30.11.13
Fame, almost
28.11.13
Students support strike
Will Hutton, writing in the Guardian, described the situation succinctly:
The real wages of academics have fallen by 13% since 2008, one of the largest sustained wage cuts any profession has suffered since the Second World War.University administrators offer 1% and refuse to negotiate. Faculty struck on 31 October and a second strike is called for 3 December. As usual, our students put the case more eloquently than our union (the UCU, University and College Union). The leadership of EUSA (Edinburgh University Student Association) write:
Tensions are high within the University, as the second day of strike action draws nearer and many continue to work to rule. At such times we may hear the oft-quoted: “For the sake of students, don’t go on strike!”The photo above shows me picketing the Informatics Forum, surrounded by students during a previous strike in 2011.
We write on behalf of EUSA and our 32,000 members, to actively encourage you to take strike action. In the short term this will indeed affect our education, but the long term benefits are significantly vaster.
It is critical that students and staff struggle collectively. Not only to ensure that the sector continues to attract the highest calibre of people, but also so staff are able to focus on the job – not worrying about the rocketing cost of living. Needless to say, colleagues at the start of their career are hardest hit, including the thousands of EUSA’s postgraduate members who help teach.
The demands are reasonable, and the more effective the action now the sooner we can get back to the reason we’re all here – education.
It is for this reason that EUSA’s academic reps, who aren’t exactly our most radical bunch, have voted overwhelmingly in favour of actively supporting the on-going strike action. This decision has been met with broad approval from across the student body. Dozens of our reps and countless other students were out on the picket lines at 7am the other week, and have vowed to be there next time.We are actively encouraging our members not to cross the picket lines and to study from home.Education continues to be progressively marketised, fees continue to rise, power continues to shift away from ordinary staff and into the hands of the overpaid in Old College. At times like these it is vital that the university acts as a community and reasserts its stake over the corporate body. We do this by working together, and recognising that our struggles are in common.
So again, we implore you, on behalf of your students, not to undermine the strike. And hopefully we'll be seeing you on the picket lines!
In solidarity,Hugh, Nadia, Alex and KirstyThe EUSA Sabbatical Team
Labels: Politics, UK, University
26.11.13
Amdahl's law for predicting the future of multicores considered harmful
The following will be the subject of a talk at Edinburgh on Thursday. I won't attend, because it's Thanksgiving, but it looks interesting.
Amdahl's law for predicting the future of multicores considered harmful
B.H.H. Juurlink , C. H. Meenderinck, ACM SIGARCH Computer Architecture News, Volume 40 Issue 2, May 2012, Pages 1-9.
Several recent works predict the future of multicore systems or identify scalability bottlenecks based on Amdahl's law. Amdahl's law implicitly assumes, however, that the problem size stays constant, but in most cases more cores are used to solve larger and more complex problems. There is a related law known as Gustafson's law which assumes that runtime, not the problem size, is constant. In other words, it is assumed that the runtime on p cores is the same as the runtime on 1 core and that the parallel part of an application scales linearly with the number of cores. We apply Gustafson's law to symmetric, asymmetric, and dynamic multicores and show that this leads to fundamentally different results than when Amdahl's law is applied. We also generalize Amdahl's and Gustafson's law and study how this quantitatively effects the dimensioning of future multicore systems.
Labels: Computing, Concurrency
20.11.13
PhD Studentship on ABCD
Please circulate the following to those who might be interested.
We are recruiting for one PhD student to work on design and implementation of programming languages. The post is on the project "From Data Types to Session Types: A Basis for Concurrency and Distribution".http://homepages.inf.ed.ac.uk/wadler/abcd-phd-advert.html
The project has particular emphasis on putting theory into practice, embedding session types in a range of programming languages and applying them to realistic case studies. The research programme is joint between the University of Edinburgh, University of Glasgow, and Imperial College London, and includes collaboration with Amazon, Cognizant, Red Hat, VMware, and the Ocean Observatories Initiative. We have a programme grant funded by EPSRC for five years from 20 May 2013.
The successful candidate will join a team responsible for extending the functional web programming language Links with session types to support concurrency and distribution. We will test our techniques by providing a library to access Amazon Web Services (AWS) cloud computing infrastructure, and perform empirical experiments to assess how our language design impacts the performance of programmers.
You should possess an undergraduate degree in a relevant area, or being nearing completion of same, or have comparable experience. You should have evidence of ability to undertake research and communicate well. You should have a background in programming languages, including type systems, and programming and software engineering skills.
It is desirable for candidates to also have one or more of the following: a combination of theoretical and practical skills; experience of web programming or cloud programming; knowledge of the theory or practice of concurrent and distributed systems; knowledge of linear logic; or training in empirical measurement of programming tasks. We especially welcome applications from women and minorities.
