     7 | 
    | Domain Classes       |     1 |    10 | 
    | Services             |     1 |    11 | 
    | Unit Tests           |     3 |    27 | 
    | Integration Tests    |     1 |    15 | 
    +----------------------+-------+-------+
    | Totals               |     7 |    70 | 
    +----------------------+-------+-------+

The Unit Tests are not listed in the book. So removing them I have 43 lines of code, which I think compares well with the 55 in the Smith and Ledbrook book :-)

[15:13] ( Comments)

ACCU is running a mentored developer project on Grails using the Grails in Action book by Glen Smith and Peter Ledbrook. The place to start is Chapter 1 which is a rush through the main concepts of a Grails application. I did it once using Emacs and command line. Thought to try the whole thing in Eclipse using STS, but never got to that before moving on to Chapter 2. This is a set of notes based on me redoing the Emacs and command line sequence, to provide notes for someone.

Installed Grails 1.4.0.M1 to a place and ensured an alias to the right place so that grails works from the command line.

"grails create-app qotd" works fine. If this is the first time of running grails it seems very slow as it is downloading stuff -- I wonder what happens if you are not connected to the Internet? – and it seems extraordinarily slow in this case if the filestore in NFS mounted rather than local disc. Fortunately this delay all seems to go away if grails has been used before.

"cd qotd", "grails run-app" work as expected. Very seriously slow starting up on NFS mounted filestore compared to local disc. Finally decides to load resources, hibernate, tomcat, and jquery.

Am seeing lots of Gant output :-) (wander around http://gant.codehaus.org and the source code to get the point).

Finally Grails is ready and . . . after a long delay . . . the expected Grails page shows on http://localhost:8080. Doesn't look anything like the image in the book of course as this is Grails 1.4 not 1.1, which is what was used for the book.

Grails and Groovy community seem to favour Git over any other version control system, so rather than fight the community by using Mercurial or Bazaar, decided to use Git. Followed the instructions at http://www.grails.org/Checking+Projects+into+SVN#Adding%20a%20new%20project%20to%20Git Methinks this should be a separate page, and there should be stuff for Mercurial and Bazaar as well – Git is not mandated by Grails. Then found out about https://github.com/github/gitignore/blob/master/Grails.gitignore so used that instead. It seems that Grails deals with IntelliJ IDEA as well as Eclipse.

OK, ready to move on, so create a controller: "grails create-controller quote" seems to do the right thing but slow on NFS mounted filestore. This comment about NFS mounted filestores comes up at all stages, so I shall now assume you appreciate this.

Added the home closure (Groovy calls them closures, but technically they are not, they are lambda functions -- the Groovy community knows that they are abusing the jargon term but it's all too late now :-( Grails notices the file change and compiles the file but there is an error of some sort. "Failed to reload file". This is reasonable as the file didn't exist when Grails was started, so a restart needed.

Did I mention how slow Grails is to start on NFS mounted filestore?

Anyway success. Added the index redirect. Updated file compiled and seemingly reloaded. Redirect works. Excellent. Commit.

Added Emacs backup files to the Git ignore list.

Added a view and it worked!

Added the style template. Begin to wonder if SiteMesh is really worth it. Now discover that snazzy.css and logo.png are not found. Reasonable really, the default Grails project shouldn't contain them! The book really ought to say something about putting these things in the web-apps directory. I didn't think I'd have to do a Grails restart but I did. Then discovered snazzy.css refers to three other image files background.png, logo_background.png, and menu_background.png. Found versions, put them in . . . failure. Restart Grails again. I get the feeling that adding any new file requires a restart.

Onto domain models. "grails create-domain-class quote" works fine. Listing 1.10 is slightly wrong, don't use what is printed there ("jdbc:hsqldb:file:devDB;shutdown=true") instead use "jdbc:h2:file:devDB".

Short-circuited the loading of the database by writing a Groovy script and then executing it in the Grails Console. Noted that qotd.Quote has to be explicitly imported, which the book doesn't mention.

Did I mention things are extraordinarily slow over NFS mounted filestore?

Apparently Grails not finding jquery at this point is not a problem.

