Release Notes

What has changed on Abolish the Islands Trust, newest first. We build in the open, and anything a visitor can notice gets written down here — including the corrections.

• The rule, lot by lot: which island lots can lawfully add a suite 2026-10-05
  The /inequity study now names its specific measure. We read all 23 island land-use bylaws page by page and matched each suite, cottage and dwelling rule to every residential lot using the Trust's own zoning map, with page-checked citations for every row. On Thetis, Lasqueti, Gambier and South Pender, 100% of residential lots have no lawful separate-suite route; Hornby 81.6%, Gabriola 82.2%, Denman 52.2%, Galiano 31.0%, North Pender 7.4% - 7,821 lots across nine islands. Islands whose rules depend on shaded map schedules not yet matched to lots (Mayne, Salt Spring, Saturna) or with no parcel fabric are named with their reasons, not silently dropped. These are rules on paper, not permits; what they cannot show - who was kept out - is stated on the page. View
• The inequity study now reads four censuses, not one 2026-10-05
  The /inequity page gained a four-census panel (2006-2021) built from Statistics Canada small areas matched island by island by the census's own representative points, with every figure carrying its year, denominator and geography method. It shows visible-minority shares, single-detached housing, turnover and renter shares, and employment rates for the islands and named controls. Island figures for 2006 and 2011 racialized shares are not published because the source files could not be obtained; islands whose census areas also cover neighbouring islets are named on the page instead of being silently dropped. Everything shown remains a disparity to explain, not a causal result. View
• Inequity study rewritten in plain language 2026-10-04
  The all-islands study now uses short sentences and explains its key terms. The census figures, source links and North Pender suite rule are still there. The page still says what we know, what we are checking and what we cannot yet claim. View
• New all-islands investigation into family housing and equal access 2026-10-04
  The new inequity section tests a specific question: do land-use limits on suitable shared homes, suites and cottages make island housing less accessible to groups that more often live across generations? It pairs official BC household data with an exact North Pender zoning example, shows the island census figures with their denominators, and names the causal evidence still missing. This is an investigation, not a finding of racial discrimination. The page and its sources are available in readable HTML. View
• Twelve bylaw letters reported sent to the Islands Trust 2026-10-03
  Tony Aly confirmed that he personally sent twelve letters on 3 October about proposed or adopted bylaws affecting Denman, Gabriola, Galiano, Gambier, Hornby, Lasqueti, Mayne, North Pender, Salt Spring and Saturna. The all-islands bylaw audit now lists the recipient, subject and prepared text for each letter on its island page; the relevant island-town bylaw pages link to that record. This is Tony's send confirmation. We have not independently checked the exact sent-mail copies or received Trust responses to these letters. View
• The honest receipt: fifteen years of the Trust's money, person by person 2026-09-21
  The money argument now has its deep dive. The finance page reads all fifteen years of audited accounts, FY2011/12 through FY2025/26, and the headline is the one neither side expected: measured per trust-area resident and adjusted for inflation, spending is flat to down, between about 2 and 12 percent depending on what the post-2021 population turns out to be. What moved is the composition. Audited legal fees nearly tripled to $755,442, 7.24 percent of FY2025/26 expenses, and one firm, Young Anderson, was paid $766,304 in FY2025/26, the top external supplier in 8 of 11 usable SOFI years with the full series printed, trough and dip included. The Trust's own 'Bylaw enforcement' budget line rose 78.2 percent plan-to-plan 2019/20 to 2025/26, while the Trust's own figure for bylaw-ticket revenue is $5,000 to $16,000 a year, waived on compliance by policy. The Conservancy segment roughly doubled, half that in real per-person terms, with the federal Species at Risk grant era named and donated land kept as a separate labelled line. The cost of Trust Council governance is printed as governance cost, with the real per-person decline beside it and named advocacy items listed separately, well under $1M a year, with our definition of that cut printed where it is used. SOFI remuneration appears only as bands: 31 employees over $75k, up from 13, with the top listed remuneration rising from about $140k to about $219k and no staff names anywhere. The strongest finding is a publication fact: the FY2024 budget existed, the audited statement says so, and its document is the only year 2011/12 to 2026/27 never published. Two blocks were added the same week. The court record behind the legal numbers: fifteen years of audited legal fees total $4,990,050, 4.25 percent of all spending, escalating from $378,713 to $755,442 in three years, with one firm paid $4,136,443 over eleven years and no other law firm on the supplier schedules at all. The win record is published plainly and neutrally: of the 18 decided matters the Trust succeeded outright in 15 and has no un-reversed defeat; the criticism the record supports is the cost concentration, the budget-low mechanism quoted from the committee's own papers, and that no per-case spend is published anywhere. And the staff complement: 59.3 published FTE in FY2021/22 rising to about 70 by FY2026/27 (later years derived and labeled), against an unattributed 1998 estimate of about 30, with the $75,000-plus headcount doubling as the independent check. Every figure traces to a published document and page, planned and audited figures are never mixed, and a test re-reads the data files on every build. The cost-per-resident page is unchanged; the finance page states the difference in method and links back. View
• Denman's new housing rules, in plain words: a flyer to print and pass on 2026-09-20
  Bylaws 260 and 264 run to hundreds of pages, and most people will never read them. There is now a one-page version written for a neighbour rather than a planner: six plain cards on what the two bylaws change, the answer to the “doesn’t every R2 lot get a second house now?” question (only R2 lots under four hectares get one small second home; bigger rural lots get houses by area, with suites inside), a drawing of the three lot sizes, and the colour map of Denman’s 934 lots, where orange means the new rules allow fewer homes — 148 lots change down, 14 change up. It says the rules are not final, it says the “about 100 fewer homes” figure is staff’s estimate of the most Denman could ever build and not a forecast, and it carries the source list on its face. Print it, leave it at the store, tape it to a post. The PDF is at denmantown.org/share-art/housing-review/denman-260-264-plain.pdf, and the Denman bylaw-audit page links the same files. View
• The Executive Committee decides Hornby and Galiano bylaws this Wednesday 2026-09-20
  The Islands Trust Executive Committee meets on Wednesday 23 September 2026, and its agenda lists three bylaws for approval: Galiano Island Bylaw 300 as item 7.1, and Hornby Island Bylaws 176 and 177 as item 7.2. Our bylaw audit found errors in all three that can still be corrected if the committee returns them with directions: Galiano’s Bylaw 300 grants its permitted uses to a zone that does not exist, Hornby’s Bylaw 177 renumbers part of the land use bylaw and leaves the old numbers in place, so that the secondary suite rule would point at the section on fences, and Hornby’s Bylaw 176 relies on a Schedule 2 that is not attached. On Hornby there is more than drafting at stake. At the public hearing on 1 September about 68 people came, thirteen spoke, nine of them opposed the bylaws and two supported them, and two asked that the decision be left to the incoming committee. The committee gave both bylaws third reading the same afternoon, and its minutes record no tally and no discussion. We oppose the two bylaws as they stand, and the Hornby page now says so, with the hearing record quoted and linked. The Hornby and Galiano pages each have a letter ready to copy, and they now say that it has to arrive before Wednesday. Denman’s Bylaws 260 and 264 are not on that agenda, so they will come to a later meeting. The letter about them on denmantown.org was sent to the committee on 20 September. View
• New: the bylaw audit, for every island 2026-09-19
  A bylaw is the law on your land, and nobody proofreads them for you. We have read the land use bylaws and community plans of fourteen island areas, from Ballenas-Winchelsea to Thetis, for references that point nowhere, rules that cannot both be obeyed, and things that were left out: 191 findings, resting on 729 quotations. Bylaws still on their way to approval come first, because those can still be fixed, and several are close. On Salt Spring, draft Bylaw 553 is up for three readings on 24 September with the public hearing waived, and written comments close at 4 p.m. on 23 September. Hornby’s Bylaws 176 and 177, Galiano’s Bylaw 300 and Denman’s 260 and 264 are with the Executive Committee, which has one month from the day the Secretary receives a bylaw and whose silence counts as approval. North Pender’s Bylaw 235 is waiting for the Minister. Gabriola has a whole new community plan at first reading. Each of those pages has a letter ready to copy, addressed to the body that can still act. Every quotation on every page is a link: it opens the bylaw at the paragraph the words are in and highlights it, or opens the Islands Trust’s own document at the page. A script checks each one before it is published, and we opened the links on the live site in a real browser as well. Not everything found is against the Trust: most cross-references in most bylaws are right, each page lists what was checked and found sound, and Bowen’s page says plainly that its bylaw is a municipal one. We also found that the Trust’s listing for Mayne Island’s Land Use Bylaw currently serves the community plan instead, so the zoning in force there cannot be read on the Trust’s website. View
