Category Archives: technology

Data junkie: a new Garmin or swim shiny, part 2

There was this.

Yesterday I went to one of the adventure sports shops and bought a Garmin VivoActive HR mostly because in the price sector, it covered most of the swimming stuff and cost less than I plan to spend on a Rado. It is now affixed to my wrist and at some point this evening, I will take it for a “run” otherwise know as a “rulk” where I go out “running” but am so run unfit, I walk most of the run. This will be the way things are for a while.

I didn’t get much sleep on Friday night which, I think, is why I went and bought both a sports watch and a toaster. The Vivoactive is my second sports watch – there is a Vivosmart somewhere in storage which was used for step counting. I hated it because like a lot of sports watches it was dog ugly and in particular, it whinged at me if I didn’t move enough which was basically Monday to Friday. I spent a lot of time running up and down stairs in my then job once an hour so that I stopped feeling guilty about the vibration alerts. I don’t want to be wearing this all the time because it isn’t really about counting my steps – my phone does that anyway – but I need some serious help on the swim data front. Mostly I estimate how long I have been in a pool, and I count the lengths and hope I don’t lose count somewhere along the way. Mostly I do but the figures are “ball park” right, I have no idea exactly how long a length is taking me, and frankly, the measure that matters most to me – the one which has always mattered most to me –  how long I rest between lengths is never anything more than a vague feeling of progress.

I wanted certain functionality – namely blue tooth synching to my phone – and it would be nice if it handled running as well. You can spend hundreds on a watch to do this. You can also spend a lot less. But you get less for your money. I wanted – at least – a SWOLF measure so that I could at least chart changes in how the swimming was going and from what I remember, the cheapest you could do this was with a Garmin Swim which didn’t sync to a phone. So next step up. If I were wealthy, the one I wanted was a Polar V800 – the vivoactive came second to it on several occasions, and the times the Garmin 1, the Polar had not been included in the group test. Price wise, however, the Polar was way out of my league. You could not call me a serious sports person at all. I just don’t have the time.

If I am truly honest, looking around the pool this morning, I wondered if I wasn’t a bit pathetic as well. Historically I swam 1600m back stroke, it sounded good, and it took around an hour. But it failed to meet a bunch of random goals I had over years in terms of feeling safer in the water. The Irish life guards require candidates to be able to swim 400m front crawl in 8 minutes. I knew when I last had a go at that goal that I was nowhere near it. I looked at swim watches then too; ten years ago, but they were rather less able to do much other than act as a stop watch than today’s lot are. I’m going to be honest and say I think swim watches still have a way to go – I feel running and cycling are much better served by wearable tech. But we are where we are .

The watch

So, the watch has a rectangular face which is about 3cm by 2cm. This marks it out a bit from a lot of other sports watches which have relatively round faces. The one which I bought has a black strap although I think you can get a white strap as well.

In my opinion, the watch is ugly to look at. Sure, it’s sleek, and it has a nice reasonably sized touch screen. But I’m female and I wear a Tissot women’s watch every day. For looks, this, and to be fair, almost no sports watch, competes – the prettiest is probably the Withings or Nokia as they will soon been known. The Withings do cover swimming but do not pick up SWOLF iirc and the ones which were available here didn’t pop up on group tests either. In answer to the question “Would I wear this every day?”, it’s a straight no. The watch is big and clunky. I’m not the slightest of females by any manner or means but I don’t find it sleek, stylish or elegant.

By default out of the box it has a digital display and to be frank, its brightness level was sufficiently low to make it practically unreadable. You can change the watch display to analogue which I did (eventually) but that doesn’t really change my opinion of how it compares on looks to a normal analogue automatic watch which I wear every day.

There are two buttons on the watch under the watch face. One is basically a back button, and the other, amongst other things, accesses activities for you to start tracking. The screen is a touch screen and you can scroll down through it to see things like the weather today, the number of steps you’ve stepped, your last sports workout, a summary of your daily activity and your heartrate. The watch has a heartrate monitor installed which does not work when swimming.

In terms of charging, the watch comes with a charging band which you can plug into a USB port on your computer. Personal view on that is it compares badly to the one on the Vivosmart – I struggle to get the watch out of it every time I plug it in.

Storing the data

The watch syncs its data to a Garmin Connect account. I already had one of these from when I bought the Vivosmart.

It does not currently, as far as I know, talk to Google Fit (Android girl, sorry), and to get it to talk to Runkeeper, you need to route it via Tapiriiki, which seems to be one of the standards.

