A couple of years ago I wanted a new running/GPS watch, and eventually settled on a Coros Pace 2. This was very much a "bang for my buck" purchase in that it certainly doesn't have as many features as the top of the line Garmin watches - but was also less than half the price.

Probably the most notable feature that it was missing was any kind of route tracking and/or turn-by-turn navigation. That was fine, I knew what I got when I bought it. But then in May this year, Coros added breadcrumb based navigation, and they're currently going through beta testing of turn-by-turn navigation. I filled in the trivial form Coros sent me to be added to the beta test and a day or so later got added to the beta test; one quick app and firmware update later and I at least nominally had the feature enabled.

The obvious thing to do was therefore to go for a run. Results were pretty good - the turn-by-turn got a little confused at a four-way junction but other than that worked absolutely fine. Maybe I'm just cynical these days but I am pleasantly surprised that a company would release what is a major new feature like this without trying to monetise it. Hey, they might even get a return customer if/when I decide I need my next watch.

Unless you've been living under a rock, you're probably aware that the UK had a General Election yesterday. One question that has come up is "what would have happened had there been a Labour-Lib Dem alliance"? TL;DR: the Tories would still have won, but with a reduced majority.

My (pretty simple) method:

• Assume that all the remain parties got together and just put up one candidate in each seat. For this purpose, the remain parties are Labour, the Liberal Democrats, the SNP, Plaid Cymru and the Greens.
• I then think it's a reasonable assumption that the Tories and the Brexit Party would have come to a more formal alliance given the alliance against them, so assume that the Tories, the Brexit Party and the (pretty much irrelevant) UKIP just put up one candidate in each seat.
• In each constituency, we can now reduce it to three parties: "Leave", "Remain" and "Other". Run the simple analysis to find which constituencies would have had a different result in this scenario than actually happened.
• Yes, this assumes everybody who voted for one of the parties in the alliance would have voted for the alliance. Close enough for government work (pun intentional).
• One other scenario can now occur: if Leave would have won a constituency, but (Remain + Other) would have got more votes, we at the very least need to do more analysis to see what would have happened as we don't know how those "other" voters would have voted. Similarly if you swap Leave and Remain.
• This analysis ends up ignoring Northern Ireland. This is because Northern Ireland is complicated (understatement there...)

Given all this, there are 42 constituencies that would have ended up "Remain" but actually returned a "Leave" MP:

• Aberconwy


• Altrincham and Sale West


• Berwickshire, Roxburgh and Selkirk


• Bolton North East


• Bridgend


• Bury North


• Carshalton and Wallington


• Cheadle


• Cheltenham


• Chingford and Woodford Green


• Chipping Barnet


• Cities Of London and Westminster


• Clwyd South


• Delyn


• Dumfries and Galloway


• Dumfriesshire, Clydesdale and Tweeddale


• Filton and Bradley Stoke


• Finchley and Golders Green


• Gedling


• Hazel Grove


• Hendon


• High Peak


• Hitchin and Harpenden


• Kensington


• Lewes


• Milton Keynes North


• Moray


• Reading West


• Rushcliffe


• South Cambridgeshire
(where I spent a good number of years)

• Southport


• Sutton and Cheam


• Truro and Falmouth


• Warrington South


• Watford


• West Aberdeenshire and Kincardine


• Wimbledon


• Winchester


• Wokingham
(where I grew up!)

• Wrexham


• Wycombe


• Ynys Mon

and 12 constituencies which would have flipped to "Leave":

• Barnsley Central


• Barnsley East


• Dagenham and Rainham


• Doncaster Central


• Doncaster North


• Hartlepool


• Houghton and Sunderland South


• Kingston upon Hull East


• Kingston upon Hull West and Hessle


• Normanton, Pontefract and Castleford


• Stockton North


• Wentworth and Dearne

On top of that, there are 9 constituencies which might have flipped to "Remain" (but "Other" votes could have changed the result - see above):

• Bury South


• Derby North


• Dewsbury
(my constituency!)

