The stairway is gone, the printouts aren't

A stack of paper and a ghost#

Imagine the only photo of your great-grandmother is a photocopy of a photocopy, and the studio that took it burned down. You'd treasure it. You'd also squint at it a lot and ask: how do I know this is really her, and did anyone write on the back?

That's where we are with Heaven's Stairway. We were handed 18 A4 printouts, scanned in, of a long-gone online seed shop's catalogue pages. They cover 42 strains from four suppliers, with prices, flowering times and little shop codes like "ws" and "P005". As far as we know, it's the only copy of those pages. Some corners are smudged. A few lines have faded half out. On the Mr Nice set, the shop code for Red Devil has faded off the page entirely.

The shop itself is a matter of public record. Heaven's Stairway was a Montreal seed company that began importing seeds in 1998 and, according to Wikipedia, supplied around 300 strains worldwide [1]. In late January 2006 police raided it. Six websites went dark, and about 200,000 seeds, US$183,000 and three gold bricks were seized. Seven people were charged; the news reports we read say nothing about how the cases ended [1][2]. We couldn't find the shop's own site anywhere online.

So this is a piece about a quiet problem in cannabis history: it's weirdly hard to find the original, accurate source for almost anything. The thing you want was often on a website. Websites die.

What the pages actually are#

Each set opens with a supplier's logo, then runs strain by strain: a blurb, then a "Specifications" line with flowering time, price per pack and a shop code. The suppliers are the Mr Nice pages (which carry a lot of Greenhouse Seed Company copy: the White Rhino blurb says so outright), the BCGA, the Brothers Grimm, and Paradise Seeds.

Notice what this is. It isn't the breeders' own words, and it isn't a lab result. It's a shop quoting breeders. That's a perfectly useful witness, but a witness is not the original document. Think of a postcard from a holiday: great evidence somebody went, weak evidence about the weather.

The pages are full of the sort of detail that disappears first: a 40-order limited edition of Beauty & the Beast, "only enough seeds for 100 orders" of Killer Queen, a Brothers Grimm "In/Out Mix" the breeders confess got "mixed up on the sorting table" so "now we don't [know] which strains they are". The Come Back Early entry carries a note that "Neville's Threeway has now been given it full name after meeting with Neville himself, sorry for any confusion". Nobody's databases keep that.

The priciest packs are the Brothers Grimm headliners: Cinderella 99, Rosetta Stone, Apollo 11, Ice Princess and Mantanuska Thunder are all $150 for ten seeds, which is $15 a seed. The cheapest are the mixes at $20 and the BCGA Blueberry at $30. A reasonable guess is that the shop was pricing what was hard to get, but the pages don't say.

Checking the claims against the public record#

We went through the printouts' boldest claims and tried to match each against something that's still around: Seedfinder's cup history, Wikipedia's Cannabis Cup list and the breeders' own later write-ups. Here's the scorecard.

Notice what the scorecard does and doesn't show. The cup claims check out, and that's worth something: the shop wasn't making things up about trophies. But the "check" is mostly other people repeating the breeder. The more times a claim is copied, the more it feels confirmed, and the less it is. That's the real trap, and it's the whole subject of our award-check piece.

The graveyard is bigger than one shop#

Heaven's Stairway isn't alone. A short, fact-checked tour of things that used to be there:

Overgrow.com. A cannabis grow forum that went online in 1999. Per Wikipedia, at its peak it held more than six million posts from more than 133,000 registered members, and it was among the six sites taken down in the 2006 raid. It's back as a forum since 2016, according to Cannigma, but the old threads are a different story [1][3].

The Open Cannabis Project. A nonprofit that aimed to keep the cannabis genome out of patent trolls' hands. It spun off from Phylos Bioscience in late 2017 and dissolved in May 2019 after a backlash over Phylos's breeding plans [10][11]. The news pieces we found don't say what happened to its data.

A door that shut. In February 2016, Leafly cut off outside developers' access to its strain data, and people building on it had to find another source overnight [12]. The data didn't vanish. The way in did.

And a counter-example: Seedfinder, the strain database that still answers when you look for a 1998 cup winner. It's the reason most of our scorecard above could be checked at all. Long-lived sources are rare, and the rest of us live off them.

Not everything stays lost#

Here's the hopeful bit. Some records that disappeared haven't stayed disappeared. A handful of lost records have been recovered in full, and they're now on CannaLib, where you can search them like anything else. We won't bore you with how. Think of a library that quietly gets a burned book back, page for page, and puts it on the shelf.

It's also why we keep saying "as far as we know". The Heaven's Stairway scans are the only copy we have. If you have a better one, or a different page, we'd love to see it.

What to do with all this#

Three small habits that make you hard to fool:

Ask who said it first. A shop page, a strain database and a magazine can all repeat one breeder's line. Count the breeders, not the pages.

Look for the date. The best thing on these scans is a line like "As of May 2000, these are hot off the press". It tells you exactly when the claims were made, and by whom.

Save things. Print the page, screenshot it, scan it. In 2026, a stack of printouts is the last witness standing, because someone in 2000 pressed Print.