100 ideas so far

  • RSS reader that prints out a custom newspaper. One or two sheets a day. PDF
  • ?Railway station database with freely usable images
    • hostnames: station.wiki or railstations.wiki
    • make available in JSON-LD
  • hard drive price tracker with alerts: started
  • twitter for transit vehicles
  • build something with the Netflix API - maybe with the affiliates programme
    • Netflix shutting down API
  • good neighbourhood conversations
    • neighbourhood.wiki / hood.wiki / community.wiki
  • wikipedia keywordiness - see https://github.com/robennals/think-link / https://github.com/robennals/think-link/blame/master/wiki/wikigathernames.rb
  • Enterprise content management
  • watch video, use subtitles to show definitions for infrequently used words
  • run face detection across illustrations in IA books scans
  • peer-to-peer money exchange with twitter and facebook friends
    • register for site using social account, link facebook and twitter
    • make offers like I want to exchange GBP for USD
      • suggest that users tweet there offers
    • trade with friends, once trade is agreed transfer money
    • rate transactions
    • users build reputation trading with friends
    • maybe strangers with reputation will be willing to trade
  • search book covers for similarity to wikimedia commons featured pictures
  • new service to replace 'out of office' e-mails
  • annotate a mediawiki article to show which user added each image
  • system for managing Frequent-flyer programs. create accounts, keep track of miles and expiry dates
  • personal dashboard should show upcoming holidays, bin days and shop open times
    • keep track of all pay as you go accounts, for example:
    • Contactless transit cards: London Oyster card, Bay Area Clipper card
    • PAYG Mobile phones
    • Localphone
    • temp sensor: https://tessel.io/ $99 + $15 shipping, includes wifi
    • nice: http://shopify.github.io/dashing/
    • monetize: adverts, sell anonymised data
  • street party database/wiki
    • hostname: streetparty.wiki
  • text interface to OSM geocoder and road details, works from the web and command line
  • crawl The Economist for obituaries and ensure each person has a Wikipedia article
  • web site to send e-mail bin reminders: take postcode and e-mail address, maybe provide API
  • attach bitcoin to e-mail: "i sent you 5 BTC" like how paypal works, charge for service
    • coinbase has done it
  • monthly / weekly / daily bitcoin payments using coinbase API
  • timeshare management and calendar software - initially for noncommercial holiday home shares between friends
  • facebook app to invite friends on holiday. calculates the time and price of their travel
    • hostname: friends.travel
  • corporate photo gallery
    • enterprise.pics
  • book site compares prices between UK and US
    • AcrossThePond.books
  • do something with Land Registry Price paid data
  • proper UK comparison site, tesco direct, sainsbury's and amazon UK. electric toothbrush price varies a lot
  • battery powered devices should report charge levels to a central server
    • mightytext does this, for a single device
    • monetize: collect and sell battery stats, show adds on web page dashboard
  • AR mobile phone app: point at supermarket shelves, calulate best price per kg, share online
  • put something on the web about 'future peaking', maybe a name for a company or blog. No good search results currently.
  • calendar server that solves these problems. hack on radicale, make it faster
  • calculate wikipedia keywordiness and use to build wikipedia text auto linker
    • apply to Hansard text on theyworkforyou.com
    • auto-link as a service
      • use wikipedia keywordiness to add links to text
      • provide wordpress plugin that auto creates tag pages
    • domain name available: keywordiness.com
  • geocode quality review blog
  • gecoder test suite
  • currency arb travel site, see blog post: http://rootofgood.com/smart-travel-geographic-arbitrage/
  • site that highlights session timeout mistakes
  • sesason ticket site
    • hostname: RailSeasonTicket.travel
  • affiliate page with all rail ticket affiliate's compared
  • bitcoin Consumer Price Index - new currency, like bitcoin but stable price
  • app idea: alerts for reduced items at the supermarket
    • people can report supermarket discounts, rewarded with badges.
    • app tracks discount timings, suggest the best time to go for discounts.
    • app users visit the supermarket, take photos of reduced items and the app sends out alerts to other users who might be interested
  • tickets from cambridge to london are confusing, site that makes all the options clear:
    • network rail card / LST vs. KGX / day returns / travelcards / Carnet tickets
    • destination in London, should show travel time via LST and KGX and price difference
    • hostname: cambridgerail.travel
  • database of transit route planners: TFL, transit.511.org
  • web site database of free software companies - crunchbase for open source
  • british recipe site (again)
    • crawl uk recipe sites: bbc, bbcgoodfood, waitrose, delia, channel4, uktv, jamie oliver, deliciousmagazine, riverford
    • as many facets as possible
    • write ingridents to a database, sort by frequency, simple ingredients should appear most frequently, search for simple ingredients within other ingredients to classify
    • try and add facets for different dishes: spagbol, chilli con carne
    • generate seo landing pages for different dishes, but how to sort? relevance isn't a good option, maybe voting buttons
    • parse prep times and cooking times