• Bylaw pages you can search, and links that land on the sentence 2026-09-19
  We checked our own citations the way a reader would: follow the link, then look for the words we quoted. Too often they could not be found. Our copies of the Denman bylaws had been made by a tool that put a space between every fragment of text in the PDF, so the 64-hectare rule read “6 4 .0 hectares”, section 8.12 read “8 . 1 2”, and searching the page for the real words found nothing. A link to a subsection such as 5.5(3) often landed at the top of a zone that runs for three pages. Denman’s Land Use Bylaw 186 and proposed Bylaw 264 have now been read again from the PDFs, line by line. Only the spacing changed: for Bylaw 264 the old and new text are identical, character for character, once spaces are ignored. Every subsection, every table, every site-specific rule such as A(5) and every definition now has its own address, and a link highlights that one paragraph. The check marks in Bylaw 186’s permitted-use tables, which had been showing as empty boxes, are check marks again. Following a citation no longer costs you your Back button, which used to take three presses. One correction of our own: we had quoted Bylaw 186 as listing “woodsheds”, and the bylaw says “wood sheds”. Along the way the checks turned up three numbering slips in proposed Bylaw 264 itself: section 5.6 has a subsection (20) sitting between (12) and (13), the only subsection of section 8.6 is numbered (3), and there is no section 5.13. Later the same day we did the same for every land use bylaw in the library, fourteen in all, from Ballenas-Winchelsea to Thetis and including Bowen Island’s. For each one the old and new text are identical to the character once spaces are ignored. Four quotations on our “What is a structure?” pages had copied the old damage faithfully, such as “neighbou ring” and “includ es”, and now read as the bylaws do. The twelve official community plans are next. View
• Every rule in a bylaw now has its own address 2026-09-19
  The Islands Trust publishes its bylaws as PDFs, and you cannot link to one section inside a PDF. Our readable copies could, but the link often dropped you at the top of a zone that runs for pages. Each numbered rule is now its own paragraph with its own address, so a citation such as Bylaw 264 section 5.6(17) takes you to that rule and highlights it. Three faults came out of checking this, and all three are fixed: the contents page at the front of a bylaw was claiming the address before the real heading could, so 27 of 44 links into Denman's Land Use Bylaw 186 landed on a line of dots; a subsection number that appeared twice on one page was counted once, so eighteen links said they were exact and arrived at the wrong subsection; and the cited paragraph was being scrolled underneath the site header, which on a one-line rule meant the first thing you read was the next rule. We have also written the test that catches all three if they come back. View
• We have published the marked-up Housing Strategic Action Plan 2026-09-18
  When we asked the Islands Trust about housing rules, it pointed us at its Housing Strategic Action Plan. So we read the plan against the Trust’s own minutes, staff reports, budgets and meeting recordings, and marked every claim and all 31 actions. That marked-up copy went to the Trust with our questions on 17 September, and both the letter and the document said it would be published. It is now on the housing page. Of the 31 actions we could confirm 7 as done, 2 of them late. The staff report that went to Trust Council with the plan lists 8 actions as deferred, and the published action table does not show those deferrals. Three actions are marked Completed or Ongoing and we could not find the record behind them. A deferral is not a failure and nothing in it says anyone did anything wrong; we asked the Trust to correct anything we got wrong, and any correction will appear on the page. View
• Did island land go up more than the rest of BC? We checked. 2026-09-17
  One argument for the new Denman rules is that land prices doubled since 2020, which is a big enough change to justify trying new kinds of rules. Land did roughly double here. It also roughly doubled across rural British Columbia: the Gulf Islands rose 102 per cent between the 2018 and 2026 assessment rolls, the typical rural area rose 91 per cent, and 23 rural places rose more than the islands did, Haida Gwaii by 203 per cent. Houses did not double at all, and a normal island house rose a little less than the middle community BC Assessment lists. What is different here is supply: Salt Spring and the south Gulf Islands added 13 lots to the tax roll in eight years, while the province added 9.7 per cent. The page shows every number in a chart and a table, says how to reproduce it from the two public files, and lists what it cannot show, including three things that cut against our own argument. View
• What planning costs per resident, after Dave Olsen 2026-09-17
  Dave Olsen of Lasqueti Island published a comparison in The Islands Grapevine: the qathet Regional District spends about $20 a person a year on planning, the Islands Trust about $387, and he invited every island to do the same arithmetic. We took it up, and checking it changed the answer. His Islands Trust figures are exact, and so are his qathet dollars, which we can now cite to the adopted financial plan rather than to a staff member. His denominator does not hold: the qathet figure is raised from four electoral areas of about 6,237 people, not the whole region of 21,496, so that service costs about $85 a person, not $20. The gap is about four times, not nineteen. The bigger finding is one nobody had published: every regional district we checked excludes the Islands Trust islands from the planning service it taxes for, so no regional district is currently sending these islands a planning bill at all. We also corrected our own first draft, which used a 25 per cent sample population. View
• What the record shows about housing, island by island 2026-09-17
  The housing page now sets out what we found in the Islands Trust's own records for each of the thirteen island committees: whether the plan states how much housing is needed over twenty years, and which zoning changes since 2023 allow more homes. It also quotes what was said at five recorded meetings, with a link to listen to each one, and names a speaker only where the voice was confirmed. A short section says plainly what the record does and does not show, including that it does not show a law was broken. We have put the same questions to the Islands Trust and to the Ministry of Housing, and both letters and their reply dates are on the page. View
• The housing page now explains the islands' duties in plain language 2026-09-17
  What do the islands actually have to do about housing? The housing page now answers in plain words. They must get a housing needs study on time, receive it and post it, and renew it every five years. Their community plans must state how much housing is needed over 20 years and carry policies for each kind of need, and they must weigh the study when they change the plan. They do not have to change their zoning to allow more homes, plan for the study's exact number, or update the plan by a deadline. View
• The Islands Trust answered our housing question, and we published it in full 2026-09-16
  On 31 August we asked the Islands Trust and the Ministry of Housing whether the 2023 small-scale multi-unit housing law reaches the local trust committees, and which committees have allowed that kind of housing anyway. The Trust's Director of Legislative and Information Services replied on 16 September, and his reply is now on the housing page, unedited. The Ministry has not yet replied. View
• Older islanders: the aging trend and the mortality check 2026-09-14
  The page now states, with sources, that the 65-and-over share of every island has about doubled since 2006 in the Province's own census tabulation — the same trend as the whole rural coast, which the page says first. We also tested the hypothesis that islanders die younger because care is worse: the best available data contradict it (the Southern Gulf Islands health area's life expectancy of 83.6 beats the provincial 82.0, with the lowest age-standardized mortality of Island Health's fourteen areas), and no public record measures death by island at all. A standalone study, with the full tables, verification trail, and confounders, is published at output/island-aging-mortality-study.md. The share card and story sheet are rebuilt from the updated page. View
• Older islanders: the page now covers every trust island 2026-09-14
