$ONyc : OnRe's sitemap sends Google to 7 blog posts that return 404, including "The truth about OnRe"

I build websites, so I crawled all 103 URLs in onre.finance/sitemap.xml and checked status codes, titles, descriptions and preview images.

Most of the site is in good shape. The homepage and the main pages have proper titles, descriptions and preview images, every page has a canonical tag, and nothing is accidentally set to noindex.

What I found:

1) Seven blog URLs in the sitemap return "Page Not Found":

/blog/the-truth-about-onre

/blog/choosing-an-onyc-yield-strategy

/blog/onycs-best-use-case

/blog/the-onre-market-on-kamino-one-year-of-growth

/blog/understanding-solanas-onchain-yield-coin

/blog/where-models-end-and-underwriting-begins

/blog/from-yield-chasing-to-collateral-grade-assets

Google keeps hitting these from the sitemap, and anyone who saved or shared one of those links lands on a dead page. For a yield product, a missing post called "The truth about OnRe" looks worse than it probably is.

2) /resources/style-system is public and listed in the sitemap. It looks like Webflow's internal style guide page, with no description, and it shouldn't be indexed.

3) The five blog category pages (/blog-category/announcements, ecosystem, education, insights, partnerships) are all titled just "OnRe", with no description and no preview image. In search results and link previews they are impossible to tell apart.

4) The three legal pages spell the brand "On Re" in their titles ("Privacy Policy | On Re"), while the rest of the site uses "OnRe".

5) The "ONyc Launches on Kamino" post has 4 H1 headings and the Loopscale vault post has 2. Every other page has one.

Suggestions: republish the seven posts or 301 them to the closest live article, and if the Webflow sitemap is auto-generated, it should pick up the change after the next publish. Turn off "Include in sitemap" and indexing for the style guide page. Add a title and description template to the blog category collection, for example "Announcements | OnRe". Fix the "On Re" spelling and use one H1 per post.

Question: were these seven posts unpublished on purpose? If any of them were replaced by newer ones, pointing the old URLs at the new posts would keep their links working.

$ROAM every page on weroam.xyz tells Google the real version lives on a noindexed Framer staging URL

I build websites, so I checked weroam.xyz the way I check a client's site before launch: page source, canonical tags, sitemap and links.

The basics are in place. robots.txt allows crawling, the sitemap lists 285 weroam.xyz URLs, and the main pages have their own titles and descriptions.

What I found:

1) The canonical tag on every page I checked points to the Framer staging domain, not to weroam.xyz. On the homepage the canonical is https://youthful-three-407982.framer.app/, and /use/global-esim-mobile, /use/use-earn, /resources/faq, /learn/what-is-roam-network and the August 2026 round-up do the same with their own paths. That staging domain is still live and returns "X-Robots-Tag: noindex". So the sitemap says "index weroam.xyz", while each page says "the original is a URL you must not index". Google has to guess which signal to trust, and that can cost the eSIM and Use & Earn pages their search traffic.

2) On tablet, the "Our Communities" card links to https://weroam.xyz/join_us.html, which returns Framer's 404 page. The desktop version of the same card links to /join_us and works.

3) The site has a custom script that should send mixed-case URLs to lowercase ones. It has a typo: "window.loction.search" instead of "window.location.search", so it would throw before redirecting. It also lives only on working pages, while Framer serves mixed-case URLs as a 404 that doesn't include the script. In practice /Resources/FAQ shows "Page Not Found" and /resources/faq works. The same script tries to rescue join_us.html, and fails for the same reason.

4) /join_us and /footer/about reuse the homepage title "Roam - Web3 DePIN WiFi and eSIM Network".

Suggestions: in Framer, set weroam.xyz as the primary domain so the canonicals switch over (this usually happens when the custom domain is not set as primary). Point the tablet card to /join_us. Add real redirects in Framer's Redirects settings for join_us.html and mixed-case paths instead of the head script. Give About and Join Us their own titles.

Question: is youthful-three-407982.framer.app meant to stay public? If it is only a staging copy, password-protecting it would remove the duplicate site.

$PIPE : the free Hosting plan is good, but changing one CSS file still means installing Rust

I make websites, so I looked at Pipe Hosting from the static-site side. I read the Hosting guide and the pricing page, downloaded both starter ZIPs (pipe.network/downloads/pipe-host-starter.zip and pipe-host-starter-source.zip) and went through the files.

The Free plan is more than I expected: 2 sites, 100k dynamic requests, 5 GB transfer and 100 MB SQLite, with no auto-renewal and no overage charges. I also liked expected_active in the CI flow, which stops two deploys from overwriting each other.

Where I got stuck:

1) Every frontend change goes through a Rust build. build.rs compiles html/ into app.wasm, and "pipe.py bundle" runs cargo before anything else. To edit index.html or styles.css I would need Rust 1.95.0 and the wasm32-wasip1 target installed, which is a lot to ask from someone with a plain HTML site.

2) The README in the source ZIP says to deploy with "the host's scripts/deploy.py" and mentions "prebuilt/pipe-app.zip". I couldn't find either file in pipe-host-starter-source.zip. pipe.py also says "no remote Pipe deployment API is assumed", while the guide documents that API.

3) The guide says a rollback changes the code and leaves the database as it is. It doesn't show how to do a rollback, from the UI or the API.

4) The HTML of /pricing ships "—" in place of every price, and JavaScript fills the numbers in later. Link previews, crawlers and AI agents that don't run JS see no prices.

What would help: a "pipe.py bundle --static" option, or reusing the prebuilt app.wasm, so HTML/CSS edits don't need Rust. A README that matches the hosting guide, a one-line rollback example, and plan prices rendered on the server.

One question: the pricing page bills everything in prepaid USDC credit. Where does the PIPE token come in for a Hosting customer today? Is burn-to-credit live yet, or still planned?