    • identify multiple part recipes - index ingredients together an separate
    • can try identifying ingredients and cooking methods by lookup in wikipedia, wiktionary or wordnik
    • search for cottage cheese, it should be a phrase search, shouldn't give cottage pie
    • maybe provide an API?
    • save to my binder, uses facebook/google/twitter for accounts
    • post to facebook/twitter
    • star ratings?
    • reviews
    • like goodreads for british recipes
  • CME rule book as web pages
  • bitcoin webhooks, call a webhook when:
    • coins sent from adddress, or each confirmation
    • coins sent to adddress
    • tx is confirmed
    • bitcoin price webhook
  • debian fsck should detect if the network is available then load minimal GUI with browser pointed to http://paste.debian.org, terminal as root and IRC pointed at #debian or #debian-fr
    • /lib/systemd/systemd-fsck
    • source is systemd/src/fsck/fsck.c
    • runs: fsck -a -T -l
    • work out how to trigger fsck. maybe
    • possible to create sample filesystem with dd and mke2fs. then losetup and fsck /dev/loop0
    • next step is to corrupt the filesystem, probably using dd
    • corrupt a filesystem by doing this:
      • dd if=/dev/zero of=fstest.img conv=notrunc seek=100 bs=1 count=1024
    • e2fsprogs contains test suite with samples of corrupt filesystems
  • write web page about corrupt ext2 filesystems for testing
  • pyjamas city index, how much are you allowed to walk around in pyjamas?
  • software as a service: parse excel, return CSV
  • software as a service: parse excel convert to JSON-LD
    • probably impractical
  • learn django, update django book: https://github.com/jacobian/djangobook.com
  • make a database website of books that have source code. start with books on github
  • improve test coverage of web.py
  • Multiple commit boxes for ikiwiki
  • write ebooks about: ISO8583 or DFDL
  • facebook api ebook
  • book reviews and discussion for books not on Amazon:
  • system for identifying politically exposed people (PEPs) using Wikipedia
    • JSON API, post the name, and other identity details, replies with either:
    • true, maybe or false with a link to Wikipedia if true
  • hardware idea: MP3 player, like clip zip with wifi
    • ship with rockbox
    • kickstarter
  • MP3 player OS for old phones
    • maybe android with a different launcher
  • Some companies are very dependent on google SERP rank, trade stocks based on google serp rank, for example RetailMeNot
  • fix dictionary used in Debian to include words like WiFi and JSON
  • open recommendation web site - all media, books, movies, games
  • sell shares on small and obscure stock exchanges using Bitcoin
  • web frontend to 'git grep'
  • fork spyder, add spellchecker using enchant
    • source here: https://bitbucket.org/spyder-ide/spyderlib
    • write ebook about spyder
  • google docs for latex
    • done: http://docs.latexlab.org/docs
  • oauth login for corporations
  • open price diary
    • crawl prices from supermarket websites
    • mobile app to capture prices, works offline
    • offline download of prices to mobile app
    • leaderboard, badges
    • hostnames: pricedairy.wiki - groceryprices.wiki - groceries.wiki
    • crawl supermarket sites, publish prices using JSON-LD
    • objects:
      • country: uk, usa
      • currency: gbp, usd
      • supermarket chain: waitrose, trader joe's
      • specific shop: coldhams lane sainsburys
      • need a hierarchy for describing products
        • tomato ketchup
          • heniz ketchup, different sizes
          • supermarket ketchup, different sizes
      • price capture: specific shop, product and price
  • CPAN for sqlalchemy objects
  • make dbpedia / wikipedia infobox / wikidata available in JSON-LD
  • campaign for people to include '' in HTML
    • turns out HTML5 uses id="foo" instead
  • hack on mailman, add tests, make web nicer web archive interface (pipermail)
    • git mirror isn't live yet
    • maybe make a git mirror
  • freezer inventory app
  • web payments is making a standard for receipts
    • build digital wallet that can hold legacy currencies
  • receipts as JSON-LD are going to be very important
    • web app should show all receipts
    • categorise spending
    • keep track of purchased items:
      • track gifts to calculate spend for each recipient
      • resale value of items purchased for self
      • food inventory tracking
  • ebook: JSON-LD
  • ebook: systemd
  • do something web like with debian-devel-changes:
    • feed into twitter and facebook
    • make nice web pages for updates
    • alerts and webhooks?
  • e-mail to webhook gateway - maybe a bit like mailinator for webhooks
  • social networking for commuters
    • sign up, give your commute and interests
    • maybe comments via facebook
  • web station noticeboard
    • wiki + news + photos + conversations
    • one web page for each station
    • big photo of station
    • sign in twitter/facebook/google. people can post messages
    • include news stories
    • include OSM map?
    • find people to car pool to the station car park
    • links to nearby stations
    • monetize: adverts, affliates, donations
    • include maps, use OpenStreeMap. choices: Mapnik or TileMill
    • render big static map images
  • holiday calendars: google for 'exchange holiday schedule'
  • reverse engineer skyscanner and airbnb apk to extract APIs
    • need to run proxy to find skyscanner app API
  • new file format for screenshots, text and stores vectors instead of bitmaps
    • lets the user select and copy text