We seek applicants at an international level of excellence. The School of Informatics at Edinburgh is among the strongest in the world, and Edinburgh is known as a cultural centre providing a high quality of life.
The successful candidate will receive a studentship covering tuition and subsistence. Students from the UK or EU are preferred, but studentships may be available for overseas students with strong qualifications. Applications should be received by 13 December to be eligible for the full range of scholarships. Consult the University of Edinburgh website for details of how to apply.
Enquiries can be addressed to: Prof. Philip Wadler (wadler@inf.ed.ac.uk), Principal Investigator of the ABCD project.
Labels: Computing, Distributed Computing, Functional Programming
17.11.13
The Monad Tutorial Fallacy
Brent Yorgey captures the problem with many monad tutorials: they start with monads, rather than with concrete examples of the abstraction that monads capture. Dan Piponi's introduction to monads avoids this pitfall, as does, I hope, my reprise of The First Monad Tutorial, to be presented at YOW! in Melbourne, Brisbane, and Sydney next month.
Labels: Functional Programming, Haskell
16.11.13
Economics of Scottish Independence
Would Scotland be better off or worse off under independence? I've often been told that England subsidises Scotland, that Scotland already gets "more than its fair share", and that independence would bring that to an end. Is it true? Before buying independence, I'd better perform due diligence.
I'm in favour of independence, so I wanted to examine the figures provided by someone against.
I found this report by Brian Ashcroft, an emeritus professor of economics from Strathclyde, who argues Scots would be better off remaining with the English. (His collected posts on independence are here.) Examining his argument for dependence leads me to the conclusion that independence is essential.
The graph above, taken from Ashcroft's report, shows that in recent years, Scottish spending has exceeded Scottish revenues (including from oil). Uh oh! However, he mentions briefly, and a web search confirms (see graph below, from The Guardian) that UK spending has exceeded revenue by a far larger amount.
I think this shows that Scotland is in better shape for independence than the UK as a whole, because we would have less debt, assuming it was fairly allocated. Ashcroft argues that Scotland has gotten back what it contributed in oil revenue, so that everything is hunky dory, but that doesn't ring true if one considers that over the same period the UK margin of spending over revenue was far greater than Scotland's.
Ashcroft's most significant point, in my view, is that going forward Scotland may have to pay a higher rate for borrowing than the UK, but if the Scottish economy is better managed for growth than the UK, then this would pale into insignificance.
As the graphs show, we have a problem: but it is a smaller problem than the UK as a whole, and we have a better chance to solve it on our own. Better together? I don't think so.
Labels: Independence, Scotland, Yes!
9.11.13
Wealth Inequality in America
This 2012 video from Politizane does an excellent job of illustrating the massive, well-documented gap between the wealth-distribution that Americans believe they have, the distribution they would favor (regardless of political affiliation), and what America actually has: a system that rewards CEOs at 380 times the rate of their average employees.Spotted via Boing Boing.
7.11.13
IEEE statement on appropriate use of bibliometric indicators
The IEEE recently agreed a statement on appropriate use of bibliometrics.
Recently, citation counts and proxies thereof have risen in importance as fundamental elements in the determination of the scientific impact of entire departments or universities and research centers [8], funding evaluations of research proposals and the assessment of individual scientists for tenure and promotion [9], salary raises [10], or even to predict future career success [11]. While the first use is technically appropriate, provided it relies on data collected from a sufficiently large set to provide a statistically meaningful analysis, this condition is never satisfied when applied to individual scientists.
Labels: Computing
2.11.13
Is Perl syntax better than randomly chosen syntax?
Programming language designers should perform more empirical studies of the sort published at Plateau. Last night I saw a paper which compared the author's language, Quorum, to Perl and a 'placebo' language with syntax chosen at random, dubbed Randomo.
Perl users in our study performed notably poorly, not only performing less well than Quorum, but no better than a language designed largely by chance.While Perl has never had a particular reputation for clarity, the fact that our data shows that there is only a 55.2 % (1 - p) chance that Perl affords more accurate performance amongst novices than Randomo, a language that even we, as the designers, find excruciatingly difficult to understand, was very surprising. This is especially true, we think, considering we chose to test only the syntax in Perl that is relatively common across a number of languages (e.g., if statements, loops, functions, parameters). Considering that Java syntax, which many would arguably consider to be easier to understand than Perl, uses similar syntax, we are curious how it would perform. Given this interesting first result, we plan to test a number of additional languages using the same procedures.
An Empirical Comparison of the Accuracy Rates of Novices using the Quorum, Perl, and Randomo Programming Languages. Andreas Stefik, Susanna Siebert, Melissa Stefik, and Kim Slattery. Plateau 2011, Portland, 24 October 2011.
Labels: Programming Languages, Science