HPC Courses, Heartbleed Bug, Dingo Species

Three days this week was pretty much taken up by conducting Linux, HPC, and MPI training courses for a number of postgrads, almost all from RMIT. Another good class who started with zero background in the subjects and who by the third day were working their way through MPI programming. Of some note was the attendance of the first economist in the three years or so that I've been running these courses. I've had a longstanding desire (and it really wouldn't be that hard) to compare local economic development with different council rating systems as a time series - not really requiring HPC but certainly does require a geospatial person, a valuer, an economist, and a programmer to be on the same project.

On another significant IT-related issue is Heartbleed which, hopefully, most people have heard of by now. From a technical perspective a failure in the bounds-checking by some versions of OpenSSL to malformed heartbeat requests allows for an attack vector on server memory of affected systems. From a user's perspective it means that passwords on many major sites have been compromised over the past two years. Whilst Filippo Valsorda has produced a useful tool to check whether a site is currently affected, there is no easy way to check if a certificate have been re-keyed. What can be provided is a handy list of many sites that were compromised last Tuesday, a day after the bug was made public.

Recently a study has been released claiming that the dingo is a separate species. Whilst it looks serious enough the university press release on a certain date, made me think that this could be an "poisson d'avril" as the proposition is so counterintuitive. Certainly species is a complex subject with some interesting edge cases (e.g., hybrids, ring species), but the general principle of species representing a population of organisms capable of interbreeding and producing fertile offspring is true. The paper argues - through morphological analysis of museum examples, not genetics - that the dingo was a separate species some five thousand years ago and is now being threatened by extensive hybrid speciation. Personally, I find their claim highly dubious. Not only is there significant evidence that the dingo's arrival was somewhat earlier, mitochondrial DNA analysis indicates that the dingo has much less differences with the domestic dog than wolves. So whilst morphological analysis is useful for a quick rule of thumb, genetic analysis is a much better determinant.

This entry was originally posted at
  • Current Location: Willsmere
  • Current Mood: contemplative
Tags: , ,
Re dingo: Among my wolf books are pictures of the Indian wolf, found in the area from which the dingo was brought to Australia. I may not be up on the science, but these animals are very close in appearance to the dingo. I've always accepted that the dingo, like other dogs, is descended from wolves, albeit closer to the wolf than many. So I believe that 5000 years ago, they pretty much were wolves.

[Back in my box now. Wolf research is sort of a thing.]
Well, I don't know that much about the subject matter myself, but I understand that the "lumpers" in taxonomic classification tend to argue for wolves and dogs to be considered as subgroups of the same species because of the evidence of fertile interbreeding.
Lev, Lastpass's Security check feature will actually show which sites you use are effected and when they last changed their SSL cert. Tells you to wait before changing password if they haven't yet.