According to swim.com, you can automatically sync data from your Garmin Connect account to swim.com. I have attempted to set it up and frankly, my faith is somewhat lessened when swim.com tells you you need to install a plugin which neither Chrome nor Edge will install. Although apparently it should install in Internet Explorer, it doesn’t for me, not on Windows 10. Looking at a comment about compatibility on the Garmin page for the plugin, I guess it’s years old. I am hoping it is not still necessary.

There is a comment on the Reddit swim page that it can take a few days for the connect to work which it may or may not because I eventually attempted to connect it without the plug in at all. I will update accordingly. And I will annoy Reddit Swim with questions about how to make it work because swim.com support doesn’t look like they have much feedback on that front.

What does work, however, and it works very nicely in my view, is SportTracks.mobi. SportTracks is a €€ site but there is at least 45 day free trial and currently, the annual charge is $59. I’m not sure how I came across this one but unlike the others, it has an effective automatic sync.

At the pool

I’ve currently set up the watch to be worn on my left wrist – this will probably change if I start wearing it running – because I think that information is required for the heartrate monitor. Which doesn’t matter if you are swimming as it doesn’t work while swimming.

Couple of things to note – it will track walking if you pool walk for light resistance training. But for some reason, the screen responds to water running across it when walking so you often wind up on some random screen from the front lot of apps. More than once it whinged at me that it couldn’t control any music player because the mobile device it was paired with was nowhere to hand.

That’s because it was in the dressing room in a locker. I found that vaguely annoying; on the other hand it gave me credit for 800m of water walking which I find hard to believe but I’ll take. I don’t do it often enough.

Setting up a swim is easy enough: you press the right hand button and select “Pool Swim”. This locks the screen on the pool display which is handy.

This display includes time, interval distance, total time and total distance.

To start it, you press the right hand button again. After that, if you want to pause it, you press the left hand button. I do this between intervals and it works handy enough. This gives you time per interval and elapsed time on an ongoing basis which is handy if you are under time pressure in the pool and have an idea of what you are targeting.

To restart after a rest break between intervals, you press the left hand button again.

Getting the hang of the buttons is not that hard – although if you accidentally stop an interval with the right button instead of the left button, you might as well save the session and start the next one. It might cost you a length in terms of total distance covered (at least it did for me). Going back and saying “no I want to continue this session” does not seem to be an option – it was either save or discard. 100m being worth not losing, I saved and started a new.

Statistically, I do not know how accurate the Garmin is in general although it did credit me with 50m for one length and I am not entirely sure why. It was the only error in a 24 length total, however.

Unexpectedly though, it failed to identify intervals where I was backstroking rather than freestyling. This surprised me and may be a reflection my backstroke style. As a result, my jury is out on the stroke recognition side of things. It got freestyle right but frankly, it’d get that right in 75% of cases if it just assumed everyone was freestyling.

After the swim

The sync took a little longer after the swim than I expected but I think this may have been because internet connectivity in the pool building was negligible and the connection back to Garminconnect just wasn’t that fast.

You get a reasonable summary on your mobile device. I probably would order the data differently in the summary – the key items I’d like to see up top are Total time and Moving Time. You can customise a few things in the Garmin interface so at some point, I might be able to feed that through.

You can look at the swim interval by interval as well – and it is here that I noticed that one length had been measured as 50m for some reason. Which did a lot for my average and my SWOLF at that point in time.

Linking it with other applications

I’ve had a Runkeeper account for years but I can’t see an easy way to link the Garmin swim data to it other than using a third party service which I find annoying. In theory, it is possible to set up an import for a GPS map but this isn’t very much good for pool swimming. Additionally in theory it is possible to set it up via the Runkeeper mobile device but that’s not happening for me either yet, mainly because I had to change my password to enable it on my desktop and now my mobile won’t log in even with the new password. 

So I haven’t set that up. Discussions online suggest links into Google Fit are also a no-no. However, it does automatically sync with SportTracker and allegedly swim.com. I can’t comment on the latter as it isn’t working for me [yet]. However, the former works nicely both on desktop and on web and the key metrics are easier to read on SportTracker than on GarminConnect in my view.

However you can export a .fit file from Garmin Connect via your desktop to swim.com manually if you want and to be fair to them, the display is fairly user friendly. Swim.com is useful for other reasons beyond the swimming tracker so even if you couldn’t get the automatic sync working, it might be worth considering doing a manual data dump every once in a while.