  The page had one island's numbers under a trust-wide headline. It now measures the question on each island against its own bylaw's line for a second, smaller home, using the parcel fabric and the bylaw rule table. Measured on the seven islands whose zone codes settle the line, 3,909 of 5,762 lots — 68 percent — fall below it; the line itself runs from 0.4 hectares on Gambier and Galiano to 4 hectares on Hornby's large-lot zones. Denman's published screen (604 of 858, 70 percent) is reproduced exactly by the new script before it extends the method, so that figure stands. On Gabriola and Salt Spring's rural zone the bylaw says one detached home per lot whatever the size, and the page says plainly where the answer is not knowable: half of Salt Spring and nearly half of North Pender. Pepper Lane's corrected history moves out of the argument into the news archive. The share card and story sheet are rebuilt from the new page. View
• Hornby and Bowen land-use bylaws are now readable here 2026-09-13
  Hornby Island Land Use Bylaw No. 150 and Bowen Island Municipality Land Use Bylaw No. 57 now have the same readable HTML copies as the other islands, with the PDF beside each one. Bowen’s book is a municipal bylaw, not an Islands Trust local trust committee bylaw, and the page says so. View
• What is a structure? Twelve bylaws. No shared word. 2026-09-13
  The Islands Trust regulates buildings and structures across thirteen local trust areas and does not have one definition of either word. We published the adopted land-use bylaws side by side: a tent is a sleeping unit on Denman, six tenting sites equal one dwelling on Lasqueti, a mobile home is a building on Denman and a prohibited residence on Gambier, and a tiny home on wheels is an RV on Mayne. An infographic on the page carries the quotations. View
• Every news report now has captions 2026-09-13
  Our first video report went out with none, which was wrong on a site that publishes an audit of how readable the Islands Trust's own documents are. Every report now carries captions, both presenters are named in them so you can tell who is speaking, and the words sit clear of the text burned into the picture. From now on a report that cannot be captioned is held back rather than published: if the captions are not right, the episode waits. One thing to be straight about — the wording is exactly what is said, but the timing of each caption is estimated from the script rather than measured from the audio, and the caption file says so itself. The video reports are also now linked from the top of every page instead of being two clicks down. View
• In 94.5% of Islands Trust minutes, the record does not say how anyone voted 2026-09-12
  We set out to measure whether the Trust is growing more divided, and we could not: across 2,289 sets of minutes from 2008 to 2026, only 47 record a vote tally and just two name who voted which way. The phrase 'recorded vote' does not appear once. Trust Council sometimes records a split; the local trust committees, which decide what may be built on your land, essentially never do — Salt Spring's 225 sets of minutes contain no recorded split at all. This is lawful, and the page says so plainly: meetings must be open, minutes must be kept and must be available to you, but no law requires writing down how each trustee voted. Trust Council's own rules let any trustee ask for their vote to be recorded. In 2,289 documents, nobody ever did. We publish the three votes that were recorded by name, including a 12-12 tie on the Trust's own budget, and we publish the four claims we had to withdraw along the way — three of which cut against our own case, including the study this page was meant to be. View
• The economic study: what it costs to live on a Trust island 2026-09-11
  A new section publishes median household income, monthly shelter cost, median age and core housing need for all twelve Islands Trust local trust areas, from the census cut the Province commissioned to the Trust's own boundaries. Every island's median age is more than thirteen years above British Columbia's. The page also withdraws the strong form of our own hypothesis: Salt Spring has the highest income and the lowest housing need in the set, so 'Trust islands are poorer and worse housed' is not true as a general statement and we do not publish it as one. Bowen's nineteen years of municipal accounts show self-government is not cheap, and that finding leads its section. Two islands' figures are withheld by Statistics Canada and are shown as withheld, never as zero. View
• Abolish Islands Trust News: one-minute video reports 2026-09-11
  There is now a video section. Each report is about a minute long and makes one point from the public record. The documents it relies on are listed under the video, with the full transcript. The two presenters are AI-generated characters, not real people, and every report says so on screen and on the page. The first report explains why a two-to-one local vote is not yet an adopted bylaw. View
• Sam Borthwick’s positions and results; trustee-directory correction 2026-09-10
  Sam’s profile now compares his published positions with identifiable motions, collective bylaw outcomes and unresolved questions, with sources and attribution limits. We corrected the directory and its share images: 78 sourced roster profiles cover 26 current and 52 former trustees, not every historical trustee. Unverified names inferred from minutes are excluded from public profiles. Attendance figures are labelled as archive records, which can include duplicates, and moves and seconds are distinguished from individual votes. View
• Pepper Lane’s history and today’s housing plans 2026-09-10
  Pepper Lane was a real proposal for eight seniors’ homes at 3730 Denman Road. The housing association reports that the original land was lost. The current operator lists five seniors’ homes within Denman Green’s proposed twenty, with construction funding being sought. We have corrected the pages and artwork so the old eight are not counted as a separate active project. The transparency page now uses the site’s working layout and records these material corrections. View
• Corrections across the island information and campaign sites 2026-09-10
  A six-perspective review challenged the pages, calculations, articles and downloadable graphics. We corrected claims about housing duties, ALR permissions, governance-review progress and incorporation; separated housing need from zoning capacity; and labelled parcel counts as screening estimates rather than approvals. Farm comparisons now use the same matched records on the data and farm pages. Article corrections retain earlier versions. Share images, signs and PDFs carry the corrected wording. Earlier release notes describe the versions then published; this correction supersedes their affected claims.
• Correction: three share cards carried claims the pages had retired 2026-09-10
  The share image for the enforcement page still said four of eight islands authorise proactive enforcement, a claim removed from the page on 8 September; the older-islanders card still argued the accessibility case the page no longer makes; and the I Love Islands Trust card still gave the retired land-area figure. All three are rebuilt from the current pages, the argument cards are now the size link previews expect, and a test fails the build if a page changes without its card. Shared links may show the old card until the platform refreshes it. View
• Link previews no longer fall back to a placeholder 2026-09-10
  When a page could not be rendered for a link preview, every site in the network showed a generic “Island Public-Record Network” card with an image that could not load. Each site now shows its own name, description and share image in that case, unknown archive addresses answer with a proper not-found page, and the share messages on the home, history, islands, archive, court-decisions, comparison and water pages now lead with a figure that is on the page. View
• Correction: the budget page’s link preview lost its dollar figure 2026-09-09
  When our money page was shared, the preview text read “plans to spend 1.8 million” with a broken tag where “$11.8 million” belonged. The page itself was right; the code that writes the preview treated the “$1” in “$11.8” as an instruction. Fixed for every page on every site, with a test. The home page title also carried a stray question mark in place of a dash. View
• Our email used to arrive from somebody called noreply 2026-09-09
  Every email this campaign has ever sent — the confirmations, the ten machine-written issues, the corrections, and the newsletter that went out this morning — arrived with no sender name on it. A mail app that is given no name shows you the first part of the address instead, so what landed in your inbox said noreply. The footer of those emails asks you to write to us when we get something wrong. The From line said we do not accept mail. People believe the From line. It now says Abolish Islands Trust, and the address it comes from is corrections@, which is a real mailbox a person reads. Each site in this network signs its own name: a Hornby signer sees Hornby Town Petition, and a site we have not set up yet says Island Town Petition rather than borrowing this campaign’s name. We checked it the only way that proves anything: we sent one message to ourselves and read the header off the copy that came back, rather than trusting the software that said it had worked. View
• We sent a newsletter somebody chose to receive 2026-09-09