• Esher and Walton


• Guildford


• Pudsey


• St Ives


• Stroud


• York Outer

and two constituencies which might have flipped to "Leave":

• Hemsworth


• Rotherham

So, given all that, we can say that "Remain" would have gained between 42 and 51 seats, while "Leave" would have gained between 12 and 14. It's unlikely that all the "maybe" Remain seats would have gone but none of the "maybe" Leave seats, so let's take the two limits together. The baseline results are 365 seats for Leave (all Tory) and 267 for Remain (203 Labour, 48 SNP, 11 Lib Dem, 4 Plaid Cymru and 1 Green). The "missing" 18 seats are the Northern Irish ones. If we assume none of the "maybe" seats change, we get 365 + 12 - 42 = 335 for Leave and 267 + 42 - 12 = 297 for Remain, still a reasonably comfortable majority as the effective number needed for a majority is 320 due to the 7 Sinn Fein MPs not taking their seats and the speaker and three deputies not voting. If all the "maybe" seats did flip, we'd have 365 + 14 - 51 = 328 and 267 + 51 - 14 = 304 for Remain which is getting a bit tight but still at least nominally a majority.

There you go, make of that what you will. As usual, code to generate these numbers is on my GitHub. All comments and polite criticism welcome.

As some of you will know, I recently moved house (contact me by other means if you want my new address). I didn't actually realise at the time of buying the house that it had (the potential for) FTTP¹, but for the first time that opened up the possibility of dropping the old copper phone line and replacing it with VoIP. Why? Because paying £15/month or more for a phone line which we use very little is a waste of money, so at least investigating a VoIP solution seemed like a good idea².

In terms of an ISP, I've been very happy with Zen for almost 15 years now, so there seemed little point changing there. They're probably not the cheapest in the market, but they're not far off and they're technically very competent. By signing up for an FTTP contract with them, they also gave me a FRITZ!Box 7530 for "free" which seems to be a reasonable piece of kit and a definite upgrade on my aged Netgear box.

Once the (very helpful) Openreach engineer had done his job and got things to the state where there was connectivity to the fibre modem, first task was to set up the FRITZ!Box. This was nice and easy - plug the Ethernet cable from the fibre modem into the LAN1 port on the router (this was about the only non-obvious thing). Once you've done that, connect to the FRITZ!Box's wireless (SSID and password on the bottom of the router) and then browse to http://fritz.box/ (note HTTP not HTTPS) and log in with the admin password (also on the bottom of the router). Then browse to Internet / Account Information, and put in your Zen login details. You should now have Internet. Hooray! (I now spent a bit of time reconfiguring the wireless network, but none of that was critical or hard).

The choice of VoIP provider was somewhat trickier as I've never had one before, and almost all the advertising out there seems to be targeted at businesses rather than consumers (for understandable reasons, most consumers still need a bundled phone line one way or another). After some browsing, I found Port 5060 who both looked reasonably competent and didn't require a long-term contract. Order was not entirely intuitive - while their pricing page says £3/month, clicking on the "Order Now" button presents you with a £5 one-time fee. It is £3/month for their basic "one number" service, but this wasn't entirely obvious to me at this stage. In any case, I ordered and the account was activated about 12 hours later.

You can then log into Port 5060's technical portal with a different username/password from the billing side of things, and a different-but-very-similar URL. If any of the Port 5060 people happen to read this, single sign-on and everything on one domain would be nice please :-). The next steps were not particularly obvious:

1. Click on "Manage Numbers" then "Purchase Number". You can now select a number from any UK geographical code or a range of non-geographical ones. I was boring and went for the geographical code where my house actually is. This step costs £1/month for those keeping track.


2. Now go to "Manage Extensions" and "Add Extension". Create the extension, associate it with the number you just chose and get yet another username and password (which is fair enough in this case, it's effectively a user on your account). When you've created the extension, click on the username in the table (UX? What UX? If you have to put a big hint on your page telling people where to find things, you probably need to change your UX...) and note the SIP proxy.


3. Now go back to the FRITZ!Box and go to "Telephony" / "Telephone Numbers", "New telephone number". Set "Telephone number for registration" to the full number from Port 5060, "Internal telephone number" to the local part of the number (not sure how crucial this is), the username and password to those you got in Step 2 above, and both the registrar and proxy server to the SIP proxy you got above. I haven't set the STUN server - I suspect I don't need this because I have a static IP address.


4. You can now plug your regular analog phone into the "FON1" port on the FRITZ!Box (they helpfully supply a BT connector to RJ11 adapter with the router) and should be able to make and receive calls "as normal".

So in summary:

• Openreach were very good at getting the fibre installed.


• Zen remain excellent.


• The FRITZ!Box seems pretty good so far - everything Just Works(tm).


• Port 5060 seem technically competent but the UX leaves a lot to be desired.

Notes:

1. FTTP is Fibre (Internet) to the Premise or in generic terms, fast broadband. However, if you didn't know that already this post is likely to be a bit geeky for you.