Data can be dumped out to a CSV from GarminConnect as well

In general

The watch pretty much does what I need it to do at a reasonable price point (ca €200 in Europe). It isn’t particularly pretty and in my view, is probably designed for men by default. The jury is currently out on stroke recognition but data upload is okay; the Garmin Connect application contains the data for a quick consult and review. It certainly is a useful tool in terms of identifying data that I only managed to collect approximately in the past and it provides data that I really never managed to collect manually before. From that point of view, for someone like me, an occasional, unfit, and overweight swimmer with a few goals to be sorted out, it’s a decent enough piece of kit. Jury is currently out on the interconnectivity with 3rd party sites as well but at least one works properly, there is a useful fudge for a second and the third, well check back later on that.

From the point of view of adjusting goals, I can track the rest time (and specifically the reduction of same), and length time splits (want to shorten that). In terms of number of strokes per length. I’m not terribly upset about that.

Added after the fact

Swim.com synchronisation has started although the phone app is claiming there were issues with the data upload. Looks okay in the desktop web app so far.

On the techbro and Google

At some point when I was doing my Masters, someone leaned on me to apply for an internship with Google. Going to be honest and say I thought it was ambitious for some practical reasons – they were looking for people to work in SRE and I have zero interest in that any more, and I was interested in data analytics and machine learning. They interviewed me; it went badly, they said no. I wasn’t surprised because deep down I knew they weren’t offering internships in an area which interested me in the location they were interviewing for, and what they did want, I really wasn’t. I think I got leaned on to apply by the Googler who leaned on me because I was female. But I don’t know for sure and never well. Either way, we weren’t a fit at that point and the way my life has gone since, we probably continue to diverge in interest. I use some of their products and form part of their product and that’s about it.

They were in a no-win situation in the news lately though following a 10 page essay, piece, whatever you like to call it in which a person who will never need to benefit from it cribbed a bit about the Google diversity programmes and trainings. Specifically, he seemed to argue that there was somewhat of a biological reason for which women didn’t seem to make it in companies like Google. It got me thinking, not about the fact that Google didn’t want me, but what it was like to be in tech, listening to these arguments going on. Sure there is a chronic lack of women in tech compared to men, but against that, there seems to be – in the US in particular – a high quotient of assholes working there – in the technology sector, I mean. I am not targeting Google with that comment specifically. I’m not sure the two points are unrelated.

I started working in the IT sector when I was 27 years old. IT is not a vocation, you don’t have to be born to it, and you can learn it later in in life if you want to. But it is not a vocation. This piece seems to me to elude a lot of people talking about tech and in particular, it comes to the fore when people are talking about the pipeline.

You can learn to do pretty much any job in IT if you want.

IME you don’t even have to be a maths genius. I programmed in assembly language for over 10 years. I was sysprog for a mainframe. When I was 13 years old I learned Atari BASIC. I’ve taught myself Python to some extent and have coded in R, Java and I’ve done a year of SQL dev as well.

But despite the Atari BASIC at the age of 13, I wasn’t a tech nerd all the way through school. I could solve problems (I copped how to duplicate computer games on cassettes by realising a twin deck tape recorder could do it when every one around me was messing around with computer tape decks and trying to save stuff from RAM having swapped cassettes). Mostly as a teenager though, for me, the world was music. I played three musical instruments, I sang choir, and I did dancing lessons.

I did not hide in a room doing computery stuff. When I was 14, I decided I wanted to live in France and figured knowing French would be handy. From then on, my primary motivation was foreign languages. I did my first degree and first postgrad in languages. I’m a qualified translator interpreter.

In my view given a choice between learning to program and learning to speak a foreign language the amount of work in acquiring the foreign language far outstrips any effort involved in getting functional as a programmer. But for the most part, computer programming and related tasks pays more.

When I was doing my machine learning masters, I was sitting in class one day when a fourth year undergrad told me he wanted some advice. He was about 20 years younger than me and he was thinking of doing the same Masters as I was. There was a significant amount of overlap between the modules and I basically told him he’d be wasting his time but fine if that was what he wanted. The conversation moved to the fact that as far as he was concerned, women didn’t like working hard. Did I not notice how few girls were in the classes?

I’m female and I am 20 years older than this guy plus I’ve just told him I worked in system support as a programmer for more than 10 years. Sure there are few girls in the class, although being honest, it’s not as bad as I expected, half the classes were foreign intake – in my year, primarily China – and all told, I didn’t think he was in the place for generalisations.