  Thirty-six people got an email from us this morning. It is the first one a person wrote, read, and decided to send. It says what we are asking the Province for, why we think it, and where to check every figure in it. Everyone who got it had ticked a box asking us to email them. Nobody was added, nobody was assumed, and the people who left that box unticked got nothing. It also asks a question we should have asked weeks ago: how often, if at all. Nine people have now chosen. Twenty-seven have not, and until they do we will send them nothing. Doing nothing is a real answer and it is the one we honour by default. Three people got a different email first. Each of them had been sent one of the ten machine-written issues that went out in late August and early September, and each of those carried a claim we have since withdrawn. We wrote to them separately, quoted what we had said, and said what is true now. That correction went out before the newsletter did, because a campaign that has just admitted to mailing people without reading the mail should not send them more news before it sends them the apology. View
• The site now asks a better question, and gives you something to say 2026-09-09
  The box that asks whether we may email you used to say “Email me campaign updates”. That asks you to want updates, which is not a thing anybody wants. It now says what actually arrives: we read the Trust's minutes and recordings so you do not have to, roughly once a week, every figure sourced, corrections included. The box has never been ticked for you and never will be. The screen you land on after confirming a signature has changed twice. It now asks how often you want to hear from us, with the control on the screen rather than three scrolls below it, and nothing chosen for you. And it carries one thing worth repeating: the Islands Trust has no definition of the word affordable, it told itself to write one in 2015, in 2021 its own staff said on the record that they could not, and six islands then wrote eight different ones. The sources are on the same line and the full record is one click away. The share card used to run off the side of a phone screen when a name or a link was long. It no longer does. View
• What was wrong with our emails, and what we fixed 2026-09-08
  Seven people each got an email from us between 21 August and 3 September. Ten different issues went out, one at a time, to whoever was due that night. Three of the seven were sent two. None of the ten was read by a person before it left. Sending has been switched off since 5 September. Corrected twice, later the same day. This note first said six people had received one each, every one of them different; it was seven people and three of them got two, and we had the records to check before publishing it. It then dated the first of those emails to 22 August, which is what our clocks say because they run on UTC. These islands run seven hours behind that in summer, so the first one arrived on the evening of the 21st. We are leaving the wrong version described here rather than quietly swapping the numbers, because a note about our own email honesty is the last place to correct something invisibly. The emails also broke rules we should have met from the first one. They carried no postal address, which Canadian law requires on every message of this kind, and did not say who was sending them. The plain-text version, which some readers get, had no unsubscribe link in it at all. They said “reply to this email” while being sent from an address that accepts no replies. Worse, the unsubscribe page could say “you are unsubscribed” when nothing had changed, and the scanners that big employers run over incoming mail could unsubscribe people who never clicked anything. Nothing anywhere listened for a bounce or a spam complaint, so an address that had failed would have been mailed again forever. And people who signed a petition about their own island were told they had signed the one about abolishing the Trust. Every newsletter and invitation now names Tony Aly and a street address, carries the way out in both the formatted and the plain-text version, and tells you which petition you actually signed. Unsubscribing takes one press and says what really happened. The one-press unsubscribe built into Gmail and Outlook now works too: it needed a change on the mail platform we use, which was written, reviewed and shipped the same evening. Bounces and spam complaints now reach us from the mail provider and stop the mail on their own. Two more things we found while looking. The domain we send from had no DMARC record, which is the setting that lets other mail systems check a message really came from us; it is now published on all sixteen of our domains. And every time we released a change to the site, an automated test signed the petition with a fake address and the confirmation email bounced, which quietly damaged the reputation of the address we send from. That has been stopped. Not everything is fixed. Our older admin mailer and the emails that confirm a signature still do not carry the address, and they are next. View
• Five things we got wrong, and fixed 2026-09-08
  A hard check of our own published work found five errors and we corrected all of them. We cited a Hornby bylaw at a section number that does not exist in it. We dated a Salt Spring bylaw to 2011; it was adopted in 2000, and we had trusted a website file listing over the document itself. We said Denman was the only island where a second home needs a temporary permit; the bylaw's very next clause says otherwise. We published a count of Denman lots that had accidentally counted every island's lots, so 3,245 should have read 858. And our zoning data granted a secondary suite in three Denman zones that the bylaw does not grant it in, which had made our own capacity figures one home too high on 470 of Denman's 858 R1, R2 and A lots — 440 in R1, 21 in R3, 9 in R4. The Salt Spring date is corrected with the old figure printed beside it. The other four were swapped in, and the record of what they used to say lives in this note and in our code rather than on the page. That is weaker than the standard we set ourselves, and those pages need a correction line the way the housing page has one. View
• Older islanders: the page now argues what its name says 2026-09-08
  That page was named for older islanders and argued about document accessibility, with a note at the bottom saying it was not about older people. It now carries what the Islands Trust's own 2025 housing study says: that senior citizens are 35 to 50 percent of the people on these islands, that many own homes they find increasingly difficult and expensive to maintain, and that the barriers are high costs, limited land and lengthy permitting. The accessibility audit has not gone anywhere; it is on its own page in full. The page argues about barriers rather than about ageing on purpose: our own data holds places with no Islands Trust at the low end of the same range of median ages — Tahsis 58.8, Gold River 56.4, the Sunshine Coast 55.6, against island medians of 56.0 to 67.5 — so an ageing population is not by itself evidence about the Trust. View
• Correction: Bill 44 does carve out the Trust areas 2026-09-07
  On 31 August we published that there is “no carve-out” in Bill 44 for Islands Trust areas, and that the housing duty simply never reaches a local trust committee. The first half was wrong. Bill 44’s own section 47 amended the Islands Trust Act, s.29(1)(b), to list the small-scale multi-unit housing power among the powers a local trust committee does not have. The exclusion is explicit and legislated, and the Minister gave the reason in the House on 20 November 2023. We corrected the housing page on 2 September and should have corrected this note the same day; it stood for five days after we knew better. The original wording is left in place, marked, so the record shows what we said. View
• A campaign mark, and the name on every share picture 2026-09-06
  Share images, infographics, argument cards and yard signs now say Abolish Islands Trust at the top, with a simple stake mark so a picture still names the campaign if it is saved without the link. The pictures themselves are linocut-style illustrations of islands, water, fields and houses — not just type on cream paper. View
• New share images for the home page, the petition, and the governance review 2026-09-06
  Those three pictures now lead with one idea each: who decides, the exact petition sentence, and 15 recommendations versus one in progress. Yard signs for the governance review were redrawn so they read at a distance. We also narrowed a comparison headline that had said nobody ever measured the Trust: our search found no matching study, which is a smaller claim. View
• See the artwork before you share it 2026-09-05
  The toolkit now shows previews of its share images and printable signs, with download and sharing controls beside each one.
• Petition email confirmation repaired 2026-09-05
  Fixed an error that could stop the confirmation button in a petition email from counting your signature. Your private link stays private.
• Easier sign-in and safer petition forms 2026-09-05
  Sign in with a code, even when tracking scripts are blocked. Autofill and private-link recovery work without waiting for the page to notice your email. Spam checks run behind the scenes, with a clear retry message when needed.
• A clearer petition and a shorter path to signing 2026-09-05
  The homepage now explains the ask in plain language. Optional form fields fold away, and the petition page shows the exact signed text. We also fixed a mobile map caption and removed a claim that no comparison study has ever existed: our search found no matching study, which is a narrower finding. Private petition links no longer load analytics. View
• Help getting your petition link 2026-09-05
  Confirmation emails now wait in a retry queue if the mail service is down. The recovery form works separately from signing. Check your email and confirm your signature before it joins the public count.
