Lev Lafayette's journal

Every year I do a series of invited lectures and assessment for the UniMelb master's-level course "Cluster and Cloud Computing", which is a rather massive course with approximately four hundred students primarily from computer science and data science. Anyway, this has started for 2025 and will continue for the next few weeks. In between all this I have two advanced researcher workshops next week on "Parallel Processing" (shell commands, job arrays, OpenMP and OpenMPI programming), and "Mathematical Programming" (R, Octave/MATLAB, Julia, Maxima, Stata, etc). Further, I recently gave a presentation to the Spartan Champions group on scheduler-level parallelisation. In most cases this is the easiest way to do a type of parallel data computation and, as such, makes a low-hanging opportunity for making the best use of computing resources. Finally, on a slightly related matter, I have put up a script (called "lament") designed for local viewing of Facebook encrypted messages, converting the JSON to HTML conversion of Facebook Messenger files using the Pandas and JSON extensions for Python.

I have had the pleasure of catching up with interstate visitors this week! Firstly, I had the delight to catch up with Justine, Simon, Erica, and Susie C., for brunch with "Le Cafe Flo, a rather good French cafe in Thornbury. It was conveniently located to "The Witches Wardrobe Flea Market", which had a good variety of goth, emo, etc. wares. It was a bit of a Perth migrants day for our part, enhanced with visitor Susie C., who is making a lot of noises about making the trip across the Nullabor on a more permanent basis. Another visitor this week is my dear friend (and co-owner of property) from Darwin, Lara D. Lara is blessing me with two weeks with her company, and we have quite the range of artistic, musical, and culinary delights lined up, although it will be hard for the latter to beat the dinner provided by Carol D., Lara's mother, last night.

One of the ways that affirmation occurs for those who live by the keyboard is that your essays are read and then circulated. I've had two such experiences recently. The first was my letter to members and associates of the Australia China Friendship Society (ACFS) Victoria on the 75th anniversary of its founding of the People's Republic. Apart from the hundreds of people it is distributed to normally, local members and other state branches have also asked me if they could forward and republish. The second such experience occurred just today when a member of East Timor Women Australia (ETWA) asked if they could republish my review of the Tais Exhibition and Symposium, as it includes a combination of a capsule history, personal experiences, and a review of both events.

If this wasn't enough, I have received similar feedback in recent days from multiple sources concerning three HPC workshops that I ran the week previous which started from the Linux command-line and ended with profiling and debugging MPI code, which is quite a firehose to drink from in three days if you're starting as a newbie. Over the years I have become accustomed to receiving positive feedback from doctoral and post-doctoral researchers from these workshops. After all, when you have complex problems and big datasets being introduced to supercomputing opens your eyes to computational possibilities. However, when such remarks are made by professors and lecturers with years of experience in the HPC world you can be forgiven for thinking that you've done a fairly good job. Apropos, in a few weeks I will be giving a presentation at eResesearch Australasia with the snappy title "The Spartan HPC Story: From Small Scale Experimental to Top500 and Beyond". This is story of a brave little supercomputer that started on sticky-tape and elastic bands and now is one of the most powerful systems in the world.

The third related item is that I've started work on my doctoral studies in climatology and sustainability. Being just a couple of weeks in it is an orientation period where one is introduced to the Euclid University way of doing things, which includes using their preferred tools. Coming from an open-source world it never ceases to surprise me how products like Microsoft Word (for example) are the default; a lot of this of course is because educators themselves teach how to use a tool rather than the underlying principles of learning such tools and then cycle of ignorance continues. For the shell game that is referencing systems, the University prefers Chicago-style referencing with Zotero for tracking (at least that is open-source). In the future, I may get an exemption to use my preference, IEEE style, notable for being spartan in implementation and free from unconscious biases.

The past week, in addition to my usual work, has consisted of my annual series of five guest lectures for the UniMelb master's level course, "Cluster and Cloud Computing" which covers supercomputing from a high level, the local system, job submission, parallel programming, more than enough Python, and a modicum of Linux commands. The course coordinator and I had some discussions about how the education system is prepared to teach the intricacies of object-orientated programming, but core operating system commands are learned by osmosis. After all this comes the marking of the 380 students enrolled this year. In addition, I have three days of workshops to run next week, with much greater detail and with the addition of regular expressions into the mix.

I have been blessed by a few occasions this week with a visit of my dear Darwin friend, Lara D. Hilariously, we met at the airport at 12.30am on Saturday morning, as her plane from Darwin arrived at the same time as mine did from Townsville. Shortly afterward we found ourselves visiting the Triennial exhibition for what must be the fourth time or so for myself and later in the week it was off to the Da Vinci (and friends) exhibition at Lume, which is another impressive show (although I twitched a bit on a misattributed quote, which is sloppy of them), with VR flyovers, drawing classes, a lovely cafe, and Lume's signature immersive experience of art and music, in this case mainly 19th and 20th century Italian operatic.