Amended the controller to present a random message. Worked fine. Can't say I like the coding style of Listing 1.12, far to imperative and not nearly declarative enough.

  
def random = {
    def allQuotes = Quote.list ( )
    [ quote :
      ( ( allQuotes.size ( ) > 0 )
        ? allQuotes[ new Random ( ).nextInt ( allQuotes.size ( ) ) ]
        : new Quote ( author : 'Anonymous' , content : 'Real Programmers don\'t eat much quiche.' ) )
    ]
}

Scaffolding gives a CRUD interface by adding a single line to the controller. Magnificent.

Noted that all entries created via the scaffolding are created at 00:00 -- how to add a time to the date of creation?

Grails gets confused by the .#Quote.groovy file created by Emacs.

Added constraints, but the screen doesn't quite look like the one in the book, no time, just a date.

Created the service, spotted error in book, path shown is grails-app/services/QuoteService.groovy but should be grails-app/services/qotd/QuoteService.groovy

The field transactional is not in the default service as in the book, presumably a 1.1. -> 1.4 change. Listing 1.15 still hideously imperative and insufficiently declarative.

class QuoteService {
  boolean transactional = false
  def getStaticQuote ( ) { new Quote ( author : 'Anonymous' , content : 'Real Programmers don\'t eat much quiche.' ) }
  def getRandomQuote ( ) {
    def allQuotes = Quote.list ( )
    ( ( allQuotes.size ( ) > 0 )
      ? allQuotes[ new Random ( ).nextInt ( allQuotes.size ( ) ) ]
      : getStaticQuote ( ) )
  }
}

Started adding the tests. Spotted that the code listings had the wrong class names QuoteService Tests should have been QuoteServiceIntegrationTests. I hate * includes, not sure why but I do. the Grails template integration tests is full of them. Also of course the book is using Grails 1.1 and hence JUnit3 whilst Grails 1.4 uses JUnit4 so the test files are very different looking. But this is good. Using TestNG would of course be better than using JUnit4. But then using Spock would be even better. Must find out how to use Spock for these Grails tests.

Test reports go in a different place that stated in the book, I found them at target/test-reports/html/index.html. We are green :-)

Funny I always though AJAX was a scouring powder like VIM, apparently though it works well with Grails. I use the word apparently as it seems http://jira.grails.org/browse/GPRESOURCES-66 means you have to hack a lot to get it working. Something to do with JavaScript -- or should that be ECMAScript, is there actually a difference?

I think at this point I am going to quit, I'll leave AJAX hacks until I need them. Maybe actually try to do this with Eclipse/STS.

Final note: Grails is going to be great fun, and very efficatious at creating the Web applications I am seeking to build.

[14:51] ( Comments)

Someone pointed out I had said that Jack Kreindler was at 56 Harley Street, whereas it is actually 76 Harley Street, a very different place. Correction now in place.

Yesterday I attended the ONEIS Open Day. ONEIS is a new way of managing information created through a collaboration between Ben Summers and Jennifer Smith. Ben is a programmer and Jennifer an information management expert. Between them they have created a system which may be the harbinger of doom for the relational database as the way most people manage their information.

Saying that the demise of relational database is nigh, may seem a bit over the top. But we have the NOSQL people (MongoDB, CouchDB, Cassandra, etc.) bringing new ways of managing persistence, and now there is ONEIS which is a data store, but is neither SQL nor NOSQL. ONEIS treads another path, that of the object store. The critical thing that ONEIS brings to the persistent data store game is that of not enforcing a schema that remains static throughout the life of the repository. Instead, data and meta-data are held as objects with links. Simple, obvious, and very effective. Why is it so effective in the eyes of the current user base? Because of the attention to the detail of the user interface and user experience.

Having a domain expert be half the development team, it is nigh on impossible not to have a system that works for the domain expert. In this case the domain expert has had an excellent grasp of what works and doesn't work for a wide range of people using the system.

The presentations by the ONEIS folk were interspersed with "stories from the trenches" of people already using the system. Most of these were interesting, and certainly supported the marketing of ONEIS as the next big thing for users, but one in particular stood out as a show case for what is wrong in traditional information management systems and what is exciting about ONEIS.

Jack Kreindler, Medical Director at 76 Harley Street, a private medical practice, definitely not part of the NHS, talked about how they are using ONEIS to create a complete, useful and usable electronic medical record system. The splendid thing about their system is that it works for the patient, the medics and the practice. This is something the NHS electronic patient record system can only dream of becoming. Yet Kreindler's system is exactly what the NHS system should aspire to be.