    • include multiple translations of texts, so i can screenshot a dialogue box in one langauge, others can view in a different langauge
    • pulsing progress indicators can continue to pulse in the screenshot
    • might need to embed fonts
    • maybe users can view with a different theme
    • could grab a screenshot in linux, view as if on windows or osx
  • python module, it auto-determines what blogging software is running on a given URL
  • file format for blood tests
  • include license information in image files. also web payment details, like details of royalty fee and bitcoin address
  • premium bitcoin ad network, like The Deck http://decknetwork.net/
  • bitcoin jobs board
  • take popular open source database driven web apps and publish the database schema with annotations
  • Android SMS <-> IRC gateway
  • OCR book scan corrections in the browser
  • OSM plus ATOC timetable data gives you train movements map
  • selling mTickets with bitcoin
    • UK rail mTickets infrastructure is produced by masabi
    • setup payment processor
    • need a company name
    • talk to masabi
  • open source paperwork for setting up organizations
    • health and safety policy
    • other policies
  • vacation site or calendar should track how many vacation days you have left
    • suggest holidays aligned to public holidays
  • android app bitcoin payment API
    • bicoin payments on mobile devices, need an API for an app to request bitcoin from a wallet app on the same phone.
    • example: buying rail tickets using a rail ticket app
    • waiting for confirmation screen in rail ticket app?
  • Project to ensure every UK station includes platforms on OSM.
  • Database of speakers with reviews
  • simple team whereabouts tool:
  • bitcoin + https://www.easypost.com/ shipping API
  • Adobe reader has annotation with form filling and signing. We should have the same thing on Linux.
  • Database for recording serial numbers of stuff you own
    • mark an item as lost or stolen
    • other people can use the database to check if an item is lost or stolen
  • mobile app for proximity group voice chat (party line / conference call)
    • CB radio for the 21st century
    • handsfree, used in-car. let you talk to six nearest people
    • reports location to server
    • figure out which road you're on
    • avoid brief periods where conferences merge at intersections
    • start with six nearest people
    • as it gets more popular have six nearest people on the same road
    • people pre-record a greeting for when they join a conference
      • convention is that people include a car description in their greeting
    • beep and play greeting each time a person joins the conference
    • drivers enter number plates
      • display shows number plates of people in the conference
      • highlight number plate of person speaking
    • should people have callsigns?
  • browser extension, spell checks the sites you visit, highlights mistakes
    • money from donations?
  • UK bitcoin payment processor, accepts bitcoins, send faster payment
  • bug tracking with bitcoin.
    • pay to fill bugs, refund if good bug
    • get rewards for good reports
    • pay for bugs to be fixed
  • deb file viewer instead of file-roller when you double click on a .deb file
  • UK rail network in graphviz format
  • startup job scraper - use AngelList API to find startups, scrape their sites to find job pages.
  • show me TED talks by authors who's books i've read, using goodreads.
  • distributed goodreads
  • OCR lower-thirds in TV news - good project for archive.org
  • car satnav/route planner that suggests places to stop.
    • https://news.ycombinator.com/item?id=8419658
    • hard to monetize - micropayments? android app
  • uk transport strike calendar (tube/rail/bus/airport) calendar
  • machine readable strike data - past and future strikes
  • blog post or campaign website to convince americans they need more vacation time
    • highlight vacation news stories, like unlimited hoidays at netflix and virgin
    • guides for negotiating more vacation
  • combine http://twotap.com API with bitcoin
    • account created
    • API documentation: https://twotap.com/docs/
  • PDF viewer for magazines, should show two page spread, easy to zoom in and out
    • sample PDF magazine: http://www.gsb.stanford.edu/sites/gsb/files/stanford-business-magazine-2014-spring.pdf
  • Web browser save web page should include security certificate
  • database of: delis, coffee shops, bakers and butchers that are on twitter
  • fuse or git frontend to online grocer shopping, edit order by editing files. able to diff orders
  • tweets have a tidy looking page with a nice URL, Wikipedia edits should have the same thing
  • write econoday python module - would be popular for shops using trading systems in Python
  • search engine that ingests epub files and indexes them
    • extract ISBN, search online for best metadata
    • allow searching for keywords like people and places
    • kind of done: https://edwardbetts.com/monograph
  • web app for building village newsletters
    • lots of villages make a newsletter and post it as a PDF online
    • maybe all villages should have a blog
  • web project: easy chart builder
    • let people enter data quickly to make a pie/line/bar char
    • make sure chart looks nice
    • short URL
    • offer buttons to post to twitter or web
  • database of festivals
    • somebody built it: http://thefestivalgadget.com/
  • combine Price Paid Data with Companies house data to give name + address + house price
  • use Price Paid Data to find houses that haven't sold in 30 years