“Girls” he told me, “did not like work. They really only came to university to find husbands, princesses all of them”.

There were, at the time, plenty of women in the science faculty, not just doing other bits of compsci, but doing assorted natural sciences stuff and biotechnology. TBH, I’ve never done a full engineering degree or done a chemistry degree, but what I do remember is that the people who did – many of which were female – worked harder than frankly, I’d seen any of the compsci students do. Including this guy, who was, after all, looking for a way to avoid getting into the work force by redoing quite  abit of his fourth year and then writing a dissertation to get a masters. I thought his attitude stank.

The biggest work soak of my Masters actually wasn’t a compsci module. I did two stats modules with the maths people – time series analysis and multivariate analysis. Those courses were harder work than anything else I did in UCD. Those classes were about 50:50 women/men. I don’t think it was the lack of interest in work which was the problem, somehow.

The thing is, I worked with a bunch of IT heads in a specialised business, and to be honest, for the most part, my experience is that the men I worked with then were not bastards. But quite a few of them were like me and did not come from a compsci background. I know at least one person had a history degree, one was a languages dude like me although with a different set of languages, and at least one had a background in psychology or sociology, One started off life as an electrician, and I know a couple who went in with school leaving certs as their qualification. Trained by various employers. There was a time you could do that.

When I was in school, we did not have a computer module – it was the mid 1980s when home computers were really only getting there. I had a maths teacher who…was interested in application of maths and a couple of times, he took us off syllabus and spent a week doing some basic macro economics, and he spent a week teaching us BASIC as well without the benefit of anything so much as to run code on. I tended to have a go off it when I could get my brother off the computer.

I went to an all girls school and by the time I got to the age of 16, a key objective of my maths teacher was to retain as many girls in the higher level maths class as he could. He argued massively with any girl who made known a wish to go back to what was called Pass Maths (the syllabus calls it Ordinary Level Maths) from his honours class because they felt they were not up to it. In his opinion, anyone who had made it through the higher level course at junior level was capable of the higher level course at senior level. It was very often a losing battle for him. When we eventually finished up, I think the class had 7 people in it. Out of a cohort of over 60. The school itself, while having some good maths teaching did not necessarily push young women in that direction – my class had a cohort of secretaries and legal secretaries and a few nurses. The higher level physics class was even smaller than the higher level maths class, although for some reason, the higher level chemistry class was bigger. It may have been considered useful for nursing, along with biology.

One thing the maths teacher discussed with us one day, probably after another one of us had stepped back into the ordinary level class. He told us that in general, the boys’ school had bigger higher level maths classes but even so, there were times when the girls – what few of them he got to teach – did better than the best of the boys.

This is probably linked to the benefit of an extremely low teacher pupil ratio, plus the fact that he scheduled extra classes for us every chance he got. On average, in Ireland at the moment, more boys than girls take higher level maths although to be honest, the difference isn’t massive – it is in or around the 10-15% fewer girls mark. This is seriously different to the split on higher level languages where girls far outstrip boys for most languages bar Irish and English. The statistics can be found here.

What does all this have to do with Google Boy and his essay about innate differences? Well women tend to do well in most STEM subjects even the ones who go into computer science, the few of them. So he can write all the essays he likes but..frankly, his points are probably only relevant at the extreme end of the spectrum of the need for brilliance – a location occupied by Grace Hopper and Margaret Hamilton amongst others. In the early days of computer science, most software was written by women because it was considered beneath the men who were involved in hardware engineering. The position of women in software engineering has changed over time which suggests it’s less the question of innate skill and more a cultural matter.

The other thing is this: the technology industry at the moment is not really in a great place. It has massive security problems, an issue which is going to get worse and worse as we bring more devices online via the internet of things (for example), and as cyberwarfare becomes more of a front than it already is. One of the things I despise more than anything in software engineering is the whole release early and often ethos. You don’t get away with this in certain areas and yet it probably has contributed to a significant amount of the risk run by individuals and companies with respect to their technology assets. The culture is also such that while there are a large number of systems around running old code, the simple fact is that technical debt is more of an issue for more recent systems than it is for the mainframe so called dinosaurs with inbuilt backwards compatibility and a rather different ethos about release early and often (are you stark raving made?).

