olivierlacan

FYI I’m not convinced this is easier than just letting folks find their region via GPS, postal code or city name but it’s now implemented at https://indigene.app/map for folks who prefer visually pickings a region or looking at worldwide coverage.

I don't think we disagree but none of what you describe is what Indigene does: it's not doing art with AI. You're mixing the development tools with the actual mechanism of the system, which in no way uses LLM output for suggestions presented to users. Indigene doesn't design gardens and I'm not sure I ever want it to. It's a spot and condition-specific tool that lets unfamiliar people discover the most ecosystem-positive specimens they could plant. Sure, it helps focus your efforts, but it's not a whole garden design tool. Far from it.

Landscape designers who are familiar with natives don't cost "a few bucks". I've hired one and not even in a particularly expensive area (Central Florida) just a year ago. So I think we have a different target audience in mind. If you're ready to spend hundreds or thousands of dollars on native landscape design assistance, Indigene is unlikely to be the thing that pushes you to action. You've already been convinced.

On the other hand, encouraging people to iteratively rewild home gardens with natives with the the Doug Tallamy "Homegrownt National Park" (see https://indigene.app/homegrown) frame of mind isn't something that's hinged on whether a garden would earn landscape design trophies. A few native plants can already have an outsized impact on the native wildlife ecosystem in a given area.

I'd second your pitch for befriending folks that are also into native plants (or any plants) however. That can start with planting a handful yourself and offering seeds or cuttings to neighbors too.

It's in as of 1.1 yesterday, really good suggestion thanks! https://indigene.app/release-notes/1.1/

Curious what (broad) region you're in if you don't mind sharing. Just added Scottish Lowlands yesterday via 1.1 (https://indigene.app/release-notes/1.1/), but I need to tackle some Asian or African regions at some point as well. There's obviously a lot of North & South America to tackle as well...

Thank you for the feedback, I'd love to see your thing and what angle you're focusing on. I see what you mean, but I've never trusted Claude to do analysis. And I'm still only trusting it to help me build acceptance criteria that I can validate against a multitude of region-specific sources for different data points:

- host plants (e.g. European caterpillar counts via Gaytán et al: https://onlinelibrary.wiley.com/doi/10.1002/ece3.73004)

- food sources

- pollen interactions

- broadly documented wildlife & plant dependencies

- etc

What is your skepticism based on? If you've identified a gap I'd be very obviously interested to close it. But if you think I'm allowing Claude to editorialize without verification, then I can assure you I'm not. I've definitely caught it trying to skip steps and "leave out" verification steps, especially for new region coverage where we have insufficient data for a clear cut boolean answer.

And like everything, I keep expanding the rings of what I cross-reference for each region. Hopefully the more keen eyes on this project will mean folks can spot fallacies and holes. But to be frank, I'm in this for the long run so I very mulch welcome it :-)

PS: https://github.com/olivierlacan/indigene/blob/main/DATA_SOUR... might be a good one to focus on if you hadn't picked it up yet

Should be addressed soon: https://github.com/olivierlacan/indigene/pull/262

Although the entire point of the landing page spot finder flow is not to have to browse anything, regions are there for you to explore everything, not just where you are.

Was that not clear when you landed on the homepage? You can use your device location, zip code or town name to find natives specific to your location that way. It should find the region without you have to parse anything.

I'm so glad! Anything you found to be missing? Or that could help you with the new plants you haven't tried?

Your end goal focus is fine and I'm glad it improves your quality of life. Same here whether I focus on edibles or natives in my own garden. But I'd caution you about the "throw random seeds" approach, unless you've sourced those seeds and aren't using retail seed package "for pollinators". Those are notorious for frequently including invasive species.

A few examples among many similar studies:

https://neobiota.pensoft.net/article_preview.php?id=118480

https://www.washington.edu/news/2002/04/18/wildflower-seed-m...

https://www.dtnpf.com/agriculture/web/ag/crops/article/2016/...

https://wssa.net/2017/08/make-sure-your-seed-mix-is-tested-f...

https://pmc.ncbi.nlm.nih.gov/articles/PMC8176916/

"See what sticks" is inherently biased toward fast spreaders that will outcompete and outspread other plants regardless of their ecosystem services. Even if they're natives.

But like the parent comment, I can't tell you that your own enjoyment isn't warranted in your own garden. The whole point of Indigene is to try and contribute to tilting the balance a bit more into being more aware and focused on the ecosystem benefits of what you plant or don't plant.

You might have missed filters on the result page that specifically let you exclude plants that spread (width or height), thorns, and even stingers. Those can help you avoid what you describe, but they may not be obvious. Maybe I need to make those more obvious? Or default to them?

I think it's fine to have a human-centered mindset although the point of Indigene is to challenge that a bit and re-center the focus on the ecosystem services these plant provide. Ornamental and recreational qualities of a plant don't have zero value, but they definitely don't have 10 times the value of its ecosystem services.

But like everything, since the site is for humans, I do need to find a balance there. So point taken.