northern lights notification on planes

Commercial pilots flying over the Atlantic often see the northern lights, but this happens at night when the cabin lights are dimmed and people are sleeping. The pilot doesn't want to make an announcement because it would disturb the sleeping passengers.

The answer is a mobile app with notifications. Soon all transatlantic flights will include Wi-Fi. Passengers who want to see the Northern Lights can connect to the in-flight Wi-Fi and use an app on their phones to request notification. When the pilot sees the Northern Lights they can send a notification to all interested passengers.

credit card travel notification

Could work better. Should be automated. Should complain about dates in the past.

network reporting of battery levels

Imagine a family going on holiday, they want to be sure that all devices are charged before leaving the house. Battery levels should be reported to a central location on the home network. There should be a dashboard showing the battery level of all devices.

mobile phones as microphones at talks

Talks need runners with microphones for people asking questions. Everybody in the audience has a smartphone with a microphone. Let them use it.

Instead of raising a hand to ask a question. Let people type it during the talk. The moderator can read the questions and answer ones that have obvious answers. Let people ask the good ones.

Standard for reporting telephone queue lengths

I want to phone a company, but I don't want to spend a long time in a queue.

Companies should report the length of the queue on their web site. Have an option to email me when the queue is short.

Geo-demographics

Load Price Paid Data and supermarket locations. Geocode postcodes, then find 100 nearest house sales, normalise using house price index, use trimmed mean to give average house price for area.

Also consider proximity to various supermarkets.

Database of book errors

Once place to find a list of errors in published books.

Doorstep camera

I should have a recording all of interactions with people who ring my door bell. Everybody should do this. People could publish remarks made by political campaigners.

mobile digital assistant should understand queues

Join a queue, phone numbers it, tracks position.

self checkout

Machines should have login via mobile NFC or loyalty card. Then can have l kinds of customisation. Remembered interface language, favourite fruit and veg. Give warnings based on food preferences.

station ticket machine

Should tell you when your train is and from which platform.

Mapping between OSM tag and font-awesome

This should exist. Lookup railway=station and get a railway station glyph.

Wired broke the formatting for their back issues

  • Crawl all Wired back issues from archive.org
  • Make available on a web site with full text search, run entity extraction
  • Construct an ePub of each issue

Cameras pointed at bike parking

  • Offer free parking for 2 hours, then start charging.
  • Detect abandon bikes.
  • Catch people stealing bikes.
  • Track usage to justify building more.
  • Live reporting of space availability.
  • Cycle route planning can pick nearest available space to destination.

Fon auto-login for GNU/Linux

  • Port android fon login: https://github.com/rsalvaterra/Fon

Never stop optimising

  • Packing checklist app interfaces with hotel manifesto
  • Don't forget things you need, don't bring things provided by the hotel

UX blog

There are various standard processes that users do with websites. They should be documented. Adding two-factor authentication. Lots of sites offer two-factor auth, it is new and complex and needs explaining to the user. Take screenshots of activating 2FA and login with 2FA active.

Other things to screenshot: signup, adjust email notifications, things like that.

Not sure how to monetise.

API to find current train given location and time

Database or wiki of train facilities

  • details of on-board facilities
  • does the train have power sockets next to each seat?
  • how about USB sockets?
  • is there wifi? is it free?

Database or wiki of airport wifi and power

  • Where is the power in airports
  • Is it free

Python wrapper around udd-mirror

  • There should be a Python library for accessing the Ultimate Debian Database (UDD)
  • Maybe with SQLAlchemy
  • https://udd-mirror.debian.net

Database or map of office bike parking

I'm cycling to an office, does it have bike parking? How much? Is it covered?

Rotating photos

  • eBay should let buyers rotate photos
  • it should be a browser feature

Geolocation should work in road tunnels

GPS doesn't work in tunnels because the sky isn't visible. Navigation on the phone has to make assumptions based on average speed when entering the tunnel. It should be possible to include radio beacons in a tunnel that provide location. These beacons could be wifi or something else. There should be a standard. All new tunnels can be build using the standard, old ones can be retrofitted. Beacons need power, all tunnels have power for lighting.

QR codes for appliance user manuals

Household appliances come with user guides on paper that need to be filed. They should include a QR code sticker or NFC beacon that can be scanned with a mobile to display the manual.

Kickstarter for standards

The world needs more standards, but who is going to fund and write them? Either kickstarter should include funding of standards or somebody should clone kickstarter.

Household radio time beacon

Something in my house should be transmitting the current time. Possibly my wifi router. The oven and microwave should include a radio chip that can pick up the radio beacon and set the clock. I should never need to set any clocks in my house.

Standard for location of markdown source of a web page

Lots of web pages are written in Markdown and converted to HTML. The markdown should be made available via the web with a link in the header of the HTML to the Markdown.

rss2email converts from HTML to text, it could use the markdown version to avoid converting from text to HTML and back.

List of places, sort by distance to another place

I have a list of French airports, but I don't know where they are. Can I get them sorted by distance to Morzine? Originally I was thinking in terms of driving time, but now I think direct distance is good enough.

Or show the airports on a map.

Standard HTTP error code for rate limited (done)

If I hit an API too fast how does it warn me? The standard way to report errors for HTTP is via an HTTP error code. There could be a header that suggests a delay before retry.

Actually, this exists: 429 Too Many Requests. See https://httpstatuses.com/429

Code to parse references in ePubs

No standard format for writing a reference to another work or paper in a book. Somebody should write a parser.