• Meeting transcripts: what was said at the Trust's meetings, beside what the minutes say 2026-09-01
  The Islands Trust records every public meeting and publishes the video. It also publishes minutes, which are a summary written by the recorder and adopted by the committee. Nobody publishes the words. Now we do. Each transcript page carries a machine transcript of the Trust’s own recording, timestamped to the second; every line is a play button that jumps the Trust’s video to that moment, streamed from the Trust’s own server, so you can see and hear who is speaking. The official minutes sit beside the transcript, and where the two differ — something said aloud that the minutes leave out, soften, or attribute to someone else — the difference is listed with the quotation located to the second. A quotation is stored only if it can be found word for word in the transcript; anything that cannot be located is discarded. Four Denman Island meetings are up to start (20 January, 17 February, 23 June and 24 August 2026), with the rest of 2026 transcribing and other islands to follow. Two cautions, printed on every page: these are first-pass machine transcripts, so names and homophones are often wrong and any line should be checked against the recording before it is relied on; and elected officials and staff are named in their public roles while residents who spoke are shown as “Member of the public”. Every page also carries a linked source block — the Trust’s recording ID, its player, the stream file and the minutes document — and a one-click citation for the moment you are listening to. Where a meeting discusses Camp Denman, the publisher’s personal interest is disclosed on the page. View
• Housing: the duty to count, without the duty to permit — and two letters asking whether we have it right 2026-08-31
  In 2023 British Columbia told local governments to stop banning duplexes, triplexes and townhouses. The law places that duty on “a local government”, which the Local Government Act defines as a municipal council or a regional district board. A local trust committee is neither — and the same Act, in the part about housing needs reports, expressly says “local government” includes a local trust committee. So the Legislature knew how to sweep the islands in, and did it for the duty to count the shortage while leaving them out of the duty to permit the homes. The Trust’s own 2025 assessment puts the islands 2,028 homes short within five years, and never once uses the words rezoning, approval, completions or multi-unit across 570,000 characters. This is a reading of the statutory text and not legal advice, so on 31 August 2026 we put it to the Ministry of Housing and Municipal Affairs and to the Islands Trust and asked them to confirm or correct it. Both replies will be published in full and unedited, and if either shows the reading is wrong the page will be corrected and the correction will say so. Two other things belong in this note. We also killed a claim repeated all over the internet: Trust areas are not “explicitly exempted” from Bill 44 — there is no carve-out, the duty simply never reaches a local trust committee, and the difference matters. [CORRECTED 7 September 2026: that sentence was wrong, and it is left standing so the record shows what we said. Bill 44 does contain a carve-out — its section 47 amended Islands Trust Act s.29(1)(b) to list Local Government Act s.481.3 among the powers a local trust committee does not have. See the correction note of 7 September 2026.] And we withdrew a claim of our own. We had reported that open rezoning applications average 7.1 years and called that processing time; checking every open file against the Trust’s own meeting record showed 31 have not appeared anywhere since before 2021, so the age measures an un-updated register rather than the work. The retraction is on the page. What survives is narrower and firmer: you cannot tell from the Trust’s own published record whether an application was ever decided. View
• “Just look at Bowen” — we followed the instruction, and the cost finding goes against us 2026-08-31
  Bowen Island incorporated in December 1999 and is the only island that has done what the town petitions propose, so “just look at Bowen” is fair advice. We looked. Its own-purpose taxation per resident rose from $840 in 2005 to $2,360 in 2023, and its position among BC municipalities got worse rather than better — 80th percentile to 90th, above the median comparable small municipality every year for nineteen. Anyone told that incorporation is cheap is being told something this record does not support, and that finding leads the page rather than sitting under the ones that help us. The statutory half reframes the cost debate. Islands Trust Act s.47(2)(b) requisitions a municipality in the trust area only for Trust Council and Conservancy administration, not for local trust committee operations — and Local Planning is $7,760,914 of the Trust’s $11.8M budget. Incorporating does not end your Trust levy; it removes the two-thirds of it that buys planning. Bowen’s Trust bill still grew 147% since 2015/16 against the Trust Area’s 63%, because s.47(6) apportions by assessed value rather than by service received. We also mirrored the 1999 Restructure Study, which is the only baseline that exists: voters were told total taxes would hold flat or fall slightly, and that the Islands Trust line would drop from $216 to $41 on an average home. And two results we did not expect are published anyway — the mainland side of Comox Valley Electoral Area A pays more in total local-government tax than the islands do, and Bowen has outrun Vancouver’s population growth rather than converging with it. The number that would settle the argument — what an unincorporated island spends on its own government — does not exist in any source, so nobody can honestly claim it, us included. View
• Trust islands against non-Trust islands: the counterfactual, published with its falsification conditions 2026-08-29
  Nobody has ever measured whether Islands Trust governance makes a difference. The only study that tried was paid for by the Islands Trust Conservancy and compared ferry-only islands to Greater Victoria; the Province had the data for a real comparison in 2002 and declined to draw it. This page is our attempt, and because it comes from a campaign it is built to be checked rather than believed: each of the five designs states in advance the result that would count against us, the findings that have already gone the wrong way are listed by date at the same prominence as the findings that help, and the indicators we cannot yet compute are listed beside the ones we can. The bias warning sits above the first table rather than in a footnote, because it runs toward us: the Trust Area was drawn in 1974 around the islands already under development pressure, so finding less development inside it would prove nothing. Preparing the page found two wrong numbers in our own protocol — a marine designation credited to Texada that is ten kilometres away in a different electoral area, and a Quadra figure that divided zoning from a 1,064,539-hectare region by a land area that was not Quadra's. Both are corrected, and both are published on the page as retired findings. View
• A seventeenth judgment: the damages action against the Denman trust committee 2026-08-29
  Ellis v. Denman Island Local Trust Committee, 2020 BCSC 935, had been recorded on this site as a decision we knew existed and could not find. It is now located, read from the courts' own published reasons, and added. It is another Trust win, which is why it goes in: most of the pleading was struck as an attempt to re-litigate matters already decided, and the remaining claims in negligence, nuisance and misfeasance in public office were dismissed on summary trial, the judge holding that misfeasance in public office cannot be grounded in conduct the courts have already found lawful. The counted record moves from sixteen decisions to seventeen: the Trust has now won fourteen, lost one and split two. The one point it did not get is published too — it asked for costs at Scale C and was awarded Scale B, the judge finding no complex or novel legal issue. View
• Authority to enforce, discretion to defer: eight islands, eight enforcement regimes 2026-08-27
  We read the Trust's own enforcement policies — the trust-wide Policy 5.5.1, the best practices manual, the public FAQ, and all eight local trust committee policies. The Trust describes its process as complaint-based and says so plainly, and complaints must be in writing with anonymous ones refused. But Policy 5.5.1 also allows an investigation to start without any complaint in six named circumstances, including where contraventions appear to occur in setbacks from water bodies or in development permit areas. The operative word throughout is may. Seven of the eight island policies open with the same sentence: the committee has the power and authority to enforce its bylaws, and also has discretion to adopt priorities for enforcement or to defer enforcement. On the face of these documents there is authority to enforce and no duty to. Four of the eight islands authorise proactive enforcement; the other four say nothing either way, which is not the same as refusing it. And each committee has deferred something different: Gabriola has suspended enforcement of the Mudge Island land use bylaw entirely pending a review; Gambier records a standing deferment for non-permitted seawalls on North Thormanby Island; Mayne and Saturna will not enforce against unlawful short-term rentals except in listed cases; Salt Spring defers enforcement against non-compliant dwellings until there are safe, secure housing options. Two things we will not claim. Complaint-based enforcement is the norm for BC local government, not a scandal. And several of these deferrals are humane and defensible — Salt Spring's is a reasonable committee doing a reasonable thing. The point is that these are eight separate policy choices that appear nowhere together, so nobody can see them as a system. The number that would settle it — how many files begin proactively rather than on complaint, per island, per year — is not published. View
• Four out of five Islands Trust zoning codes exist on only one island 2026-08-27
  We have ingested the Trust's own zoning and official community plan layers rather than fetching them for one question and discarding them: 2,934 zoning polygons and 1,774 land use designations across 15 local trust areas. The zoning uses 509 distinct codes, and 401 of them — four in five — are used by exactly one trust area. Only one code, W1, is shared by as many as nine. Salt Spring runs 154 codes; Ballenas-Winchelsea runs three; Denman 55 and Hornby 35. We are careful about what that means. It is not evidence of incoherence: per-island committees writing per-island bylaws is what the structure exists to do, and a zone that suits Salt Spring may be useless on Lasqueti. What it does mean is that nobody, the Trust included, can compare its own islands. There is no shared vocabulary, so there is no way to ask across the trust area how much land is zoned for housing, or industry, or conservation, without building a crosswalk that does not exist. The Trust's statutory object is stated over the trust area as a whole, and its own data cannot be added up to that object. Two things we would want to know if we were reading this, so we publish them too: the layer contains one polygon whose trust area code is a typo, and its add-dates record when the file was digitised in 2008, not when any rule changed. View
• Two-thirds of the water the Islands Trust administers has no marine zoning at all 2026-08-27