PS: swiftcoder and pavel_lishin had similar feedback about garden goals although with a different focus

Sure. But this same argument goes for every open source project ever built. At some point some convergence is in our best interest one some things, or we all burn tokens recreating the same thing until the planet is mostly data centers.

You'll note I'm not reinventing things, most of this project is connecting dots between open science and already amazing projects like iNaturalist. But if you're familiar with it, you might also be frustrated with its extremely simplistic approach to identifying natives.

The reason I suspect is why your last point exists, except I've spent an inordinate amount of time and effort ensuring that Indigene doesn't claim a plant is native to a region (vs. a whole continent) when it's not. That makes it less comprehensive than other resources, but unless you can point to specific issues it doesn't make it less reliable.

I can't make you care, but if you might this is where I'd look if I were skeptical:

https://indigene.app/sources

https://indigene.app/confidence

https://github.com/olivierlacan/indigene/blob/main/DATA_SOUR...

https://indigene.app/plants/penstemon-digitalis (example of a status conflict)

https://github.com/olivierlacan/indigene/blob/main/app/scrip... (which is what does this status checek)

And if you care to call me out on any GIS mistakes (given my own lack of expertise in that area) this where you're likely to find mistakes: https://github.com/olivierlacan/indigene/blob/main/app/scrip...

Thank you for that! Curious what feedback you get from her.

Yeah, that's not great. You missed "Don't want to use your location? Search by town or pick a region instead." because it's lower on the page (see edit below for fix).

I specifically limited map levels to reduce the app's size and that's a big downside.

I don't get your location, your browser does since everything is client-side by the reverse geocoding API definitely does for the map (and you can't take my word for it) so let me see if I can make this less annoying.

No point of trying to be privacy-conscious if the alternative is a pain in the ass, no argument there.

Edit: should be resolved via https://github.com/olivierlacan/indigene/pull/253

Recently had to fix a similar issue where a zipcode from Florida landed in Spain (https://github.com/olivierlacan/indigene/issues/219).

That's what I get for trying to limit myself to client-side APIs but I'll take your repro and try to address it. Here's the tracking issue if you're interested: https://github.com/olivierlacan/indigene/issues/251

Edit: should be addressed by this change if you'd like to give another try: https://github.com/olivierlacan/indigene/pull/252

Welp you're in luck because my friend Rikki Ocampos who's been helping with Indigene told me yesterday about CISMAs: > CISMAs are organizations for different management areas, they’re part of the Florida Invasive Species Council.

https://www.floridainvasives.org/for-homeowners/

I told her guidance on "unplanting" invasives is definitely something I want to tackle but it can be tricky to navigate so the first update is this PR (https://github.com/olivierlacan/indigene/pull/247) providing better guidance on whether, when, how to "unplant" invasives and whose permission you might need.

Definitely curious about any and all invasive management resources so thanks fo the https://woodyinvasives.org link, I don't think I was aware of that one yet!

Once you got to results you mean? I just made the numbered steps at the top of the spot result pages clickable so you can use them to navigate back to soil but that may not be obvious. So thanks for the feedback!

This one deserves a little build-focused post on the blog I referenced in the OP. But in summary, it was a frankly shocking roughly 60-minute session with a surprisingly narrow prompt in Claude Code.

I didn't use a cloud environment which is what I typically do on this project. I set it up to use OpenRouter to delegate to appropriate voice model for the narration (which still blows my mind, especially the French version as a native speaker). That cost barely $2 of credits.

The art direction was simply to remain coherent with both the tone and voice of the project but the visual identity and existing design and illustration elements.

All told I think the final cost was around $30 (at API rates) for the both the US English version and the French version. It was much cheaper to French, because it produced reusable HTML, CSS, and JS assets to animate the scenes in the video.

As someone who's done a lot of tedious video editing work and (lousy) motion graphics in the past for some educational content, it both terrified me (for all the hard-working folks who do this for a living) and made me realize how accessible producing this kind of video content is. It's a bit more palatable given the focus of this project, but I know many content producers, video editors & audio engineers (even voice actors!) who likely wouldn't find this exciting at all.

Given this project has no profit motive, I haven't paid artists (or programmers to make this hit home) to contribute to the project but if people find Indigene helpful (outside of me and a few friends) it's something I'd much rather do. I've done that in the past for Keep a Changelog for instance.

Thanks! There's actually a "Soak up rain" and "Stop erosion" goal for the spot-centered default flow to find natives, but it seems like your need is a bit more heatwave resiliency focused?

The alternatives (get it? https://indigene.app/alternatives) section focuses a bit more on the swapping non-natives or turf grasses that do poorly or require intense irrigation during heatwaves, but I really like the idea of adaptation planting.

Generally most natives are inherently better suited for heat waves and warming/drying summers since they tend to need little to no watering once established, but I agree it'd be nice to get a subset of plants that can create shade and help create a bit of an oasis.

Who doesn't love a tightly scoped request like this. Would an anti-filter mode do the trick for you?

Click twice on a filter to flip it to exclude instead? I can noodle on it for sure.

Did you try the print stylesheet? I did some work on it a few months ago but I really like the idea of being able to print out instructions, especially for the propagation technique pages.