Campaign by sending Birthday cards to MPs

Date of birth is known for most MPs, it is in Wikidata. That means we know when their birthdays are. We could have something that shows upcoming birthdays for MPs, then people can send them a birthday card and include a note about an important political issue. Will have more impact than a letter or e-mail.

Save paper, warn when printing a page with one line of text

When I print a web page there isn't sometimes a second page with only a single line of text. Any software that I use to print should warn me before this happens and offer an alternative like reducing the margins of the first page.

Can this be built in a generic way? All Linux printing goes via CUPS, could this be a feature within CUPS?

Apps where I'd like this feature: Libreoffice, Firefox, Chrome, GNOME apps, KDE apps

Next steps: write a detailed page, maybe a specification, put it on a wiki or github. File bug reports with each project.

Standard API for bug tracking software

Every bug tracker should provide an API so it is possible to search for bugs and submit bugs via the API and the differences are hidden.

Reflow man pages on window resize

See https://unix.stackexchange.com/questions/11389/dynamically-reformatting-man-pages-on-terminal-dimension-changes

script for adding books to Wikidata

Reading an authors article, it mentions books, these should items on Wikidata.

Example: https://en.wikipedia.org/wiki/Rachel_Sherman_(sociologist)

System should take an ISBN or title + author, then search online for info and use it to build a Wikidata item.

I could use it to create Wikidata items for all the books I've read.

project to add --help to all Linux commands

Write a justification. The survey commands in Debian. Give numbers. Make a nice website. Write patches.

produce word counts for IA books

Internet Archive cares lots about page counts, but what about word count? Write code to process abbyy XML and return Word count.

Free Software central bank interest rates scraper collection

Central banks post daily interest rates on the web. These numbers should be made machine readable. Different financial companies are going to end up writing their own parsers. Might as well have a collection of open source parsers. For example:

  • STIBOR: Stockholm interbank offered rate
  • LIBOR: London Interbank Offered Rate

Mobile notifications for end of adverts

People watching live TV want to do other things during adverts. Have a machine with a TV tuner watching TV and spot the end of ad breaks, send notifications to people when the ads finish. Users have to tell you what TV their watching. You can sell this data. Maybe you can guess the channel using audio from the microphone.

Strip affiliate tags from amazon.com links in comment

There are people posting book recommendations on HN with affiliate links. Shouldn't be hard to add code to forums that looks for Amazon URLs and removes the affiliate tags. Could handle plain text and HTML. Sometimes the tags are hidden by using Amazon short links, these need to be checked by requesting them. Could also include an option to warn about affiliate tags without removing them.

Add correction to codespell dictionary

Command line or web UI, paste a word, the code adds it to the dictionary while keeping it sorted. Submits a pull request.

https://github.com/lucasdemarchi/codespell/blob/master/codespell_lib/data/dictionary.txt

UK rail journey pricing by mile and by minute

Rail journey planners give you journey duration and price but don't give you distance. Need to build one that gives price per mile and price per minute. Or just make a table of all rail journeys with price per mile information.

Make page for search engines to index for each organisation registered with the ICO

https://ico.org.uk/about-the-ico/what-we-do/register-of-data-controllers/download-the-register

Pages in ico.org.uk aren't appearing in Google. Data is available for download, updated daily.

Could serve with adverts.

Python module to check for and reject reserved names

System to guess web search engine syntax

Many web sites featuring search don't include details of the search syntax. What if we had a system that given a search engine URL would do a series of searches to guess the search syntax.

Used Tesla site has poor filtering

Filter on price, battery and colour. Detail page shows year of manufacture. Would like an Autopilot 2.0 filter.

Could scrape the site and add these features and alerts.

Use codespell dictionary for spell checking in web apps

People write prose in web forms and make spelling mistakes. Spell check is hard because of false positives. Could use the codespell dictionary to suggest corrections. Add to django and wtforms

Add place of birth and place of death to Wikidata using infoboxes

Wikipedia has lots of person infoboxes that contain place of birth and death, sometimes this info is missing from Wikidata. Could try parsing every person infobox on Wikipedia to find place info, or use dbpedia.

Automate rotating credit card offer activation

US credit card require activation every three months for 5% cashback on a rotating set of categories. Activation could be automated. This could be procmail + python on Linux, or some kind of gmail plugin.

Birthday app

Track birthdays. Have I sent them a card? Have I sent a present? Include postage deadlines.

Postage deadline API

Sending a letter or parcel from point A to point B, how long does it take? Information is spread around the internet on different websites, there should be something generic. Example query: send a parcel from California to Berlin. I want to know the various options, how fast they are and what they cost.

Should be aware of Christmas posting deadlines.

Example: https://www.usps.com/holiday/holiday-shipping-dates.htm has last posting dates for:

  • Contiguous US
  • Alaska & Hawaii
  • International Mail
  • Military Mail

Interface for adding corrections to codespell dictionary

Web or command line. Enter a spelling mistake, code searches github code to check the frequency of the word. Find correction via ispell, aspell or hunspell. Clone codespell, warn if dictionary isn't sorted. Create a branch, add the correction in the right place. File a pull request.