What is so different? Well two things. First Kreindler's system has evolved from a small system that worked to a larger system that works, albeit with a very different schema to the original. Secondly, it has been evolved by the users as they use the system, i.e. it has changed in response to perceived user needs. Contrast this with the NHS system. Firstly it is a government procurement which implies massive upfront specification and huge documents, almost none of which is actually useful -- even though very large amounts of money are given by the government to the contractor. Secondly, there is an attempt to create a huge working system by direct fiat, and as John Gall has pointed out in his work on Systemantics, big upfront design, and direct creation of big systems, just never works. The government is seemingly unable to learn this simple lesson. Time and time again, huge projects are let and fail. Or is there a conspiracy theory here that the contractors have a deal with the government to extract vast amounts of our tax money in exchange for doing nothing useful?

The moral is of course that Kreindler's system looks like an excellent prototype for the NHS to pick up on and evolve, stage by stage, into a national electronic patient record system. Sadly this is highly unlikely to happen as government procurement is so entombed in waste and failure, and isn't going to change.

But let's go out on a high note. The beauty of ONEIS is that it allows the schema of the repository to change over time according the the needs and desires of the users. Relational databases just do not have a chance in the face of user evolvability to match need that the ONEIS way of doing things provides.

[14:11] ( Comments)

I can't remember how exactly I got there, but I alighted on this webpage: Enough is Enough, by Fred Wilson. The article is a reaction to the ongoing cases brought by Lodsys against Apple and Google and, more terrifying, against individual applications developers. Lodsys appears to be a patent holding company that does nothing but licencing patents, i.e. it is the archetypal patent troll.

Wilson opens his article with:

I believe that software patents should not exist. They are a tax on innovation. And software is closer to media than it is to hardware. Patenting software is like patenting music.

So here we have a venture capitalist (VC), i.e. a person who buys innovation for a living, saying that software patents should not exist. He is not the only one. Most VCs I know think patents are a waste of time for startups unless you want to get into the patent trolling business. If you actually want to create, market and sell innovative product, you are far better off using copyright and commercial secrecy as tools to protect IPR.

Big Money will of course object to this since patents have become an integral part of their business model. Fine if you are Big Money, useless if you are a small startup generating innovation. And everyone agrees startups are where innovation happens not in big organizations. So a patent-based business model is pro Big Money and anti startup, and hence anti-innovation.

Sadly I fear the patent system is so deeply enshrined in Big Money's view of business that it will not allow the system to be changed, no matter how much the VCs complain. To those that have I guess.

[12:59] ( Comments)

For reasons which are unlikely to become apparent any time soon, I find myself in need of being much less laid back about income generation -- even though I am currently involved with a pre-money startup and writing a book. I have ongoing training gigs, but I need to increase income from that level. For various reasons I need to telework rather than commute.

I have a CV (of sorts) on my personal website http://www.russel.org.uk/cv.html. The short form summary is skilled and interested in: Python, Groovy, Java, Scala, D, Go, Ruby software development generally; parallel and concurrent programming in Python, Groovy/Java/GPars, Scala/Akka, Go, D, C++; SCons, Gradle, Maven and Ant build frameworks; Bazaar, Mercurial, and Git distributed version control systems; LaTeX typesetting.

My accountant tells me it is best for me to work as a sole trader, which I do under my own name or, where the people letting the contract cannot work with named individuals, under the trading name "Winder Information Systems Engineering" -- which is my sole trader trading name registered with HMRC.

If the people letting the contract require to deal with a company, or there are liability issues which preclude working as a sole trader, then I trade through It'z Interactive Ltd (http://www.itzinteractive.com) -- though formally this is dormant just now as I have not undertaken any of this sort of activity for 5 years or so.

So if you have any consulting or training needs which I can fulfil, please contact me. If you have funds for projects, but no internal resource, where a project is amenable to being undertaken on a contract basis via teleworking, I would love to hear from you.

[10:45] ( Comments)

The United States Patent and Trademark Office (USPTO) has been active again in trying to amend processes. This time it is all about “re-examination”, and how to make it more streamlined. See for example http://www.uspto.gov/news/pr/2011/11-28.jsp and this conference http://www.law.berkeley.edu/10988.htm – thanks to Groklaw's "Groklaw Latest News Picks" of 2011-05-29T05:55+01:00. Also of course this Groklaw article about recent events in the case colloquially known to everyone as “Paul Allen vs The World”. I am sure there are many other examples.