  We measured every marine zone the Trust has drawn, using its own zoning layer — 541 polygons across 25 zoning areas — against the provincial trust area boundaries, with land subtracted from the parcel fabric. Across the twelve local trust areas that have a resident parcel fabric the Trust administers 453,582 hectares of water. 152,470 of them, about a third, carry a marine zone. The rest carry none. And the share is not broadly similar with a couple of outliers: it runs from 99.6 per cent on South Pender to 5.2 per cent on Gabriola, between committees of the same body under the same statute and the same object. The sharpest case is next door to us. Denman zones 78 per cent of its water; Hornby, four kilometres away in the same channel, in the same regional district, on the same ferry route, zones 16 per cent. Two things this does not show, stated as plainly as the finding: unzoned water is not unregulated, because federal and provincial authority over navigation, fisheries, aquaculture licensing and the seabed does not depend on a local zone; and more zoning is not automatically better, so an island that zones little of its water may have decided that deliberately. What it does show is that the variation is very large and, so far as we can find, has never been published or explained anywhere — including by the Trust, whose object is to preserve and protect the trust area as a whole. View
• We measured Denman's W1 line against the rule that made it. It matches. 2026-08-27
  Denman's Official Community Plan has fixed the seaward extent of the island's water zoning since 1991: out to the 30 metre bathymetric contour off Vancouver Island or mid-channel, whichever is furthest west in Baynes Sound, and elsewhere to the local trust area boundary. That is a testable claim and, as far as we can tell, nobody had tested it. So we did — 180 transects radiating from the island, measuring where the mapped W1 zone actually ends against where the plan says it should, using the Trust's own zoning polygon, the provincial boundary layer, the BC Freshwater Atlas coastlines and the parcel fabric. In Baynes Sound the mapped edge reaches at least mid-channel on 45 of 56 transects, typically 220 metres beyond it, and stops roughly 900 metres short of the trust area boundary — which is exactly what the policy directs there. Everywhere else it sits on the trust area boundary, with a median gap of one metre. The line was drawn to the rule. One anomaly is reported rather than buried: on 18 transects at the north end the mapped zone extends past the mapped boundary, by up to 731 metres, which is most likely the difference between the provincial depiction and the legal description in Schedule A to B.C. Reg. 119/90 plus ordinary digitising error in both polygons. So are the limits: we could not obtain a bathymetric contour layer, and we measured W1 alone rather than all five water zones, so the shortfall figures are an upper bound. This is the third finding in a row on this question to come out against the campaign's instincts, and it is published for the same reason as the other two. View
• Correction: Denman's marine zoning boundary was decided in 1991, not in 2009 2026-08-27
  On 26 August we published the 2008–09 file for Denman's Official Community Plan and Land Use Bylaw and reported that the record contains no discussion of how far out to sea the marine conservation zone reaches. We said at the time that a silent record is not proof that no decision was made, and that the zoning may simply have been carried forward from an older bylaw — which was the next document to obtain. We have now obtained it, and that is what happened. The 1991 Official Community Plan, Bylaw No. 60, states that the Water designation covers the marine area "as far west as the 30 metre bathymetric contour off the shore of Vancouver Island", and its Policy 4 commits the Trust to extend zoning out to that contour or mid-channel, whichever is further west in Baynes Sound, and elsewhere to the local trust area boundary. The 2009 plan carries that forward almost word for word. So the 2009 bylaw was implementing a commitment made eighteen years earlier, and the silence we found is explained: the seaward extent was not a change, and unchanged policy does not get debated. Anyone using our work to argue that this boundary was never decided, or appeared without discussion, should stop — that argument is finished, and we are the ones finishing it. What remains open is narrower and still worth asking: the bylaw states no seaward distance, the line that governs is a GIS file kept at the Trust's Victoria office, the published map is expressly a "generalised diagrammatic representation" at 1:10,000, and whether the mapped line actually matches the 1991 rule is a question we can measure. We publish this because a project that asks the Trust to disclose inconvenient records does not get to sit on its own. View
• Six Denman court decisions added to the case law page 2026-08-27
  The case law page was built by searching the Islands Trust's own document library for anything mentioning a court. That method missed every Denman Island judgment, because the Trust's library does not contain them — which meant the one island this project is run from was the one island with no cases listed. Six are now included, traced through a private archive and then read from the courts' own published reasons at bccourts.ca: a 2001 Court of Appeal decision that struck down Denman's forest cover bylaw while reviving four others, two decisions about tree cutting in the Komas Bluff development permit area (2005 and 2007), two about whether a party who has already litigated a bylaw's validity may challenge it again (both 2013), and a 2015 decision enforcing a costs award against the land itself. The counted record moves from ten decisions to sixteen: the Trust has now won thirteen, lost one and split two, and it won five of the six added here. We publish the losses and the wins on the same page, and the earlier blog post carries a dated note rather than quietly corrected numbers. None of the six changes the finding that no court has yet been asked where the trust area's outer boundary runs. View
• New Trust meeting records now reach the archive's subject pages, not just its search 2026-08-27
  The archive checks the Islands Trust's document library every day and adds whatever is new, so minutes, agendas and hearing records arrive here shortly after the Trust posts them. Keyword search picked those new documents up immediately, but the subject pages did not: which documents belong to a topic is worked out in advance rather than on every visit, and nothing was recalculating it after a daily pickup. A new document was therefore searchable and filed under its island, type and year, but listed under none of its subjects — with nothing visibly broken to give it away. Adding a document, or reading text off a scanned one for the first time, now refreshes the subject listings straight away. We also compared our holdings against the Trust's library end to end: the library reports 7,459 records, but 358 of those are the same document listed more than once, leaving 7,101 distinct documents. Eight were missing here, each apparently a fetch that failed on the one day it was new. Those are being collected, and the archive now re-checks the full library weekly so a single failed download can no longer become a permanent gap. View
• Correction: we briefly listed 449 working Trust documents as broken 2026-08-27
  The broken-record ledger exists to show which Islands Trust documents the Trust's own links no longer reach. On 27 August a new automated check ran across the archive for the first time and, for a few minutes, added 449 documents that are not broken at all. The check was requesting too many documents at once, the Trust's server responded with errors rather than files, and our code wrongly treated "the server did not answer" as "the document is gone". Of the 453 entries it added, 449 were server errors and 4 were real. The 449 false entries were deleted the same day and the genuine ones kept. The check now counts only a 404 or 410 — the responses that actually mean a document has been removed — and anything else is reported as a failed attempt rather than recorded against the Trust. It also makes far fewer requests at once, which is what caused the errors in the first place. A re-run of 200 documents afterwards found two broken links, and every one of the nine entries this check has produced was then confirmed by hand. We publish this because a ledger that documents someone else's broken links has no business hiding its own mistakes. View
• The Denman 2008–09 bylaw adoption record is now published 2026-08-26
  Nine documents obtained from Islands Trust on request are now in the archive, covering the passage of Denman's Official Community Plan and Land Use Bylaw 186 from first reading in June 2008 to adoption in May 2009 — the bylaw that created the island's W1 to W5 water zones. Three of them are image-only scans with no searchable text at all, including the submission package the Executive Committee actually decided on, so the pages that matter have been transcribed by hand and published alongside the originals. They were requested to find out where the marine conservation zone on the outer edge of the Denman Local Trust Area was decided. What the record shows is that on 3 February 2009 the Executive Committee was told the bylaw regulates the use of land "including the surface of water", approved it, and recorded no discussion of how far out that reaches. We are publishing the limits of that finding as prominently as the finding: a silent record is not proof that a decision was never made, this may not be the complete file, and the zoning may simply have been carried forward from the older bylaw — which is the next document to obtain. Records obtained on request are labelled as our copy, never as a Trust-published source, because the Trust does not publish several of them. View
• The Conservancy and output pages now show their evidence without JavaScript 2026-08-26
  "The Conservancy, measured" and "What did they do?" were publishing a headline and two links to anyone reading without JavaScript — including search engines — while every sourced figure sat in the browser bundle. On two pages whose entire claim is that the evidence is public, that was the wrong way round. Both now render in full: the Conservancy page carries the protected-hectare totals, the 8.51% share, the holdings, the pace of growth, the cost against the whole budget, the NAPTEP uptake, the named places with their documented counterfactuals, and the concession that the work is real. "What did they do?" carries the per-area document counts, the caveat that a published document is not an action, and — most importantly — the list of things a local trust committee is legally barred from doing, which is the part that stops the page being unfair. View
• Correction: two archive documents were listed as live after the Trust deleted them 2026-08-26
  The broken-record ledger could only ever re-check documents already known to have failed, so it could clear a repaired record but never notice a new casualty. A document that we mirrored successfully and the Trust deleted afterwards stayed off the ledger while the island pages went on listing it under "Each links to its exact source". Two were found this way — a Gambier agenda package and a Hornby agenda package, both now 404 at the Trust. The ledger now has the other half: a sweep that walks the documents we hold and opens a record for any source URL that has since died, using the same test as the repair check, so a deleted file cannot hide behind the Trust's habit of answering with a 200-status wrapper page. View
• Cortes and Quadra zoning citations pointed at a moved domain 2026-08-26
  The Strathcona Regional District moved from strathconard.ca to srd.ca. Every zoning value we hold for Cortes and Quadra — 2,503 lots — cited the old address, which no longer resolves. All of them now cite the district's live land-use page. The zoning data itself is unchanged and still comes from the same Strathcona bylaws; only the address moved.