Menu parser

Given a URL the system can parse a menu and turn it into structured data. Handle different types of menus, like individually priced items or set menus. Could include definitions for tricky words. Understand menus in French and translate to English.

OSM object tag editor

Existing editors have a focus on editing the map, sometimes mappers just want to edit the tags. This works okay in ID for small objects, but fails for large objects like admin boundaries. At the moment I use JOSM, but there could be a web interface that shows the map without allowing edits and includes editing tags of items.

Next level is editing multiple items at once, for example moving a wikidata tag from one relation to another.

Database of hippie places

Starting list: Ballymena, Bath, Brighton, Bristol, Cheltenham, Hebden Bridge, Glastonbury, Hove, Leicester, Lewes, Llanidloes, Machynlleth, Monmouth, Oxford, Saltaire, Stroud, Totnes, Wivenhoe or London's Stoke Newington

Include population, details of supermarkets, railway stations and typical rail journey time to London.

FOSDEM key signing party android app

Most participants make notes on paper, then use caff later to sign keys. Key signing could be done with a mobile app that e-mails signed keys at once, no delay. It does mean having signing keys on the phone which some will disapprove of.

People are normally in order, but sometimes not. Need to be able to find people by entering a number.

Could show key images.

Parser that turns twitter screenshots back into text

Twitter screenshots are popular but disappointing. Want to turn them back into machine readable format. Could even search twitter by time and handle to find the original tweet, if it hasn't been deleted.

Audiobook with illustrations

The narrator of an audiobook refers to a diagram or table in a book by suggesting the listener check the accompanying PDF. There should be

Hotel booking/review site scraper suite

Checking out hotels is easy via a web browser, but I want to automate it. This is so I can write code that finds flights plus a hotel. Sites to scrape: hotels.com, booking.com, tripadvisor.com, kayak.com

Include filtering by hotel star rating, customer review score and price.

Python module to test if a name is famous using Wikipedia

Somebody running a busy site is interested if anybody famous joins the site. There should be a software module that checks the name to see if there is an article on Wikipedia, then message an admin to have a closer look. Ignore common names.

Common name metric

Software module, given a human name, tell me how common it is in a given region.

Subscription information service

Provide data about different companies that offer subscriptions (netflix, spotify, amazon prime, blue apron). How many subscribers, what price for each plan. Data to be used by investors or competitors.

Website for tracking subscriptions

Subscription data aggregate. Consumers can track their subscriptions, see how much they spend in total. Monetise by suggesting new subscriptions. How does one track? One of these maybe: 1. ask the user which subscriptions they have 2. request user login/password for each service (probably not sensible) 3. via credit card transactions, needs access to credit card somehow 4. browser extension to access saved passwords and cookies - this should work

Visiting new city, what transport apps are available

Need a database of cities, enter a city name and see get a list that might include: * local transit operator - can I buy a ticket with the app, does it work with my foreign payment card? * ride hailing (uber / mytaxi / and so on) - any local restrictions? can i use ride hailing at the airport * municipal bike sharing * dockless bikes * food delivery * citymapper type apps

Flight search that takes getting to the airport into account

Budget airlines fly in the early morning or late evening. Travelling to Stansted by rail is harder at these times. It would be useful if when looking for flights the price included either the cost of the rail ticket, airport parking or taxi.

Database of human settlement housing stock information

What kind of homes are there in a given place. Different house types.

Rail journey planner for the future

How long does Reading to Cambridge take once Crossrail is active? Or routes that use HS2 once it is live. I want a journey planner that knows the routing, frequency and timing for future projects.

Subscription tracking website shows you the real cost of subscriptions

Site aimed at making people think about the cost of subscriptions and using the money to pay off credit card debt or invest in index funds.

Users enter all their subscriptions (Netflix, Amazon prime, cable TV, snack delivery, recipe meal kit services). We calculate how much money they'd make over a decade by cancelling the subscription and putting the money in a low-cost index fund.

Also show how much money could be saved by using the funds to pay off credit card debt.

Include broadband subscription? Yes.

Ask for country, then ask if there is a cable/satellite TV subscription. For US ask for either the monthly payment or ZIP code, then show packages available from the local cable company. Side project: API for available TV packages given ZIP code.

Let users enter new subscriptions, site e-mails admin, admin can accept, reject or edit.

Include nicely formatted page for printing, so users can fill in the site for a friend or family member and hand them a print-out.

Add 'located in the administrative territorial entity' information using Overpass

Given a Wikidata item with coordinates it should be possible to add information about the administrative territorial entity in which it is located. We can use the Overpass API to find all the containing areas, and pick the smallest one that has a Wikidata identifier.

Campaign against security theatre

  • hostname ideas:
    • securitytheatre.fail $38.34 per year,
    • securitytheat.re $15 per year.
    • securitytheatre.lol $38.34 per year
    • securitytheatre.wtf $32.29

Simple site explaining ticket restrictions

  • For example: http://www.brfares.com/#faredetail?orig=STD&dest=PAD&grpd=1072&tkt=SVR

  • Nicely formatted version of this info:

    RETURN TRAVEL

    Not valid on trains timed to depart: London Paddington after 04:29 and before 08:10 and after 16:40 until 18:31 Reading after 04:29 and before 08:35 and after 17:10 until 19:01 Slough after 04:29 and before 07:50 London Waterloo after 04:29 and before 08:15 Not valid on trains timed to arrive Salisbury after 04:29 and before 09:30.