All excellent feedback thank you! I was just talking with a friend who's helped me test a few things about the wildlife support focus, which excites me too because I've been trying to see what particularly tickles Steller's jays. They show up at pretty specific points in the season but other than bird seeds I've never really noticed a specific plant tie.

And for now Indigene only has a group category for Jays, turkeys & woodpeckers: https://indigene.app/wildlife/acorn-birds

Ditto on that "good for wild turkeys" idea.

For deer hardiness, do you mean a way to find more deer hardy plants directly?

Love the idea for fast growers, especially to counter boxstore-bought yews and other non-natives people pick up because they want fast-growing shrubs for privacy hedges.

Bloom-timing is a cool one too, I already love the info I get from the propagation section (https://indigene.app/planting) on timing specific techniques and I added plant-specific propagation timing fairly recently (https://indigene.app/plants/symphoricarpos-albus#propagation), but having something equivalent for what to expect in a given season so you might have overlapping ecosystem support instead of one big bloom and nothing for months.

Thanks so much for all these great ideas!

That actually may be more due to the fact that I've tried to ensure we have sufficient cross-referencing before we claim things are native to a specific region.

Curious which regions you looked into.

That's true, good reminder to prioritize that. Thanks.

That's not accurate. H-1B is a non-immigrant visa. It's not intended for tourists.

From CBP's own website[1], emphasis is mine:

    An immigrant visa (IV) is issued to a foreign national wishing to live permanently in the United States.

    A nonimmigrant visa (NIV) is issued to a foreign national with permanent residence outside the United States but wishes to be in the United States on a temporary basis for tourism, medical treatment, *business,  temporary work, or study*, as examples. Note: There are more than 20 different categories of nonimmigrant visa classifications.

[1]: https://www.help.cbp.gov/s/article/Article-1027?language=en_...

Southern California is in. Took a bit longer than planned since as you'll see there were quite a few other improvements today: https://indigene.app/release-notes/0.24/

I'll move on to other parts soon, like the Bay Area.

Some are pretty crude but it was one of the hardest things to determine when dealing with a fairly small garden. I agree they’ve been very useful before and sometimes after planting when it becomes clear you might have to relocate a plant.

The next thing I’d like to do is categorize some of the plants that are ideal for foundation planting (water management, vertical taproots) or curb islands (between sidewalks & streets).

Fixed via https://github.com/olivierlacan/indigene/pull/105

Thanks again for the suggestion. It’ll take some refining but I also included nearby cities so it’s easier to get a sense of scale and proximity.

Fair criticism, this is based on EPA Level III/IV Ecoregions (cited in https://github.com/olivierlacan/indigene/blob/main/DATA_SOUR...) but I agree there should be a way to see it in a clear map.

Level III/IV Maps are here: https://www.epa.gov/eco-research/level-iii-and-iv-ecoregions...

At the very least I'll make sure each region page links to this given it's the source that determines what species are shown.

It’s so nice to know you’ve worked on this. I’m very curious what sources you used.

The most fun I had with Indigene was when I decided to add French ecoregions so my parents could use the app too. Most of the work is in this PR where I used a recently published study’s plant host species data to offer ecosystem value/services metrics since the Tallamy data doesn’t cover that: https://github.com/olivierlacan/indigene/pull/53

That stuff reignitated the kind of deep data hunt fever I haven’t had since my COVID Tracking Project days when I was finding random ArcGIS dashboards and extracting data from their JSON calls to find some reliable case data.

I’m bummed I didn’t find your project before. I kept looking for something when I moved from Florida to Washington after converting my entire yard there. Having to learn a whole new ecoregion was rough and it made me think the barrier should be drastically lower for people who aren’t as motivated.

On the financial opportunity side, I hoped for a while that there may be a way to align our best interest as a species (and our ecosystem) with what our modern societies value but the last decade has made it clear to me that neither governments nor private enterprise are capable to remain as focused as they need to be. Plus leveraging worldwide public-funded science to make a living seems a bit odd if I can instead use my engineering expertise to not require unnecessary waste of resources (servers, more LLM tokens, etc.) as the ecosystem rapidly degrades.

This is all the motivation I needed. I want to add more species to Florida first because this project finally came to life thanks to a visiting friend who still lives in Florida and struggles to find what to plant around her house.

Thank you, it’s very explicitly the result of a relatively tight but fast TDD development process with Opus (4.8 & 5 mostly), a lot of cross-referencing with established sources (see https://indigene.app/sources), and my own 30 years of amateur gardening experience and roughly 10 years of Florida & PNW native planting & observation experience.

So I take this feedback to heart and just like with the (sourced & generated) French localization I’m working my way through validating everything. This is another benefit of the project being entirely public, you can see precisely in the repo what I did and didn’t think to do in order to ensure accuracy.

When it comes to prose & style, I’ve set some clear accessibility goals in the agent config, and I’ve rewritten some things but it can be hard to fight the needless verbosity of LLMs at times.

I would deeply appreciate if you took the time to send a PR for any kind of improvement or addition). Thanks in advance.