• Link rot swept out of the citations 2026-08-26
  A full check of every outbound source across the network found six links that had gone dead because the site on the other end reorganised, and all six are now repaired and re-verified. The Province moved its Islands Trust overview, which is the source for the 5,200 km² land-and-water scale on the water study; the Islands Trust renamed its annual-report and financial-reports pages, which carry the Conservancy figures; the Trust also moved the Hornby Island Official Community Plan; the Land Title and Survey Authority moved its title-search page, which is linked from every lot page; Lasqueti's community site moved its ferry page; and the Strathcona Regional District moved to a new domain. In every case the underlying facts were re-checked against the relocated source and none of them changed — only the addresses did.
• Correction: a fairness receipt pointed at a page the Trust had deleted 2026-08-26
  The fairness page cited the Islands Trust's Bylaw Enforcement Review Project page for the statement that the Trust approved revised enforcement policies in March 2026. We recorded that link as verified on 24 August 2026; by 26 August the Trust had deleted the page and it returned a 404, so a published claim was resting on a source a reader could not open. The claim itself was correct and is unchanged: it now cites the adopted policy instead — Islands Trust Policy 5.5.1, Bylaw Compliance and Enforcement, whose own header records Trust Council approval with an amendment date of 11 March 2026. Because the policy document carries the date, the citation no longer depends on the Trust keeping a landing page alive. The ocean page's federal-role receipt has also gained a second source, the Pacific Aquaculture Regulations, after the DFO regional site failed to resolve twice during the same check. Every source on all eight argument pages is now checked for liveness by a script (npm run check:campaign-links), which is how this was found and how it will be found again. View
• The water study now shows the full marine scale 2026-08-26
  The water study now leads with the official 5,200 km² land-and-water scale and explains how Islands Trust zoning overlaps with provincial Crown tenure, federal fisheries and aquaculture licensing, navigation review and shellfish sanitation. Its new sourced Salish Sea infographic can be shared or downloaded uncropped in portrait, square, story and full-resolution formats, with a complete accessible text edition on the page. The campaign's argument pages and share cards now use the same navy evidence-field, fair-counterpoint and real-question visual system. View
• The democracy page now shows who decides what 2026-08-24
  The democracy case now distinguishes decisions a Local Trust Committee can conclude from the controls that remain Trust-wide. It shows the local two-trustee majority; local permit, resolution and enforcement powers; the Executive Committee approval route for every local bylaw; and Trust Council control of budget, policy, procedures and staff. It also says plainly that local trustees are not powerless, the appointed member cannot defeat two participating local trustees who agree, and the bylaw approval chain does not apply to every permit. View
• Trust-wide evidence first; island evidence filed by island 2026-08-24
  The Abolish homepage and case index now separate institution-level evidence from local case files. The Trust-wide case comes first; local disputes are labeled and grouped under the island they document. The farming article now leads with all twelve measured local trust areas instead of a Denman worked example. The water study no longer defaults to Denman, and its three demands apply to every controlling boundary and zoning record. Detailed Denman material remains available as a clearly labeled Denman case study. View
• Eight arguments, eight evidence dockets 2026-08-24
  The abolition campaign now opens with the democratic-accountability case and carries eight separate evidence dockets: democracy, farming, ocean jurisdiction, housing, fairness, older Islanders and accessibility, public spending, and families. Each page separates the political verdict from the facts that support it, states what the evidence does not prove, links primary sources with pinpoints, and generates its own sourced 1080-by-1350 share image. We also retired legacy posters whose motive, causation, age, map, or accessibility wording could not survive a hostile fact-check; corrected the FOGI point number and the case-law tally; and kept the full water-map study beneath the new divided-jurisdiction argument. View
• A network-wide accuracy and publishing sweep 2026-08-24
  We corrected host canonicals and sitemaps, strengthened no-JavaScript campaign pages, fixed pre-existing map, media-gallery, crawler and route type errors, and regenerated the current social and poster artwork from deterministic source files.
• Every argument can now travel as a sourced image 2026-08-23
  Every petition homepage, substantial argument section, and public post now has a top-right share control. It opens a 1080-by-1350 image made from that exact page heading, excerpt, and deep link. On a phone or supported desktop browser, native share sends the PNG and link to installed apps. The same panel covers Facebook, WhatsApp, Bluesky, Threads, X, Reddit, LinkedIn, Telegram, Pinterest, email, and SMS, with copy-image, copy-caption, PNG download, and print for everywhere else. The four property families keep distinct art direction. Their illustrated backgrounds make no factual claim: every word and source destination is drawn from the live page in deterministic code. View
• Exact petition versions, corrected claims, and ads paused 2026-08-23
  The island-town petition now says exactly what it asks and what it does not: a town with a council residents elect and can replace; current law requires an island municipality while the Trust area exists; abolition is a separate decision. Earlier signatures have not been rewritten or silently moved. They remain attached to version 1 unless their owner reads and explicitly accepts version 2 from the private management page. Public totals now separate current-version signatures from earlier-version supporters, current PDFs include version 2 only, and the full wording history is public. A one-time correction notice shows both texts, contains no cross-petition pitch, and says plainly that doing nothing preserves version 1. The abolition petition demand itself is unchanged. We also retired an 86% water creative after finding that it mixed different land and area geographies; parcel-residual estimates are now labelled with their road and unsurveyed-land caveat. Marine zoning is valid, reference maps and GIS exist, and the narrower demand is to publish the controlling records online. Campaign ad routes are paused while election-period registration and authorization requirements are resolved.
• We went looking for the court cases. They mostly go against us. 2026-08-23
  There is a new page carrying every court decision we could find on the Islands Trust, and the short version is that the Trust nearly always wins. Of the ten decisions we counted it won eight, lost one and split one — and both times a court went against it at trial, the result was reversed on appeal. We read eight of them in full; two are old enough that we have them only through later judgments quoting them, and one we could not find at all. The page says which is which rather than blurring them together. We are publishing all of this because we spend a lot of time asking the Trust to disclose things it would rather not, and we do not get to keep our own awkward evidence in a drawer. One finding matters more than the rest. Our water study asks how far out to sea the Trust's rules reach, and until now our own notes marked it an open question whether it had authority over water at all. It is not open. A local government in British Columbia can zone the foreshore and the water above it, and the two leading cases saying so are both Islands Trust cases, both upheld on appeal — one about a floating marina office on a leased water lot at Ganges, one about a shed at Port Washington sitting half in the sea. Anyone who tells you these marine bylaws are invalid is wrong, and we have corrected our own pages to say so. None of the ten decisions counted in this review addresses the question the water study asks. In each, the argument was whether a bylaw reached one particular building or activity. None cites the regulation that is supposed to contain the Trust's outer boundary. These decisions therefore do not approve or reject that boundary. In the one case where a judge did have to work out where a water zone line ran, he looked at the zoning maps and found they did not match the coastline. The page quotes the judgments, links each one to the court's own published text, and says plainly where we have not read something rather than implying we have. View
• We broke the petition for several hours. Here is exactly what happened. 2026-08-23
  On the morning of 23 August, every attempt to sign the petition failed. The form returned an error and, worse, printed the database's own error text at whoever was trying to sign. It lasted several hours and we found it because the signature count stopped moving, not because anything alerted us. The cause was ours: a change adding comment screening passed an empty list into a database query in a way that produced invalid SQL, and because almost every signature has no comment attached, it failed for everybody rather than for a rare case. It is fixed, and confirmed working on phones and desktop. Two things follow that matter more than the bug. If you tried to sign during that window, your signature was never recorded — nothing was stored, so we cannot even email you to apologise. Please sign again; it takes a minute. And every release from now on is gated on a check that fills in the real form on the live site and confirms a signature can actually be created, at three screen sizes. Nothing we had — the type checker, the build, 234 tests, the accessibility and mobile audits — could catch this, because a page looks identical whether or not the thing behind it works.