Removals comparison site with video survey and quotes

Customer visits site, requests video survey. Comparison site is responsible for the survey, provides details to the movers who give quotes. Nobody needs to visit, only one survey required. Survey video is stored for reference.

National Trust and English Heritage membership price history

Build dataset of historical pricing for membership of National Trust and English Heritage

Web database with free license photos of hardware

  • Provide a service that allows lookups via ID like USB ID or PCI ID.
  • Useful to show photos of hardware in error messages.

UK postcode lookup to find TV region

iCal to PDF converter

councillor list web page parser

There are over a hundred councils in the UK. They should have a list of councillors on their website, but they write them in different ways. It would be good to have software that could parse and extract a list of councillors and party affiliations. The software needs to understand different ways of writing party names.

Data to extract: name, party, ward, photo

convert UK company name from upper case to mixed case

Companies house gives names in uppercase, better to show in mixed case. Can get a rough approximation using Python str.title(), but need special case to avoid turning abbreviations like UK into Uk.

Hotel site based on where famous people have stayed

Show me all the hotels where Winston Churchill has stayed. Could just be some web pages, index by Google, affiliate links for the hotels.

Hotel site based on rail routes

  • I want to book a hotel within walking distance of a station on the line between Waterloo and Reading

Wikidata mirror with adverts

  • System for building web databases covering particular topics:
    • countries, big cities, US states, grape varieties
  • possible hostnames: datatable.directory
  • hostnames for each data type:
    • datatable.country: $38.38 per year
    • table.country: $38.38 per year
    • archive.country: $38.38 per year
    • infobox.country: $38.38 per year
    • factbox.country: $38.38 per year
    • databox.country: $38.38 per year
    • grapevarieties.info: $20.72 per year
    • datatable.directory - not available
    • infobox.directory: $21.17 per year
    • factbox.directory: $21.17 per year
    • databox.directory - not available
    • information.country: $38.38 per year
    • databox.page - not available
    • factbase.directory: $21.17 per year
    • countryl.ist: $23.33 per year
    • datal.ist: $23.33 per year

Python module for postcode to constituency lookup

Maybe we can compress the postcode to constituency lookup table, by merging postcodes with a common prefix

If all postcodes that start with SN7 are in the same constituency it takes less space.

Tool for watching data files on the web

Somebody publishes a CSV or JSON file on the web. I want to watch it for changes. I can write code that downloads a CSV once a day and reports any changes. This should be a generic tool, no programming required. Could implement the same thing for JSON files. Tool should allow users to specify the check frequency. Changes are reported to the user by e-mail.

API for food hygiene ratings.

User story: driving past a restaurant, we wonder about the food hygiene rating, so we ask a voice assistant, and it tells us.

The first step to making this work is to provide an API that voice assistants programmers can use.

There could also be a mobile app. Another idea is a browser extension to add hygiene ratings to tripadvisor and similar review sites.

Another idea: more fine-grained category, type of restaurant instead of just restaurant.

Restaurant name classifier

Given a restaurant name determine the category of restaurant

Measure railway line in chains

The UK rail network likes to give locations in miles and chains from a given point. There should be a way to use OSM to convert these numbers into coordinates or convert in the other direction.

Reading station is a given number of miles+chains from Paddington, should be able to use this information and OSM location of rail lines to give coordinates. Same in reverse, coordinates of station plus OSM rail line info to give miles and chains from given station.

Maps for realtimetrains

Map of train journeys, based on information from realtimetrains. Use TIPLOC location data.

Web front end to taskwarrior

The urgency calculation in taskwarrior is clever. Something like it should be available in a web browser.

Use taskw to access taskwarrior. Each users needs files stored in file system.

This could be a flask blueprint, or a service that anybody can sign up for.

Airport parking scraper interface

If I'm comparing flights from different airports I should factor in the cost of airport parking.

Each airport has parking as part of the airport website. Enter dates and times and get a price for different options.

Would be nice to have a Python library that given airport, dates and times can go scrape the appropriate airport website and return the price options.

Next step is to combine flight search with airport parking prices. We can calculate the parking times from flight times.

Need to consider which airport terminal as well as airport. All BA flights from Heathrow go from Terminal 5, so look for parking at that terminal.

Could then show a list of flights which include the cost of airport parking.

JSON file format to represent file listings

Way to share a directory structure with names and metadata for directories and files.

Python module to extract article title from web page title

Web page has a title like this:

"‘Extremely active’ jumping worms that can leap a foot raise alarm in California | California | The Guardian"

I want to extract the actual article title:

"‘Extremely active’ jumping worms that can leap a foot raise alarm in California"

Self hosted TripIt clone

Somebody else has asked for it. https://www.reddit.com/r/selfhosted/comments/419wcl/selfhosted_alternative_to_tripit/

Mostly about writing parsers for booking confirmation emails.