There is a desperate lack of women in technology. I wonder really how much of it is being culturally driven in a different direction and how much of it is realising that you’d have to work with a bunch of guys like the guys in Uber as a useful example. There are a lot of very bright women around the place. The technology industry is probably missing out. Instead of trying to make excuses for why it is missing out, and in addition to its diversity programmes and efforts to advertise itself to women, I think a period of self reflection in terms of the kind of organisation that seems to float to high profileness in the tech sector might be in order. If one thing is obvious from this debacle, it is that at no point do any of those busily suggesting that it’s alright that there aren’t that many women, or minorities in general in their sector consider that maybe their behaviour contributes to it.

Pool review: Bonnevoie, Luxembourg

I’m currently without a real home pool at the moment which makes the building of a swimming habit somewhat difficult. My would be home, dCoque, which is Luxembourg’s National Sports Centre, is currently closed for renovations until 1 October which is a nice chunk of the year. I’ll be honest and say I have not really managed to get a swim habit. No swimming till October is  pushing it a bit though. The next replacement, which I haven’t managed to check out yet, Badanstalt, about a 10 minute walk from home, is also closed. The pool 5 minutes from home isn’t open to the public. And so on and so forth

Next on the list is Bonnevoie. Bonnevoie is not too far from the railway station in Luxembourg, and public transport wise it is served by buses 3/30. You need to get off at Leon XIII. It is open every day except Wednesdays, and weekdays, it tends to close at 20.30. On Sunday it is open from 8.00 to 12.00.

There are two pools in the swimming centre in Bonnevoie; what they call the large pool, and the small pool. The large pool is 25m long, and ranges in depth from 1.8m to 3.8m at the deep end. There is a shelf at around 1.5m around the edge under water. The steps are set into the pool wall. Temperature wise it is reasonable – it is cool enough to swim comfortably in, but not so cold that you fight against going into it. When I was there, 2 lanes were separated off for lane swimming. The rest of the pool was occupied by people swimming lengths anyway so basically the choice was yours.

The small pool is about 10 by 10 meters, and at its deepest is 1.25m. One side consists of steps rather than some ladders, and at one end, there is a barrier to hold onto when getting into the pool. There is a small slide as well. I understand the pool is also used for things like aquaerobics. Temperature wise it is around the same as the big pool.

Entry to the pool without discounts for an adult is currently 3.80E and the ticket is in the form of a chipped band which I wore on my wrist and which also serves as a key for the lockers. In my view, there are not a lot of dressing rooms, but they work okay. The lockers are tall and narrow, and while you did not have to fumble for a euro or a coin of some description to lock them. I did not like them. I struggled to get my clothes bag into them and eventually pulled stuff out. My personal preference would have been for wider but shorter lockers and they could have increased the number of lockers available had they done this.

From a layout point of view, the lockers are a distance from the showers which means that really, it’s wise to have a bag to carry towels and shampoo or whatever you want in the showers afterwards. I tend not to like this because it’s just another thing to remember. That being said, one item of design in Bonnevoie recognises the needs for that, and as a result, there are shelves poolside to store stuff, like small swimming bags with towels for example. I liked that touch because typically, in other places, to find somewhere to leave a bottle of water and flipflops or on other occasions, everything can be a hassle.

The other thing which the pool in Bonnevoie has which is not a common feature in pools in my experience is a fine big readable digital clock. This makes it handy to plan around bus services and dates and the like.

Given that currently, dCoque and Badanstalt are closed, it is likely that I will default to Bonnevoie as the alternatives are a) rarely open (Bel Air) or b) expensive (Les Thermes). I’d be happier if it were open for 30 minutes longer during the week and maybe a few hours more on Sunday but then the world does not revolve around me.

Bonnevoie Swimming pool
30 rue Sigismond

Bus 3/30, Place Leon XIII

Monday: 08.00-20.30
Tuesday: 06.45-20.30
Wednesday: CLOSED – although apparently open 8-17.30 while Badanstalt is closed until 1 September – I cannot check this myself.
Thursday: 06.45-20.30
Friday: 08.00-20.30
Saturday: 08.00-20.30
Sunday: 08.00-12.00

Cost of a swim: €3.40

Website: Ville de Luxembourg

Treasa’s Map of Pools

 

 

Things every house should own

I was reading a fascinating article on Ars Technica the other day – mostly to be honest, I read these things because I am intrigued to see what the comments will be like – if I am honest. Anyway, the article concerned the hacking of an emergency notification system in Dallas Fort Worth. Here’s a link to it.