• Maps you can move, a mistyped-email catch, and the other side's petition 2026-08-23
  The water study now lets a reader choose any local trust area and inspect its provincial boundary on a movable street map. It does not default to one island and it does not invent marine-zone lines. The sourced chain spans eight bylaws using sea-surface or air-space language, Galiano and Saturna's boundary clauses, Salt Spring's Vancouver Island carve-out, and the provincial Schedule A office-map provision. The global demand is to publish and reconcile every controlling boundary and zoning record; detailed single-island examples are filed under the island they govern. Also new: if you mistype your email domain when signing, the form now offers a correction. Somebody signed with an address ending .comm, which means their confirmation could never arrive and no reminder will ever reach them — from their side they signed, and we have no way to tell them otherwise. It suggests and never corrects, because plenty of real domains look like near-misses of the big providers. And there is a new section carrying the Friends of the Gulf Islands petition: their case in their own words, all six of their asks quoted in full, with a link. We disagree with it and we would rather you read it than take our summary of it.
• Rebuilt for the phone, and a way to tell people you know 2026-08-22
  Most people read this network on a phone, so it has been measured on one. Every button, link and form field is now at least 44 pixels tall — the size a thumb actually hits — and no text field renders under 16 pixels, which is the point where iPhones zoom the page in and do not zoom back out. The campaign material page no longer scrolls sideways: the ad previews were rendering at their true widths, up to 970 pixels, and pushed a 320-pixel screen more than 400 pixels off to the right. They now scale to fit at full fidelity. Three of these faults were invisible on a desktop browser and none of them broke the build, so there is now a script that loads each page at phone widths and measures what rendered, and tests that fail if the old traps come back. Separately, after you sign the petition you are offered — once, not pushed — the option to tell people you know. It sends from your own email, everyone goes in BCC, and the addresses never reach our servers: nothing to keep, because we never receive them.
• Every island now shows its census population 2026-08-21
  Until today this network held no population data at all. Each island site now states how many people Statistics Canada counted there in 2021, how that changed since 2016, and how many of its dwellings are nobody's home — each figure with its census year and a link to the StatCan table it came from. Two things worth knowing about the numbers: they cover the Trust Area designated place rather than the island's shoreline, and Statistics Canada publishes no age breakdown at that geography, so anything we say about the adult population is an estimate and is labelled as one.
• Salt Spring Island: the deep dive 2026-08-21
  A long, fully footnoted piece on the largest unincorporated community in British Columbia — bigger than 107 of the province's 161 municipalities, and twice a rejector of incorporation. The centrepiece is a 2015 provincial analysis, now surviving only in a web archive, showing that Salt Spring's property taxes fund roughly a quarter of the Islands Trust, and that if it ever stopped, a benchmark household on every other island would pay about $62 a year more to keep what the Trust does. Two of the article's findings run against our own argument, and are published anyway. View
• You can now choose how often we email you 2026-08-21
  Sign a petition and you can pick a weekly digest, daily, every two weeks, monthly, or no email at all — and change it any time from the private link we send you. The digest covers what changed on the sites, what turned up in the public record, and how the petitions moved, with every figure sourced. If you signed before this existed we will write once to ask what you want; if you ignore that email, you stay off the list. Unsubscribing is one click and never removes your name from the petition.
• Release notes now live on every island site 2026-08-21
  Until today only DenmanIsland.co published a running list of what changed. Every site in the network now has its own Release Notes page, linked from the footer, showing the changes that affected that site. It is the same standing rule we already followed — anything a visitor can notice gets written down — just published where the visitors are. View
• Corrected the land-area figures on the island sites 2026-08-21
  Our parcel totals were counting statutory rights-of-way — leases, licences of occupation and similar interests that are registered over land rather than being parcels of it — as if they were lots. Because they overlap the lots beneath them, they inflated every hectare total, badly in places: Galiano was showing 23,229 hectares for an island of roughly 5,700. They are now excluded from every total. We have also relabelled the figure as mapped parcel area rather than land area, because the parcel fabric covers the whole Local Trust Area including foreshore and marine parcels. Individual lot pages are unchanged — a right-of-way is a real registered interest and still shows on the lot it crosses.
• Zoning now cites its source on every island 2026-08-21
  Gambier and North Pender were showing each lot's zoning with no citation behind it, and 819 of Cortes's 820 lots were citing the wrong island's bylaw — the regional district publishes Cortes and Quadra zoning from one layer whose shapes reach across the channel between them. All three are corrected, so every zoned lot on the network now names the bylaw it came from.
• Water pages now say when the Province has not mapped an aquifer 2026-08-21
  Lasqueti and Texada have drilled wells on record but no provincially mapped aquifer at all. The water page used to say no record was "available yet", which read as something we had not got round to fetching. It now names the gap for what it is, shows the well count and the median depth, bedrock and yield the Province does record, and notes that none of the usual provincial ratings — vulnerability, productivity, demand, water quality — exist for these islands. View
• Every island site now shows who has signed 2026-08-20
  Each island-town site has a public signatures page listing the people who chose to show their name, with the date they signed and any comment they left. Signatures marked private are counted but never named, and email addresses are never published. Islands with no signatures yet publish that honestly too, with an invitation to be the first. View
• Per-island water reports 2026-08-20
  Every island site now has a full water report at /water: each aquifer the Province maps beneath the island, its vulnerability, productivity and demand ratings, wells on record, water licences, and the provincial observation wells tracking long-term levels — plus median well depth and yield from wells actually drilled on that island's lots. Gabriola and Salt Spring lead with written analysis; every other island gets the same data. View
• Lost your petition link? Email it to yourself 2026-08-20
  Every petition form now has "Already signed? Email me my sign-in link". Enter the address you signed with and the private link comes straight back to your inbox — no account, no password. From that link you can switch your name between public and private, edit your comment, or withdraw entirely. View
• Confirm a signature, then share it — and sign both petitions in one click 2026-08-20
  Confirming a petition signature now lands on a share page: your name in a signature card, ready-made posts for X, Facebook, WhatsApp, Reddit and email, a downloadable share image, and QR codes for flyers. Signed the abolish petition? One click also adds you to your island's own become-a-town petition (and the other way around) — no second confirmation email, because your address was already verified. View
• Petition signatures are now email-confirmed 2026-08-19
  Signing the Trust-wide abolition petition or any separate island petition now works in two steps: add your name, then click the confirmation link we email you. Your signature counts the moment you confirm. Every signature made before this change stays counted; nothing was removed. Why the change: it stops anyone else from signing or altering a signature using your email address, and it makes every published count one a skeptic can trust. Each island site now also sends its mail from its own domain. View
• Signature ledger: look up public names across every petition 2026-08-19
  abolishislandstrust.org now has a central signature ledger. It totals the Trust-wide abolition petition and every separate island petition, and lets you look up names people chose to publish. Private signatures stay in the counts only. View
• Camp Denman / Elkhaven interest disclosed; portfolio map published 2026-08-19
  Tony Aly's involvement with Camp Denman (also known as Elkhaven / Elk Haven) is stated on the ownership and transparency pages, the directory, and the Elkhaven archive. The live portfolio — landintelligence.co, denmanisland.co, denmantown.org, abolishislandstrust.org, the island-town sites, and iloveislandstrust.org — is described as a private network, not a federation of local chapters. View