On a more culinary and personal aesthetic dimension I hosted a very nice dinner for Liana F., Julie A., and Erica H., during the week which included confit byaldi, the signature dish from the film "Ratatouille", which turned out pretty well for my first attempt at this; it looks fancy, but it's pretty straight-forward). Another event of note was attending Anthony L's wonderful annual gathering at "Life's Too Short", attending with Ruby M. We took the opportunity to visit the NGV at Federation Square beforehand with the wide collection from the Joseph Brown collection. Anthony, as host of his event, noted that it was International Submarine Day and offered plentiful AUKUS-based "yellow submarine" cocktails (rum, tequila, citrus, with more than a dash of imperialism). It dovetailed well with my attendance the day prior at a Labor Against War meeting with former Qld Senator Margaret Reynolds visiting in support.

I am now down to the final week of editing the 15000-word research project for my Master's in Climate Change Science and Policy, following that and a short presentation on the topic, that will be the conclusion of that degree, although I strongly suspect what is the beginning of a lifetime of further study and activity on the subject, starting with the initial subject of my research project; climate change impacts, adaption, and financing for Pacific developing nations. As much as climate change mitigation is absolutely necessary to prevent future impacts, some impacts are and will occur regardless - hence the need for adaptation. For the Pacific, we are talking about nations that have contributed very little to greenhouse gas emissions (both in absolute and per capita terms), but will feel the brunt of its effects; the moral principle of torts should apply.

In my working life, I have two impending HPC-related trips interstate pending in the next two weeks. The first is a visit to Sydney for Supercomputing Asia, where I have a presentation to give as part of the international HPC Certification Forum and a poster on HPC training metrics and usage. Apparently, if you train people, they submit many more jobs - who knew? Two weeks after that, and with an interesting dovetail with my studies, I'll be spending a week in Townsville conducting training workshops at the Australian Institute for Marine Science; it's been almost ten years since I've visited said people and it's pleasing to see that some of the staff are still there!

Even with all this on the agenda, I have had some opportunities for social occasions. Liana F., and I caught some the Triennial exhibition that's on at the NGV, which makes me want to go again, and for Invasion Day we went to Connection at Lume; our third visit to said exhibition. On Wednesday night I received an impromptu ticket to see the Pet Shop Boys' movie "Dreamworld" with Robbie K, which I observed featured quite a demographic of men of a certain age and greater than zero on the Kinsey Scale, and on Tuesday night caught up Nathan B., who is visiting from interstate with a lovely collection of old and true friends at Naked for Satan. For someone who is finishing a master's project I've had zero evenings this week when I've been in my own company; possibly not the most sensible decision, but I'm rather glad that I did manage all this.

Last weekend I had the pleasure of spending some time in the home of Carla with her child and friends; we played the boardgame "Root" with its various anthropomorphic animals seeking to control a woodland. It's quite a clever design, with each faction having its own style and victory conditions (I played the cantankerous Eyrie). The youngsters are very keen on the game and explained their own modifications and alternative factions; I couldn't quite shake the feeling that it was rather like being part of the "Stranger Things" D&D group - and Root does have its own RPG, which I am keen to try. Appropriately, the following Monday, spent time at the Red Triangle for the regular fortnightly game with far more aged players (we haven't changed over the decades) of Burning Wheel where - as an extreme mashup - our setting is the Thirty Years' War, the story is following the classic "The Enemy Within" campaign from Warhammer FRPG, and with scenarios from Lamentations from the Flame Princess.

Betwixt these events I had a linner event with Rob and Angela, with Erica turning up to graze for dinner. Mention must be made of Rob's Rookwood Distellery whose Yuzu gin was awarded a Silver in the Australian Gin Awards last week. Of course, we sampled said gin, which comes up with quite a subtle and beautiful blue sheen when mixed with tonic. Later we moved on to drinking tequila out of skull glasses, which Rob (he has taste this man) cleverly paired with gouda and cummin. In the course of the discussion Angela and I decided we are going to write a psychology paper together, with her professional doctorate clearly taking the position of lead author, and my mere GradDip and wordsmithing helping the process (working title: "Catatonia: a misunderstood and missed diagnosis"). We both have a few publications under our name and we're making some initial progress on that project as well.

The fun and frivolity of the weekend had to come to a close of course and now the nose is to the grindstone of the ongoing preparations for the operating system and software upgrade on the Spartan supercomputer. Whilst applications are constantly being updated, we've managed to avoid doing a major version upgrade to the OS since the system was first initiated way back in 2015 (our system version of gcc is 4.8.5, the environment module version is 11.3.0, for example). The change is necessary (RHEL 7.x will be unsupported soon), but it is going to come with some pain as users will have to reinstall and recompile many of their own applications, and some will just not work. But with the exception of building containers for old software to live in, there's not much that we can do - and the timeline is going to be incredibly difficult with the amount of work we have to do.