As it happens, the comments were the usual mix of rank ignorant and rank knowledgeable and a scale of non-rankness in between. One of the key questions from the “we recognise we are ignorant” is why the area still used this kind of alert system – basically sirens given that the technology existed for things like cell phone warnings and this is where it got interesting.

Dallas Fort Worth is in Tornado Alley and so the alert system actually gets used, and often at short notice. And the reason it continues to be used is that a) it works and b) it is reliable. There is already a lot of experience to demonstrate that in emergency situations, cell networks are not reliable (although landlines are tending to stand up to disaster a bit better). Well, this system is reliable as long as no one tries to hack it but that’s a wider moral discussion. The point is, cellphones are often switched off, on silent, in the room downstairs, or whatever other excuse you are having yourself. In the context of the onset of a tornado that’s never going to be enough.

A couple of things struck me about the conversation – one was the usual tendency of people who live somewhere with gigabit fibre to assume that everyone has gigabit fibre, and similarly, the tendency  of people to assume modern might always be better than old tech. Sometimes and often times it isn’t. Anyway, another point which popped up almost in passing in the comments was identifying other sources of information in the event of an emergency. A siren is one thing to tell you hunker down, avoid, escape to high or low ground, or whichever is appropriate for your situation (CF earthquake drills in Japan as another example). Someone mentioned that they would always have a battery operated radio in the house just in case power was taken out.

This caused me to pause and think. I have a torch in the house at all times, and it has a place where it lives. I’ve always felt that you should always have a functional torch in the house and that’s why there were 4 in the last house. But if I lost electricity, I have maybe mobile phone for up to 24 hours if it was fully charged when I lost electricity and provided I switch off cellular data. I live in a temperate area and normally, you’d expect that I might be out of electricity for a relatively short time.

It’s just lots of people have found out the hard way that this might not be a completely safe assumption.

So I added a battery operated radio to the shopping list and found a small Sony for 25E in a local electrical shop. I need to get some batteries for it because of course it’s a different size to the set of batteries I have here already.

You can wander around Youtube and find all sorts of survival kits, squeezed into tiny Altoids boxes if you want to be really creative. But in the context of living in a city with no desire to go off-grid, as it were, possibly your shopping list should include a torch, a radio, spare batteries for both and a few litres of water at the very least. And at that, I’d like to think we’d be able to see the biggest problems coming from a long way off.

Leaving it to the engineers

It is possible that this belongs on my other site but…Fine.

I think this was caused by a tweet linked to a comment made at the SCIC21 interpreter training conference during the week. It is an annual conference involving the interpreting service of the European Commission and the universities which run courses that supply the interpreter pipeline. The comment, which I now cannot find, grosso modo, amounted to highlighting the risk in allowing the engineers to be in charge of the future of the language industry. In other words, Google.

EDA: and the tweet is here from Marcin Feder:

C. Tiayon – do not leave languages in the hands of engineers, a reference to Google, etc.

(quick thanks to Alexander Drechsel who found it for me)

I have a background in artificial intelligence and machine learning and as part of getting it, I did a machine learning course where the lecturer confidently asserted that translation as a problem had now been solved by applying statistical methods. This basically means, loads of data and learning from that. There are well known issues with this; it tends to be good, as in better than what went before, but anyone who has seen both Google Translate and Bing Translate in action would really be admitting that it’s still not great. In many cases it’s terrible. Twitter’s language recognition engine is a bit hit or miss too which makes its translations hilariously absent. One of the things which annoyed me in general when I did my CompSci masters was the extraordinary tolerance computer scientists have for faults in things which are not necessarily computer science. Woe betide you if you don’t comply with someone’s pet programming style peeve (usually in the area of variable names) but a 60% success rate in whatever the code is trying to do, like, translate something from French to English is tolerable. The computer end of business they are shockingly fussy about, the business end of business, less so. Human interpreters and human translators are still far, far better than computers in terms of transferring meaning from one language to another. This is because meaning is not all verbal and computers are not good at nuance.

So we keep hearing how great computers are at something or other – lately it has been GO – or diagnosing some illness or other. I have doubts about the last one because often that’s a question of judgement rather than straight binary… – Anyway we keep hearing how great computers are at some task but when you drill down, it is because monumental assumptions and allowances have been made. You can read, for example, that Duolingo is better than college courses for learning languages and that this is scientifically proven. The relevant study had a 75% drop out rate. That means 75% of people who started learning a language to in a programme to measure the effectiveness of Duolingo dropped out before the end of the study.