2. We did consider the idea of not having a "landline" at all and just using our mobiles, but as our new house has solid stone walls approximately the thickness of those in a nuclear bunker, mobile reception in large parts of the house is dodgy at best.

Don't use the same password on more than one site. That won't be news to the more security conscious of the people that read this (if it is news to you, read the link) but even for my more techie friends, the practical management of secure passwords can be an issue. I occasionally get asked how I do this, so here's a blog post...

For avoidance of doubt, this is very much just how I do things. It's by no means the one and only one correct way to do it, but it works for me. The other slightly "odd" (although maybe less odd among people that read this) requirement that I have is that it needs to provide a solution for when my laptop is running Linux. Anyway, what I do is:

• Create a password database with KeePass. Use something really strong as your master passphrase.
• Use KeePass to generate every new password you ever create. Use secure settings for this - my default settings are 32 characters and all the character classes KeePass allows except "High ANSI characters".
  
  • Occasionally this breaks a website which doesn't like that kind of password. This is annoying, especially when they don't tell you why they don't like your password. Not much that can be done about that though.
• Put the password database onto my Google Drive.
  
  • On Windows, install Google Backup and Sync to ensure that I always have the latest version of the database.
  • On Linux, install Insync to sync Google Drive to my local machine. Yes, this involves paying money but as a one-off payment I think £30 (or $30 if you're lucky enough to be in the US) is a worthwhile investment.
  • On Android, the Google Drive app gives you sync automatically. (This assumes you're running full blown Android, rather than AOSP as you get on e.g. Kindle Fire tablets. I haven't tried to solve that problem)
• On Android, install Keepass2Android Password Safe to access the database.

This now means that 1) I have the latest version of the password database on all my devices and 2) I'm not reliant on any cloud provider's security for my password security. Even in the worst case that Google were hacked (or subpoenaed) and my password database was obtained by an attacker, the database is still encrypted by my strong passphrase so hopefully useless. I am reliant on KeePass's security, but I'm happy with that. (Also, thank you to the EU for making us safer).

The only gotcha I've found with this system is that I have to be careful not to have the password database open on device A and editing it on device B, or the syncing understandably gets confused. There aren't many valid use cases for having the database open on more than one device at once anyway, so this just encourages good practice :-)

Historical note: until March 2019, I used Dropbox for the sync rather than Google Drive, mostly because Dropbox provided a free Linux client. However, Dropbox now restrict their free accounts to three devices which broke all that. If you pay for Dropbox already for other reasons, you can use that and not pay for Insync.

It is apparently "Green GB Week". The cynical among us may say this is a publicity stunt to distract from the a certain summit happening on Wednesday which seems likely to go horribly wrong, it's as good a time as any to reflect on the state of our world. Some assumptions I'm not interested in debating here:

1. Climate change is happening.
2. Climate change is significant.
3. Climate change is primarily man-made.
4. We have an ethical duty to do something about it.

Given all that, the question on my mind is "how can I, a moderately intelligent, moderately well-off person make the largest positive difference to the environment?" For anyone that doesn't know me, I already cycle to work every day so that takes out one of the obvious changes I could make. I also have two children; I'm well aware this is in many ways a significant increase in "my" environmental cost but it isn't something which is going to change. It also means that I am time-poor. I am prepared to invest potentially non-trivial amounts of money in something if it could significantly affect my environmental cost. Not sure what "non-trivial" means yet.

So... tell me what things I can do which will have the largest benefit to the environment. If they just affect me and my immediate family, that's OK but what would be wonderful would be things which affect a larger group of people. Particular attention will be paid to ideas with quantitative numbers behind them - feel free to refer to the amazing "Sustainable Energy - without the hot air" to get those numbers. Stuff which I've considered:

1. Going (mostly) veggie. I know I should but... I like meat.
2. Voting Green. This is a tricky one - firstly because our electoral system means a green vote is almost certain to be "wasted" (I live in a Con-Lab marginal), and secondly because I have serious problems with the Greens' energy policy. "Do it all with renewables" just doesn't add up (see "without the hot air" for the numbers); as far as I can see, the only way we can in the short term get low carbon energy generation is with nuclear. It's not perfect, but it's the least bad option for base load generation right now.

What else can/should I do?