The past two days have been taken up with delivering training workshops and the two days previous a hefty proportion of the time spent generating new content and revision for said courses, along with marking assignments for the university's Cluster and Cloud computing unit. From the latter there were a few assignments that made me think "What on earth are they doing?" and a few where the students quite clearly understood what was going on; which I would hope for a master's level course. As for the training workshops, the Parallel Processing was in need of a decent revision and I had new content for GNU Parallel in particular. Alas, this meant that the content that I had for the debugging and profiling part of the workshop was cut shorter than what I would have liked. Then, for the new course, HPC Databases, I delivered pretty good content on the file system, data management and resources, and especially embedded databases with job submission, but when it came to the rather major component of database servers I'd left out a line in the configuration in the midst of a live demonstration - so nothing worked after that.

In the midst of this, the kitchen sink suffered a blockage. I tried the usual methods - plunger, hot water, baking soda and vinegar, wire, caustic soda - before resigning myself to some home maintenance. I'm not afraid of such things having lived in houses with questionable plumbing in the past (ahh, those Accelerated House days), so working step-wise from the drain basket I removed the piping piece-by-piece to discover that there was a fine collection of immovable gunk in the trap that was causing the problem. It's a wet and messy business, so I put it all back together and thought the problem was solved. Alas, my enthusiasm for pipe removal meant that the twenty-year-old plumber's putty had given out and now I have leaking pipes, so a trip to collect some silicone as a replacement is in order.

I've taken both events in rather good humour. It all rather reminded me of one of my earliest forays into this profession (IT in a general sense) when I worked as a volunteer for the Ministry of Foreign Affairs in Timor-Leste twenty years ago. That had more than a few challenging issues, to say the least, given the lack of basic infrastructure or even a semi-reliable electricity supply. In the early months I my time there I had the opportunity to spend some time with one Gabriel Accascina who was responsible for setting up the satellite link so that the (refounded) new country have at least some sort of link to the wider Internet. His previous background included developing a sewerage system in Bangladesh. "It's all shit through pipes", I quipped. But it really is; the data (like the spice) must flow. A lesson plan that doesn't quite work out, a blocked sink. All of these fade into the category of "first-world problems" in comparison to getting just basic infrastructure in place.

Most people use their computers (which includes mobile phones) for communication, social media, games, entertainment, office applications, and the like. Most of the time these activities are not particularly onerous in terms of computing as such or do not lead to enormous benefits in productivity, inventions, and discovery. There is one field, however, rarely discussed, that does do this - and that is supercomputing. It is through supercomputing that we are witnessing the most important technological advances of our day, including astronomy, weather and climate forecasting, materials science and engineering, molecular modeling, genomics, neurology, geoscience, and finance - all with numerous success stories.

Usually, I draw a distinction between supercomputing and high-performance computing. Specifically, a supercomputer is any computer system that has exceptional computational power at a particular point in time, many (but not all) of which are measured in the bi-annual Top500 list. Once upon a time dominated by monolithic mainframes supercomputers, in a contemporary sense, are a subset of high-performance computing, which is typically arranged as a cluster of commodity-grade servers with a high-speed interconnect and message-passing software that allows the entire unit to be treated as a whole. One can even put together a "supercomputer" from Raspberry Pi systems, as the University of Southhampton illustrates.

How important is this? For many years now we've known that there is a strong association between research output and access to such systems. Macroeconomic analysis shows that for every dollar invested in supercomputing, there is a return of forty-four dollars in profits or cost-savings. Both these metrics are almost certainly going to increase in time; datasets and problem complexity are growing at a rate greater than the computational performance of personal systems. More researchers need access to supercomputers.

However, researchers do require training to use such systems. The environment, the interface, the use of schedulers on a shared system, the location of data, is all something that needs to be learned. This is a big part of my life; in the last week, I spent three days teaching researchers from the basic of using a supercomputer system to scripting jobs, to using Australia's most powerful system Gadi at NCI, along with contributions at a board meeting of the international HPC Certification Forum. It is often a challenging vocation, but I feel confident that it is making a real difference to our shared lives. For that, I am very grateful.