Hosted version at myitinerary.app

TripIt has an API: http://tripit.github.io/api/doc/v1/index.html

No development is happening with TripIt, how hard would it be to clone it. Look for APIs to access the big booking websites, or failing that just web scrape them.

National Rail route planner based around maps

The National Rail route planner doesn't show any maps. Would be nice to build a site that showed results from the National Rail route planner on openrailmap. Google maps kind of does this already.

Street shop type counts

Find what types of shops are on a street in OSM. Input is an OSM way, includes connected ways with the same name.

Similar problem: find top streets for a given type of shop. Within a bounding box show me the top streets for a given type of shop.

Calculate gender balance from conference schedule pentabarf XML

Use gender-guesser Python library

Journey planner that just lists the countries you need to drive through

What countries, and is it safe?

Example: UK to India, can you do it overland? There are routes that take you via Ukraine and Russia. Most routes go via Afghanistan and Pakistan. Can you drive across the border between India and Pakistan?

How much have I spent at Deliveroo and Uber Eats?

Web scraper that downloads your order history and calculates statistics like monthly spend and all time spend.

Journey planner via tourist attraction

I want to plan a route from Cambridge to Weymouth and see National Trust gardens near the midpoint of the journey. How much of a detour is each one.

Find a place to live near National Trust

Which location is within a 1 hour drive of the most National Trust places.

Mobile first web based Pentabarf XML viewer

Like giggity, ConfClerk or confy, but web based. Use the giggity menu JSON to show upcoming conferences. Allow people to star talks of interest. Send notifications when a starred talk is about to start.

Open Source rail mileage engine

Like https://my.railmiles.me/mileage-engine/, but open source.

Plan me an evening in Manhatten

I'm staying in rural New Jersey, visiting Ted Nelson to work on Xanadu. There's not much to do or see in rural New Jersey, it is boring. NYC is one hour away by car, or two hours away by train. I was able to plan an evening in NYC which involved driving to the NJ side of the Hudson River, parking the car, taking a ferry and walking around Manhatten for 2.5 hours. I took a cross town bus from the UN HQ back to the ferry, then drove back to rural NJ.

While in New York I visited the high line, empire state building, times square, rockefeller center, grand central station, New York Penn Station, flatiron building, UN HQ.

Can I have software that automates my planning? I should figure out the journey to NYC is it best to drive or take public transit. If I drive, where do I park. What sites should I visit, what order? How do I get about. It could include where to buy dinner.

It should look at the weather and tell me what outfit to wear.

GPG keys on a web page that can be indexed by search engines with

For example Steve Wozniak is 5796CED69AC67EC1B42B56CB79338524A90BB342

More automated web version of flickr mail command line tool

When I find an article on Wikipedia without a photo I look on Flickr for images. If I find one then I send a flickr message asking the photographer to change the license. I use a command line tool, that I wrote, called flickr_mail. Then paste the generated message into flickr.

It should be possible to use Playwright to search Flickr for photos, then select the ones I like and send the message to the owners of the images.

Ultimately this tool could be opened to others people. Would need to keep track of abuse. Don't want to annoy photographers by sending them lots of similar messages.

Could generate stats about how successful the requests are. I should download all my sent and received flickr messages, then check to make sure all relicensed images are actually on Wikipedia, and see what percentage of requests are successful.

Flash card spaced repetition system or learning airport codes or UK railway station codes

HPB information site

Build a site that shows more information for each of the units at HPB. Include a way for people to comment or review on a unit and upload photos. Let people record when they stayed at a given unit.

Command line currency conversion

This command can get me the current exchange rate:

curl -s "https://open.er-api.com/v6/latest/GBP" | jq -C .rates.EUR

Local petrol prices in home assistant

Fuel price data in JSON: https://www.gov.uk/guidance/access-fuel-price-data

Web tool to spell check websites

Web tool shows a form that takes a URL, mirrors a website then runs codespell and shows spelling mistakes.

Bin day API

There is a home assistant module that gives details of when bins will be emptied. Put it behind an API and offer it as a service. Free trial with a certain number of postcode lookups in a fixed period, then start charging. Might be useful on estate agent web sites or some kind of home dashboard.

Use for an old tablet: push images to the web browser

I've got the original kindle fire and an old iPad from my Dad. Both of them have an old OS and browser and can't be upgraded. I often want to show things to people, but I don't want to hand people my phone or laptop, it would be nice to hand them a cheap tablet. The display is bigger on the tablet on the phone.

The ultimate is if I could run an android app shown on the tablet, running remotely on my phone. I think this is too hard.

An alternative would be a web application that pushes images to a browser. Run the image server on my laptop or server and enter the URL on the tablet. Then on the host machine I can push new images and they load automatically in the browser. Could be built with web sockets or Server Side Push. The browser can report back the screen resolution and the image server can shrink the images to the right size.

Web app for discussing books

At the book club meeting people were talking about books they've read. I'm feel like there could be software to assist this discussion. Maybe a page showing book covers of books that the book group has read in the last year.