Didn't get through as much as I'd expect while on holiday for a week in the North York Moors because I got distracted by some coding. Aside: coding without Stack Overflow is hard :-( Anyway, in August I read:

• A Woman of the Iron People (Eleanor Arneson): space explorers find a planet of iron-age people and send down a few people to investigate. Now, I do get this book is to some extent a commentary/investigation on our modern day anthropology, but to me it wasn't either a good story (too many "with one bound, Jack was free" type moments) or actually a particularly deep or insightful commentary on anthropology.
• Gridlinked (Neal Asher): I got on pretty well with Asher's Transformation series (see last June for the end) so grabbed this when I saw it in a charity shop. And... it's not as good as his later stuff. I've never particularly liked military fiction and while the Transformation series did go into that kind of stuff, the rest of it was interesting as well. Here, the military bits are still there, but the plot and the rest of the writing aren't there to support it.
• Extracted (R. R. Haywood): invent a time travel machine, "extract" three people from different times for... something. The "something" is revealed pretty early on, but the book then just drags on... and on.... and on.... through a fairly interminable slog with one of the characters. Chop that out, reduce the book to a third of its size and this would have been a much better novel. (Oh, and the physicist in me insists any time travel novel at least has to try and deal with the paradox issues. Hand-waving it away along the lines of "nah, let's not try that" really doesn't work).

 July started off very slow as a reading month, but picked up at the end while I was away in Pembrokeshire for a week. Anyway, I read:

• The Night Clave (Monte Cook and Shanna Germain): a Numenera novel so while set in an RPG universe, probably never going to be a straight up dungeon crawl - and it certainly isn't, being essentially a voyage of self-discovery as to what really matters for one of the characters. Enjoyable enough, but felt a bit too padded to me - maybe would have been better in a novella type format. As an aside, I dare anyone to watch Strand and not want to play in the Numenera universe afterwards...
• The Abyss Beyond Dreams (Peter F. Hamilton): back in Hamilton's Commonwealth universe, which is never a bad start and again sees people (well, in this case, mostly Nigel Sheldon) battling the Void, although this time from the inside. A lot of this book felt like a rehash of the Void trilogy itself - following the rise of individual through a Void-influenced society, much as Edeard did. Not quite sure where Hamilton's going with this one, but I'm sure I'll read Night without Stars, the second half of the series sometime.
• The Long Mars (Terry Pratchett and Stephen Baxter): while I was really taken with The Long Earth, I wasn't so grabbed by The Long War in which not much really happened. Fortunately, The Long Mars rectifies that with two major plotlines, both of which grab the attention and show how the Long Earth (or as it's not a spoiler given the title of the book, the Long Mars) really starts to shape societies. Still obviously setting some things up for later books, but I'll be back to read them pretty soon.

It's been too long. Haven't read that much in the past six months, sometimes for good reasons, sometimes for bad. But here's some of the things I did read:

• The Medusa Chronicles (Stephen Baxter and Alastair Reynolds): fairly typical Baxter or Reynolds hard sci-fi. Enjoyable enough, but with a GREAT BIG SIGNPOST to the ending a third of the way through the book which left a bit of slogging through waiting to get to that bit.
• Can & Can’tankerous (Harlan Ellison): I don't remember very much about this. Sorry.
• Neverwhere (Neil Gaiman): pretty dark, pretty twisted, generally quite weird. Or in other words, just what you'd expect from Gaiman. Enjoyable.
• Binary Storm (Christopher Hinz): entertaining action romp through a dysfunctional near-future Earth. Written as a prequel, and either the ending is very forced, or was hamstrung by the need to remain consistent with the later books. I suspect the latter.
• Nemesis (Alex Lamb): starts off as a fairly standard "conspiracy in space" - but has enough twists along the way to keep the interest levels up. Probably worth reading Roboteer (the first in the trilogy) before this though as while there's just about enough exposition so that things make sense, I felt I was missing a few bits along the way.
• The Three-Body Problem (Liu Cixin, translated by Ken Liu): Hard sci-fi, but with an Eastern spin this time. Hard going in places and jumped a bit quickly from the "investigative phase" to the "everything happens at once phase" for my liking, but I wouldn't be averse to reading the rest of the trilogy at some point.
• The Book of Heroes (Miyabe Miyuki. Apologies if I've got that name in the wrong order): young girl enters a story to save the world. A bit too young adult for me really, but probably enjoyable if you like that sort of thing.
• My Night in Freeport (Anthony Pryor): Freeport is one of my favourite D&D settings, so a new story set in the City of Adventure? Grabbed me straight away - but it's really, really, really short. Not worth the asking price.
• First Person Peculiar (Mike Resnick): a short story collection. Had some good bits and not so good bits in it. Probably.
• Poseidon's Wake (Alastair Reynolds): yes, more Reynolds so more hard sci-fi. Didn't realise it was the third one in the trilogy when I got it from the library and definitely suffered from that.
• Miniatures (John Scalzi): another short story collection that I don't remember very much about. There's a theme here, isn't there?
• The Nightmare Stacks (Charles Stross): back in the Laundryverse. More programmer/maths geek dark horror fun. Good, but not up to the level of The Atrocity Archives from my point of view.
• Mechanique: A Tale of the Circus Tresaulti (Genevieve Valentine): take a traditional circus. Add some fairly dark Iain M. Banks-style body horror and you have Mechanique. Not going to be a book that makes you smile, but still worth a read.
• Echoes of Earth (Sean Williams and Shane Dix): an off-shoot colony of humanity receives an apparently great boon from another species... but it comes with a price. A lot of set up for what was obviously always intended to be a multi-part story, and doesn't really get going itself.
• Blackcollar (Timothy Zahn): a bunch of freedom fighters strike out against the evil Galactic Empire using non-conventional weaponry. Sound familiar? Actually written before Zahn wrote what is apparently some of the best stuff in the Star Wars EU (sorry, "Legends"), but all the same... military fiction is never really my thing, but this is okay as it goes.

