I had similar problems for a while. Especially Safari tabs that were dumped from memory on iOS. Logging in again was annoying. A quite over-engineered solution that slowly evolved in my Kagi usage uses Tailscale Golink [0].
Case 1: Kagi extension for Safari redirects from Ecosia to Kagi when on the good path. I rely on this mostly because it’s first-party so there’s a bit more trust with my queries.
Case 2: Safari Redirect Extension [1] redirects unauthenticated Kagi pages to Golink. Golink redirects back to the original page with a Kagi Token appended. Instantly authenticated.
Also, I highly recommend setting custom DNS records on Ecosia/DDG/whoever-you-redirect-from to prevent query leakage.
% go run ./cmd/featuretags --min
ts_omit_ace,ts_omit_acme,ts_omit_advertiseexitnode,ts_omit_advertiseroutes,ts_omit_androidbin,ts_omit_androiddns,ts_omit_appconnectors,ts_omit_aws,ts_omit_bakedroots,ts_omit_bird,ts_omit_c2n,ts_omit_cachenetmap,ts_omit_captiveportal,ts_omit_capture,ts_omit_cliconndiag,ts_omit_clientmetrics,ts_omit_clientupdate,ts_omit_cloud,ts_omit_colorable,ts_omit_completion,ts_omit_completion_scripts,ts_omit_conn25,ts_omit_connreject,ts_omit_dbus,ts_omit_debug,ts_omit_debugeventbus,ts_omit_debugportmapper,ts_omit_desktop_sessions,ts_omit_dns,ts_omit_dnsresolvecache,ts_omit_doctor,ts_omit_drive,ts_omit_exitnodehealth,ts_omit_favorites,ts_omit_flashappliance,ts_omit_gro,ts_omit_health,ts_omit_hujsonconf,ts_omit_identityfederation,ts_omit_ipnbus,ts_omit_iptables,ts_omit_kube,ts_omit_linkspeed,ts_omit_linuxdnsfight,ts_omit_listenrawdisco,ts_omit_logtail,ts_omit_nattraversal,ts_omit_netlog,ts_omit_netstack,ts_omit_networkmanager,ts_omit_oauthkey,ts_omit_osrouter,ts_omit_outboundproxy,ts_omit_peerapiclient,ts_omit_peerapiserver,ts_omit_portlist,ts_omit_portmapper,ts_omit_posture,ts_omit_qrcodes,ts_omit_relayserver,ts_omit_remoteconfig,ts_omit_resolved,ts_omit_routecheck,ts_omit_runtimemetrics,ts_omit_sdnotify,ts_omit_serve,ts_omit_serviceclientprefs,ts_omit_ssh,ts_omit_synology,ts_omit_syslog,ts_omit_syspolicy,ts_omit_systray,ts_omit_taildrop,ts_omit_tailnetlock,ts_omit_tap,ts_omit_tpm,ts_omit_tundevstats,ts_omit_udptransport,ts_omit_unixsocketidentity,ts_omit_useexitnode,ts_omit_useproxy,ts_omit_usermetrics,ts_omit_useroutes,ts_omit_wakeonlan,ts_omit_webbrowser,ts_omit_webclient
The feature I don't want to pay for but would be nice to have: deeper level proxied wildcards.
> If you want to proxy a wildcard DNS record on a deeper level like .www.mycoolwebpage.xyz you can subscribe to Cloudflare Advanced Certificate Manager and get a certificate that is covering that wildcard ¹ ²
When building out services in my smart home, I started with something like 'home.domain.com'. I used Let's Encrypt for services like 'smart.home.domain.com' (and I think Cloudflare proxied these for free?) but realized I was losing some privacy since the subdomains were describing my network services. I also wasn't aware SSL certificates are a permanent record (though what isn't these days) and I had to be careful what names I was using.
I realized instead I could use Let's Encrypt wildcard feature '.home.domain.com' and at least limit the private details potentially leaked in certificate history. But because Cloudflare doesn't proxy deeper wildcard subdomains for free, I'd have to manually create each subdomain in Cloudflare DNS - defeating the privacy objective.
For now, instead of wildcards, I'm just using URL path prefixes. For example 'smart.home.domain.com' becomes 'home.domain.com/smart/'. This works pretty easily in a docker host with traefik handling domains, paths, and certificates. Some apps don't work under path prefixes without a lot of tweaks. And it works fine if it's one-to-one server to subdomain - but if I want to host another server at home I have to come up with another subdomain like 'app-home.domain.com'.
I could also move everything into something like tailscale. Then at least the domains are private. Right now I use an oauth2 proxy infront of my services - but most of them don't need direct internet access. The issue is, for the services that do need the convenience of direct access, tailscale funnel doesn't support custom domains.³ And even if eventually they do, maybe they'll also limit subdomain depth or wildcards.
Anyway for privacy reasons it'd be swell if deeper level subdomain certificates were part of the free tier.
And if any tailscale employees are reading this - https://github.com/tailscale/tailscale/issues/15724 please fix this too. Regular users not using some sort of enterprise saas DNS (whatever their thing is?) deserve DNS privacy too.
This would be more convincing if there were a simple stat on the home page of the dispositions of the reported bugs (fixed/no-repro/won't-fix).
As a random sample of one, I looked at one of the bugs this reported on Tailscale (first thing on the homepage) [0], and the pull request ends with "Apologies for the lack of due diligence here. I'll go ahead and close this out."
Peers might not want to use third party rendezvous servers or relay servers
It appears that Tailscale attempts to persuade peers not to run own "DERP" servers (cf. encouraging peers to run own servers)
"In general, you should not need to or want to run this code. The overwhelming majority of Tailscale users (both individuals and companies) do not."
"Running your own DERP services requires exeprtise [sic] in multi-layer network and application diagnostics. As the DERP runs multiple protocols at multiple layers and is not a regular HTTP(s) server you will need expertise in correlative analysis to diagnose the most tricky problems. There is no "plain text" or "open" mode of operation for DERP."
2. (Optional, but you really should..) At least one discovery node with a routable IP address, which we call a lighthouse."
The comparison to nc seems off. Original netcat has no required third party dependencies, no recommended third party dependencies, and not enough complexity to be linked to a company selling associated SaaS or consulting