In a sense it is great news that USPTO are looking to the processes involving patents in the US. Anything that appropriately streamlines useful processes is to be applauded. You are though waiting for the “but . . .”

The point that should be obvious is that this is all about re-examination. They are looking to the processes that involve the courts due to litigation bringing doubts as to the validity of a patent. Of course it is good to make these processes simpler and easier, but why is this the focus of attention? Surely the focus of attention should be on why there are so many patents out there that have so much doubt and uncertainty surrounding them: why are there so many (software) patents approved that then come under challenge from the courts? The obvious conclusion is that the initial examination is inadequate, that (software) patents are being granted without proper “due diligence”. One has to doubt whether there is any technical examination at all undertaken by the USPTO, that they only consider the syntactic correctness of the documentation and the following of mandated processes in awarding a patent.

So the USPTO should look to its examination processes as well as its re-examination processes. There should be assessment of potential prior art. There should be assessment of validity of the claims. There should be assessment of the obviousness of the invention. Even cursory examination of the patents involved in all the high profile software patent cases currently before the courts in the US lead to the conclusion that very little notice was taken of the technical content of the application in awarding the patent.

I believe that software and processes should not be patentable since they are not product with a physicality, they are just a way of expressing ideas. However, even if such things were acceptable, the USPTO do not seem to have employed any form of proper quality control over the patents they have issued. One can only infer that they “approve and let the courts decide”. Is this really a sane way of implementing monopolies approved by the state? (*) Well if you are a patent lawyer, of course it is. It is a “gravy train”, a licence to print money. If you are an inventor or entrepreneur, it is definitely not, it is a massive threat, with potentially huge legal costs. Big businesses may be able to set aside monies as contingency for this, but the lone inventor, and the small and medium enterprises (SMEs) that are supposedly the darlings of the western economies certainly cannot. So the process as it is favours big businesses; it is by and for them.

I think the obvious conclusion is that USPTO should undertake a root and branch re-examination of all of their processes, especially the ones at the beginning of the cycle. The US government clearly needs to step in and ensure that this quango gets put onto a proper technical footing. Unless of course it is culpable in preserving the status quo.

(*) Doesn't it strike you as ironic that states have so much legal infrastructure to stop monopolies from existing – anti-trust laws, competition laws, etc. – and at the same time have a legal infrastructure (patents) for enforcing monopolies?

[06:27] ( Comments)

The original post was wrong in some fundamentally fundamental ways, so much so that I have taken the liberty of editing the content to be correct.

In JSR166 and the Maven repository I reported that JSR166 snapshot artefacts were newly present in the Codehaus snapshots repository. In breaking news, I can report that Doug Lea (helped by Tim Peierls) has tagged the master JSR166 repository with the tag release-1.7.0 so that a non-snapshot release can be made, and artefacts appear in the main Maven repository. I have employed my Git clone of the master JSR166 repository, with its gradleBuild branch, to build artefacts and upload them. They are in the Codehaus repository now and will appear in the main Maven repository at the next synchronization.

Details of the 5 artefacts:

Each of the artefacts has source code and javadoc files as artefacts in the repository using the now standard mechanism of using the classifiers “sources” and “javadoc” &ndash with the binary artefact having no classifier.

The jsr166, jsr166tck, and jsr166x artefacts are there for completeness, the ones people will be interested in are jsr166y and extra166y ones: jsr166y comprises all the new classes that will be standard in JDK 7, whilst extra166y comprises the classes that will appears in JDK 8.

For further details or any questions about the artefacts, contact me directly. For information about JSR166 see the project home page.

[15:47] ( Comments)

There was a regression in the 1.9.4 release of Gant that could not be ignored. The problem has been corrected, and a new release candidate checked by the people who found the regression (thanks due to Jeff Brown and the Grails folk). There is therefore a shiny new Gant 1.9.5 release. Everything should be in place at Codehaus already and Maven will get updated as soon as the Codehaus -> Maven sync happens.

The regression whilst essentially trivial, was sufficiently fundamental that I have taken the drastic step of removing the 1.9.4 release from everywhere. This should have no effect on people using Groovy 1.7 or 1.8, they should use Gant 1.9.5. People still using Groovy 1.6 will not see a Gant 1.9.5 and will have to stay weith Gant 1.9.3. Of course people still using Groovy 1.6 should upgrade to 1.8.0 immediately and therefore not have a problem with Gant :-)

Thanks, and sorry for the hassle.

[09:54] ( Comments)