Contact details on LinkedIn or mail@stevesimmons.com.
Previously did military research; was a management consultant; ran marketing, profitability, credit risk, and operations for credit card companies; built trading systems and ML platforms for investment banks.
BE/BSc in Math/Stats/Physics/ElecEng; PhD in Elec Eng, MBA.
World champion programmer, a long time ago (ICPC, 1992), together with Andrew and Craig. Python programmer since 2000. Find my PyData and PyCon UK talks at PyVideo.org.
- You can also use the =LET(...) formula to define named variables:
There must be an odd number 2D + 1 of arguments. The first 2D are D name-expression pairs and the final one is the expression whose value is returned.=LET( filterCriteria, "Fred", filteredRange, FILTER(A2:D8,A2:A8=filterCriteria), IF(ISBLANK(filteredRange),"-", filteredRange) )The end result - as you see - is quite readable.
- Am I going mad or do some of those old.reddit comments slope downhill?
- > The Outlook is Superficially Stable, defined here as “By outward appearances stable unless, you know, things happen. Then we’ll downgrade after the shit hits the fan.”
- Really only if you ask for your data to be deleted too
- Why do you think the current government would be the slightest bit interested in solutions to housing, inflation or healthcare if Epstein wasn't an issue?
- A nice Easter egg in the Gemini 3 docs [1]:
[1] https://ai.google.dev/gemini-api/docs/gemini-3?thinking=high...If you are transferring a conversation trace from another model, ... to bypass strict validation in these specific scenarios, populate the field with this specific dummy string: "thoughtSignature": "context_engineering_is_the_way_to_go" - In contrast to Microsoft, who puts Copilot buttons everywhere and succeeds only in annoying their customers.
- It's worth noting these notes are 11 years old. The first give-away was the comment that in Python 3/2 is an integer, which is indeed true in Python 2 but not in Py3.
For modern users of Z3, you'd want to do `pip install z3-solver` rather than use `Z3Py` mentioned at the very bottom of this doc.
- > gas sets the price in the merit order so we don’t want it on 24/7
I never quite understood the logic for this. Sure, if you overlay a simple upward sloping cost curve on a downward sloping demand-price curve, the market-clearing price is where they intersect, and that in practice much of the time is a gas generator.
But there must be a million other aspects that can affect what price needs to be paid to secure the capacity below that point. Surely only part of the total area under that market-clearing price needs to accrue to the generators?
And if generators are getting windfall profits, can't the market rules be adjusted so more of it can given to the consumers in the form of lower energy prices?
Can someone explain this? Maybe that is what actually happens, just it is too complex for the mass media.
- WSL gives a pretty great Linux dev experience under Windows.
As for "a large portion of the software industry is iOS native apps"... How about plugging in some assumptions here and then multiplying them together:
1. Mobile app share of total software development. 30%
2. iOS share versus Android: 30%
3. What % of iOS app dev is native apps? 40%
My assumptions here give something like 4%. You should put whatever numbers you feel are right here, and I'm pretty sure it won't be close to "a large portion".
- Surprised their ads don't show customers getting pissed about a huge, distracting Copilot icon right in the middle of their Word document, with no ability to turn it off...
- For Python, you can read the PEP documents - Python Enhancement Proposals - and see the discussion of what was suggested, pros and cons, work done to determine the preferred implementation, and the final decision.
To get the best historical sense, pick a language feature that has evolved in several steps over a number of years, e.g. async/await, type annotations, the GIL, etc.
For ordered dicts, for instance, which is topical because of [1] posted two weeks ago, the relevant PEPs are:
- PEP372: The original collections.ordereddict proposal in Python 3.1 [2]
- PEP468: Making kwargs ordered in Python 3.6 [3], which also made standard dicts ordered.
[1] https://www.hackerneue.com/item?id=45756058
- And paradoxically, the more he tries to influence British politics, the more people become aware of his efforts, and think more negatively of him.
Plus who in UK would buy a Tesla now? Other than because it was discounted.
- I am in the UK.
I got Microsoft's emails, did not want Microsoft's forced imposition of Copilot in my Office subscription (regardless of price), found the classic option mentioned in online forums, and managed to switch to it just before my renewal.
My 89 year old aunt on the other hand got stung for the unwanted forced upgrade. I had to call Microsoft, complained about them unfairly exploiting vulnerable customers, and eventually got a downgrade and the difference refunded.
What really annoys me about this - quite apart from the initial deception/misrepresentation - is I now expect Microsoft to pull similar tricks in future. A real disincentive to sign up to any other 'value-added' services.
Why make subscriptions so full of traps that consumers end up hating you? (Yes, I know, so some GM can hit this quarter's bonus)
That reminds me, having just cancelled Spotify (due to their price rise), Disney+ is next on the list. Maybe Netflix too.
- As someone who up to now has mainly watched Django from the sidelines, as it were, this looks great.
What's are your plans for supporting Django v6? (I appreciate it's just gone alpha now, and planned prod release is not for two months.)
Likewise for Python 3.14.
- This is the kind of comment that makes HN special and precious. Thank you.
- If you want to add signatures, the best option is to use a PDF viewer that supports that directly (FoxIt Reader, for one). That saves the original PDF plus an annotation for the signature.
Second best is print using "Save as PDF". That saves the PDF as processed through the browser's PDF engine (PDFium, in the case of Chrome or Edge).
Worst choice is usually print using "Microsoft Print To PDF". That saves the PDF processed into a more generic way. Often the original text characters are replaced by drawings, and the final file size can be 10x what you expect.
- Same for me...
I had been tracking Datastar for months, waiting for the 1.0.0 release.
But my enthusiasm for Datastar has now evaporated. I've been bitten by the open-source-but-not-really bait and switch too many times before.
- If in Python, have you looked at dictdiffer or JYCM?
If a compressor can compress every input of length N bits into fewer than N bits, then at least 2 of the 2^N possible inputs have the same output. Thus there cannot exist a universal compressor.
Modify as desired for fractional bits. The essential argument is the same.