This is not just true for languages but in general, if computer science or technology is getting involved in your industry, it is worth paying careful attention to what they consider to be adequate performance. It may well  be significantly less than is considered adequate in general in your industry and you will want to know the rationale for that.

Amazon Driven by Data

Possibly this belongs on my datablog and I might write it up there later (or I might not).

The New York Times has published a piece on working conditions in Amazon. It does not make for comfortable reading and certainly does not paint Amazon as a company which I would want to work for. But I did want to pick up on one comment in it.

“Amazon is driven by data,” said Ms. Pearce, who now runs her own Seattle software company, which is well stocked with ex-Amazonians.

That comment about driven by data is something that worries me. Sure, data is very sexy at the moment in the tech but data in itself can be meaningless and what matters is the information you can derive from it. But the information you look for is generally skewed by humans who decide what questions they want to ask of that data. Data isn’t the answer and how it is queried is not benign or completely rationally independent.

The average tenure in Amazon, apparently, is one year. You can argue that this might be as a result of a toxic culture. You can argue that Jeff Bezos is a genius. But to my uncertain knowledge, Amazon does not turn a profit and I think, has never turned a profit. We have an ongoing assumption that this is somehow okay, that it’s a new paradigm, and times have changed. With an average tenure of a year, you have a company which cannot possibly have stability in output quality, and you have a company whose knowledge base never develops.

Amazon’s search is atrocious and its recommender system has deteriorated badly in my experience. A hypothesis for why that might be is that they appear to have rapid staff turnover and by definition, limited continuity.

Many of the datapoints in Amazon’s evaluation system appear not to be datapoints at all. They are entries in the AnyTime Feedback Tool:

Ms. Willet’s co-workers strafed her through the Anytime Feedback Tool, the widget in the company directory that allows employees to send praise or criticism about colleagues to management.

However, many workers called it a river of intrigue and scheming. They described making quiet pacts with colleagues to bury the same person at once, or to praise one another lavishly.

Anyone who implements a tool like this either a) expects it to get gamed and considers that a value or b) doesn’t expect it to get gamed in a high octane organisation is naive at best and making decisions on potentially faulty data. Either way, they are unlikely to wind up with the best staff. Such organisations, however, can only operate on the comfort blanket of never admitting this.

Amazon can call itself a data driven organisation, and if it wants to measure everything down to the nth degree, they should be bright enough to know the limitations of what they are doing. Going by the content of the NYT’s article, they probably aren’t.

Automated newsfeeds

One of the things I liked about Facebook over Google News (and in practical terms, there are few enough of those) is that I could set the newsfeed up to deliver me stuff from a lot of different language sources. Google assumes monolinguality; it’s constantly offering me tips for searching for English results only and for someone who is multilingual and interested in what’s happening in the communities of her other languages kens, it’s a bit frustrating.

However, as an advantage, it really is waning and the reason for that is the famous Algorithm. Now, in simple terms, algorithm is just just a method for achieving some result. The result which Facebook allegedly wants is for me to have the most relevant material turning up in my feed as that enhances engagement, and engagement is a handy asset for getting money out of advertisers.

Whatever way it works, it’s serving me a lot of drivel I don’t want, particularly on the news front, to the extent that shortly, I’m going to unlike pretty much all the news sources. I’m sick of them. I get pages of Royal Weddings, pages of Bruce Jenner becoming Caitlyn Jenner, pages of entertainment rubbish that I really don’t care about, don’t want to see, and can’t switch off without switching off news altogether. Note to twitter: do not mess with the time driven method by which you serve me content.

Facebook is a time sink. Today, it’s been highly negative. There has been a preponderance of news stories that annoy me, and drivel that, if I’m truly honest, I don’t want to read, but still wind up reading. Much of it is repetitive with the same stories coming from several different sources, and the occasional marginal different angle. A lot of what turns up serves to make me feel inadequate too as there’s the stuff that tells you what you need to do to have a good career, what vegetables you should eat to get healthy sleep, the ten, fifteen or a million habits of Steve Jobs.

I loved computers and technology but I realised last night that possibly the two highest profile tech companies in the world, namely Facebook and Google, make money from advertising. You can spend all the money you like developing a better newsfeed selector but it’s not going to make the world much better. The tech industry often solves the wrong problems.

