Jamie Matthews

Jamie Matthews

@mtth.org

Web engineer. Python, Django. Sustainability, clean tech, decarbonisation. Photography, fuzz pedals and dynamic languages. Co-Founder, Technical Director.

80 followers67 following48 posts

View full thread
Jamie Matthews
@mtth.org

- Keep views thin: treat GET as "read some data" and POST as "do an action". - Design endpoints around real frontend use cases (Backend For Frontend) instead of idealised resources. - Control queries carefully to avoid performance traps.

Jamie Matthews
@mtth.org

It's a free, opinionated guide to structuring real‑world Django codebases so they stay understandable and malleable as they grow. It’s not a new framework, and it doesn’t try to hide Django. Instead, it leans into Django’s strengths and focuses on a few simple ideas:

Jamie Matthews
@mtth.org

- Keep models thin: data definitions, not “god objects”. - Organise code by responsibility (readers, actions, interfaces, data) rather than by app. - Put most business logic in plain functions, not in fat models or deep class hierarchies.

Jamie Matthews reposted

Greener Charge
@greenercharge.uk

Read more about the motivation and methodology behind Greener Charge here: dabapps.com/insights/gre... P.S. We’re officially introducing Greener Charge at the “Charge & Drive EV Experience Day” tomorrow (3rd April) from 10am-4pm on Brighton Seafront. Come say hi!

dabapps.comGreener Charge: Under the Hood - DabAppsRecently, we launched Greener Charge - a simple tool that gives EV owners information about when to charge their vehicles to minimise carbon emissions. But what powers this tool, and how do we know it...
Jamie Matthews
@mtth.org

The site can be added to the homescreen on your phone so it behaves like an app, which means it only takes a second to check just before you plug in.

Jamie Matthews
@mtth.org

We use carbon intensity forecasts from National Energy System Operator to tell you at a glance whether you should plug in your EV tonight, or if it’d be better to wait until the grid is greener. We think this approach could reduce the emissions from charging by up to 25%!

Jamie Matthews reposted

Guy Shrubsole
@guyshrubsole.bsky.social

A "potential" 0.43% increase in GDP *by 2050* - in return for torching the UK's climate leadership and adding vast amounts of carbon to the atmosphere - is simply not worth it. No third runway. Grow the green economy, not airports. www.theguardian.com/politics/liv...

theguardian.comRachel Reeves confirms backing for third runway at Heathrow – UK politics liveThe chancellor says she wants plan for runway brought forward by summer and criticises Conservatives for not acting to build runway
Jamie Matthews
@mtth.org

What I’ve been calling “carbon leverage”: “if AI helps bring forward the electrification of heating, transport and industry by a single year, that would more than offset any negative climate impact from its own relatively limited power demand.”

Michael Liebreich@mliebreich.bsky.social

What happens when techno-optimism meets energy system realism? My latest for @BloombergNEF is a long read on AI, called The Power and the Glory. So that's Boxing Day sorted! about.bnef.com/blog/liebrei...

about.bnef.comLiebreich: Generative AI – The Power and the Glory | BloombergNEFThis year will go down in history as the year the energy sector woke up to AI. This is also the year AI woke up to energy. Is the data center power frenzy just the latest of a long line of energy sect...
Jamie Matthews
@mtth.org

39.3M GPU hours at 700 watts is 27.51 million kWh. 11390 tonnes would be about 410g/kWh, which sounds about right for the US grid. Right now, according to grid.iamkate.com, UK wind farms are generating 21GW (it’s windy), so would generate 28MWh in less than 5 seconds! (I hope my maths is right…)

Simon Willison@simonwillison.net

Added this provocative note to my Llama 3.3 post concerning the CO2 impact of training the model, which Meta calculated as 11,390 tons CO2eq "I believe that's equivalent to around 20 fully loaded passenger flights from New York to London (at ~550 tons per flight)." simonwillison.net/2024/Dec/6/l...

simonwillison.netMeta AI release Llama 3.3This new [Llama-3.3-70B-Instruct model](https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct) from Meta AI makes some bold claims: > This model delivers similar performance to Llama 3.1 405B with ...
Jamie Matthews
@mtth.org

Narrator: “but his maths was not right.” The first bit is correct. But the second bit should be 28GWh, not MWh. So 83 minutes of UK very-windy-day wind output, not 5 seconds! That’ll teach me to do maths in bed. cc @simonwillison.net

Jamie Matthews
@mtth.org

Solar results are in for a dull and grey November… 110kWh total, average 3.7kWh per day. 58% of last year and 54% of the PVGIS estimate. Rubbish! Still, looks like we’re going to pay less than £50 to run the house (including the heating) AND the car once export credit from the summer is taken off.

Jamie Matthews
@mtth.org

I chatted to @chrisbaraniuk.com about thermal cameras after he tracked down one of my old tweets about an uninsulated bit of ceiling. The article is currently on the BBC News homepage. Fame at last!

Chris Baraniuk@chrisbaraniuk.com

I have something else to share with you this morning! A story I've been working on for a while - on the growing popularity of thermal cameras with homeowners desperate to cut their heat loss. BUT the data can be misleading!!! Click to read about common pitfalls. www.bbc.co.uk/news/article...

bbc.co.ukCould a thermal camera reveal your hidden heat loss?Thermal imaging tech has become cheaper and homeowners are using it to track down heat loss.
Jamie Matthews
@mtth.org

Arctic air slowly arriving in the south of England. Seven degree temperature drop between 5am and noon today (purple line). Heat pump now chugging along continuously, pulling about 500 watts from the battery (blue line) and pushing nearly 2000 watts of heat into the house (yellow line).

Jamie Matthews
@mtth.org

(ignore the dip at the end of the yellow line, that’s just because the Vaillant API can only return heat data at one-hour resolution).

Jamie Matthews
@mtth.org

Mulling the idea of “carbon leverage”. The ability for an object or idea or process to disproportionally reduce (or, for that matter, increase) emissions far beyond the amount of carbon “invested” to create it. Example: software that controls a heat pump.

Jamie Matthews
@mtth.org

Since last posting on Bluesky, we got a heat pump. So far this month it’s cost about 1.7 pence per kWh of heat, or nearly 4 times cheaper than gas, and run entirely off our home battery or solar. Stop burning stuff!

You are up to date.