Weekend social activities were pretty good, visiting a good friend in their new home as their first official guest (of course I brought champagne and chocolate-coated strawberries, what sort of person do you think I am?), and receiving the ever-loyal Erica H., for some episodes of "Archer". But a personal highlight was going to The Astor Theatre on Sunday evening to watch Godard's À bout de souffle (aka "Breathless"). Unable to rustle up fellow film-lovers, I went by myself to see this in true critic style and dressed up for the occasion as one should when attending The Astor. I love "Breathless", it's one of my favourite films of all time; jump-cuts, semi-improvised, a tragic narrative, comedic, heavily referencing other films, and of course, Jean Seberg still leaves me weak at the knees. But most importantly, it is shortly after Jean-Luc Godard's euthanasia, and this will be his lasting contribution to film art.

It's a curious phrase "À bout de souffle" ("to the end of breath", literally translated which is very appropriate how the film ends) and as life imitates art, I have a fair bit of breathlessness myself at the moment. This week I've been running three supercomputing workshops, Introduction to Linux and HPC, Advanced Linux and Shell Scripting for HPC, and High Performance and Parallel Python, and the latter really isn't an oxymoron but does require work. Unfortunately, I have also come down with quite a cold with a bit of a headache, low energy, and a phlegmy chest and cough. Running three days of workshops in such a state is quite a test of my own dedication and endurance, fortunately, both of which I have in spades. Also, with due diligence, I have taken a RAT and it tells me I'm corona-19 free. I will certainly be spending Thursday on sick leave, however.

Also leaving me utterly breathless in another manner was the recent discovery that a distant friend is not receiving the appropriate medical treatment for a chronic condition. The condition is the sort that will, on average, shave some 25 years off their lifespan. The treatment, known to be pretty successful, isn't horrendously expensive, but they just don't have two red cents to rub together. I just cannot help but think why their well-paid beau apparently hasn't come to the party. Maybe it's just me but I would have thought their life would be worth saving. Anyway, I am thinking out loud, and from a perspective of some ignorance. Maybe there are other circumstances at play, and I'm not going to inquire (or elaborate) any further.

Following my release from isolation (for the good of the realm), I managed one night out for a Darwin pub meal with Lara's awesome neighbours and was kindly driven to the airport. I managed, as I usually do, to sleep through the flight and arrived at 5.30 am to a balmy 2 degrees, a far cry from Darwin's daily range of 15 to 30, and the spectacular dawns. My plants are alive, at least as much as they were when I left them. But the absence of my dear Sabre cat is both haunting and sad. It is now here, in the environment where I expect to see her, that her presence is most missed: "Nothing to be done".

The week in isolation was, nevertheless, somewhat productive. I was aware that my brain wasn't really in the right place to play around extensively on a production supercomputer, but I was certainly capable of working on getting our coursework up to scratch for the HPC Certification Forum by making sure that all the items on their examinable objectives are actually included. This is one of the main items for my work objectives this year, so getting that done will be a welcome completion of a task.

Apart from that, however, I made excellent progress on my current Psych course, "Changes across the Lifespan" and managed to finish all the readings for that subject; the fact that it is officially week 3 of the 12-week course simply means that I have plenty of time to fine-tune the assignments and prepare for the final exam. On a more practical level, I confess that I am concerned about the state of two of my friends who are showing signs of not being in a good place. It's at the point where I make myself available, to listen, to understand, to help (and it must be in that order). All this study on understanding how minds work has to come with a practical aspect, right?

The week ahead looks quite busy; I have about 15 assignments to mark for Cluster and Cloud computing (COMP90024), fortunately with drafts already done. On Wednesday, Thursday, and Friday my days are taken up with HPC workshops. Apropos this, I have recently posted an articles on Microprocessor Trends for HPC in 2022-23. I actually wrote it a couple of months ago and have curated it for public consumption. In the realm of learning rather than teaching, I've reached the point of being so far ahead in my GradDip of Psych that I'm not actually allowed to submit any new content to discussion boards, which had me amused. Fortunately, a friend has pointed me (and as a reminder) to the "How to ADHD", a very practical and educational resource for those who care about such things. It is extremely well-presented, informative, sometimes funny, sometimes saddening, but always very informative. On a related topic, on Thursday evening I'm speaking at SoFiA at the North Carlton library on "From Stoicism and Natural Pantheism to Effective Altruism".

With other matters to comment upon it's been a while since I've reviewed what has been another sociable week. For example, Justin A and family took me out to see on Thursday to see the Columbian psychodrama, Memoria, which won the Cannes Jury Prize last year. As an exploration of a descent into madness it does very well with occasional forays into magical realism, although I think it jumped the shark at the end. I've been spending a bit of time in the company of Erica H. as well, with a particular highlight for both of us being spotting a rakali on a coffee visit to the Docklands harbourside, but even more so a visit to the Imaginaria Now immersive light and sound experience, also at Docklands. Finally, late last week Penelope A., hosted a birthday dinner where I also caught up with Liana F., Annie S., and met Danielle, an AsPro with interests in transnational policy (and yes, I've already started plotting).