In the meantime, I’ve been thinking about the things that made me happy about the internet 20 odd years ago when my life got Netscape Navigator. It was the pretty things. The things that made me happy. The things that opened the world to me, and opened the possibilities.

It wasn’t a constant feed of news that I couldn’t directly influence myself. I haven’t worked out what to do to make FB give me politics not celebrity rubbish from the Guardian. And so, the news orgs are going to have to go and just maybe, FB will no longer become a pit of stuff that irritates me.

Neolithic monuments in Ireland

20150415_142156
Newgrange is one of the highest profile historic sites which we have in the country and when most people talk about going to Newgrange, they mean they want to see this one.

When I go to Newgrange, I always go to Knowth as well. Yes, you can actually go into the passage in Newgrange, and yes, it’s extremely well done but it’s always very busy.

Knowth is generally much quieter and, on occasion, no matter how busy Newgrange might be, you might have the site at Knowth more or less to yourself. There’s a lot to be said for this.

Knowth is bigger than Newgrange, but it does not look anywhere near as perfect. It hasn’t been restored (or reconstructed) in the same way as Newgrange was, and some different decisions have been made about the site. A key one is the question of the quartz stone. At Newgrange, this was built up as a wall. At Knowth, the view was taken that it was probably a terrace around the entrances. I’ve mixed feelings. Certainly Newgrange looks more complete but….

That aside, the reason I would still favour Knowth over Newgrange is the art. Knowth has significantly more external art than Newgrange and it is stunning.

Yes, the entry stone for Newgrange is iconic:

20150415_142355

but then, there’s this:

20150415_155545

and this:

20150415_160718I find what’s around the base of Knowth simply to be on a scale which is borderline unimaginable at Newgrange.

I didn’t have time to go to Dowth yet and it’s not included amongst the options you can get to from Bru na Boinne. However, if you are interested in neolithic art in Ireland in that area, I would strongly recommend Knowth as a seriously underrated site. It is wonderful. You can actually look down the passageway although access down it is not permitted to the public and you can see some public access work done on the eastern end. You can also walk to the top of it and the view from it is quite impressive.

I find the whole idea of pre-history in Ireland fascinating. If you go to the National Museum on Kildare Street, you’ll find examples of 3 and 4 thousand year old jewellery which contains carvings not dissimilar to some of the carvings on these stones and it’s extraordinarily beautiful. I really do wonder about the societies that were able to access the gold, shape it, carve it. It seems to me those societies, however on a smaller scale than is currently on the case, must have been extremely sophisticated, particularly with respect to their ability to use tools to achieve tasks which would probably challenge us today.

 

 

 

Random new info

http://www.era.europa.eu/Core-Activities/ERTMS/Pages/home.aspx

I occasionally edit pages on Wikipedia which involves a certain amount of research. Today I was updating a page on a rail accident and linked to research I did with that, I spent some time on the page of the European Rail Agency. The link above captured my interest.

Today, there are more than twenty signalling and speed control systems operating at the same time in Europe. The Thalys train sets, in particular, linking Paris and Brussels, have to be equipped with seven different signalling and speed control systems and radio communication variants.

I used to travel on the Thalys from Brussels to Paris the odd time, some years ago.

Seven. Seven. Seven different signalling and speed control systems and radio communication variants.

I know Thalys now travels between France, Belgium, Germany and the Netherlands but even so, you’d have to wonder how in that framework, they manage to need seven different signalling systems.

Deutsches Museum

On an island in the middle of the River Isar in Munich is one of the greatest museums in the world. I can say that advisedly. The Deutsches Museum on Museum Island is overwhelming.

It is one of the earliest museums of science and technology in the world, and, I am told, if you were to walk every exhibit, you would walk more than 17km.

In truth, what happens is you walk into the first section, which is full of boats and model boats, you get knocked backwards, and never really recover. They have a terrific aviation section. They have an amazing aeronautical section. They have a mindblowing collection of clocks and weights and balances. They have every sort of textile weaving system. Every sort of printing press that you can imagine. Every sort of ceramic you can imagine. A terrific model railway. A terrific collection of keyboard instruments.

They have holograms.

Most importantly, they have 2 Enigma and one Lorenz cipher machine, plus a bunch of other cipher machines. And an IBM 360 with a punch card reader. Every sort of adding machine and calculus machine or analogue calculator that you can imagine.

The entry fee is eight euro fifty. It is worth every cent and you will come out a complete wreck having not seen everything.