 October's reading:

• Gene Mapper (Taiyo Fujii, translated by Jim Herbert): a bit of a random pick up from the library, but also part of a slightly deliberate attempt to read a bit outside my normal comfort zone of moderately hard sci-fi. Very simple plot summary is near future and dealing with the socio-economic impacts of genetically engineered crops - and I enjoyed it enough to munch through in a couple of days, which is pretty rare these days. Can't let the glaring crypto hole go uncommented though: no, you cannot reverse engineer the input given only a hash, no matter how much brute force computing power you have.
• Skeen's Leap (Jo Clayton): your typical RPG thief ends up stranded on a medieval technology world and has to get back to her nice high tech environment. Along the way, she acquires a party of diverse companions and has various adventures. Unfortunately, the plot is about as generic as it sounds, and the writing isn't good enough to elevate it. It's also only half a story, ending just as it should be getting going. Meh.
• Yesterday's Kin (Nancy Kress): apparently peaceful aliens have landed in New York. What do they want, and are they truly peaceful? Sounds like yet another bit of pulp fiction, but Kress spins it through the eye of one (dysfunctional!) family and how it affects them. Enjoyable enough in what is effectively a novella format, but just lacking that... something which would make it want to pick up the trilogy its become.
• The Stars are Legion (Kameron Hurley): I know this one has got a bunch of adoration for its feminist themes, but... sorry, it's just not a very good story. Just too many points at which the "grand plan" could have got one or other of the protagonists killed, and then the whole thing would have fallen apart without its message meaning very much at all. A real struggle to get through.

Currently reading Gridlinked (back to hard sci-fi and back to the start of Neal Asher after doing the Dark Intelligence trilogy earlier in the year) and The Three-Body Problem (Cixin Liu: continuing the eastern theme), and finished off Word Puppets which I was reading through October.

September's reading:

•  The Atrocity Archives (Charlie Stross): as noted last time out, you'd probably expect I'd have read this already, but I never had. I really enjoyed the first half ("The Atrocity Archive" itself) as I'm always a sucker for Lovecraftian sci-fi, but wasn't so taken by "The Concrete Jungle". Overall the whole thing reads as what Pratchett might have written if he were writing stuff set in the modern day, which isn't a bad thing (and I do very much like Good Omens as well...). Will probably seek out some more of the series sometime.
• Flood (Stephen Baxter): I have a funny relationship with Stephen Baxter. He's an author I always think I should like more than I do, which leads to me owning quite a lot of his work... but never actually enjoying it very much. Flood was OK - an interestingly different take on climate change, but still not one that really grabbed me at all.
• Moons of the Solar System (James A. Hall III): a largely tedious gazetteer of every known moon in the Solar System (i.e. a natural satellite which orbits something other than the Sun). While some of these bodies (Enceladus, Europa, Ganymede) are definitely interesting, some have some cool orbital mechanics (Epimetheus and Janus) and it's vaguely interesting to know that some asteroids have moons, the vast majority of moons are just boring lumps of rock. Then for added confusion, some almost entirely unrelated appendices.

As of 1st October, the only thing on the reading list was Skeen's Leap, but I've actually worked my way through Gene Mapper and started on The Stars are Legion since then.