mrambossekandClaude Fable 5
a7dccf7da2
Letting the kernel fragment an oversized datagram answers one question — do
fragments get through. It cannot answer the more interesting one, because the
kernel always emits them in order, first one first.
The classic middlebox fault is exactly about that ordering. Only the first
fragment carries the UDP header, and therefore the ports; a stateful firewall
or NAT that has not seen it has no flow to match the rest against, and many
drop them. That is invisible to any in-order test and shows up in the field as
"large DNS answers fail on this network" or "the tunnel breaks when the MTU
drops" — it works until the network reorders, then fails intermittently, which
is the hardest kind of fault to chase.
So the server now builds the fragments itself (raw socket, IP_HDRINCL) and
controls their order: in_order as a baseline, reversed, and first-fragment-last.
The datagram is assembled and signed whole before being cut up, so what the
client reassembles is indistinguishable from an ordinary packet — otherwise it
would be measuring our sender rather than the path.
Two details that would silently produce wrong answers:
- The UDP checksum is computed rather than left zero. A zero-checksum datagram
is dropped by some middleboxes, and that drop would be recorded as a
fragmentation failure, which is the wrong conclusion entirely.
- Fragment offsets are in 8-byte units, so non-final fragments are rounded to
a multiple of 8. A 100-byte fragment is not an error, it is a datagram no
host will ever reassemble.
frag-send is advertised only when a raw socket can actually be opened — checked
by opening one, since a permission model has more ways to say no than a
capability bit has to say yes.
Fragment header arithmetic is unit-tested (reassembly coverage, MF flags, shared
IP ID, 8-byte offsets, checksum verification), cross-compiled and run on Linux
since the code is build-tagged.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Echolot
Free software for detecting and debugging local network issues from an Android phone — built for people who actually know what a neighbor table is.
Most "wifi analyzer" apps show you signal bars. Echolot aims at the layer where home and office networks actually break: duplicate DHCP servers, broken IPv6 RAs, MTU black holes, NAT64 weirdness, multicast that dies at the AP, DNS that answers differently than it should. It records what it observed, separates observation from interpretation, and exports the whole run so you can argue with it later.
Status: pre-release. The capability prober runs on real hardware; the production app and the probe server are not built yet.
Repository layout
docs/ design docs — the contract for everything below
echolot-prober/ capability prober: validates the no-root feasibility matrix on real devices
The Go probe server and the production app land here as siblings.
Design docs
The three specs are draft-complete and reviewed; treat them as the contract.
| Doc | What it defines |
|---|---|
| docs/feature-catalog-and-feasibility.md | Full feature list + the no-root feasibility matrix |
| docs/measurement-schema.md | Archived/exportable measurement JSON (observation vs finding, two-clock rule, anonymization) |
| docs/probe-protocol.md | Client↔server wire protocol (pinned TLS control plane, binary UDP data plane, STUN, canary DNS) |
| docs/build-status.md | Running log of decisions and next steps |
Privilege tiers
Every result records which tier produced it:
app— no root, no special setup. The bulk of the functionality.shizuku— ADB-shell privileges via wireless pairing, no root. Shipped in v1.root— future optional module.
Licensing
| Part | License | Why |
|---|---|---|
| All code (app, prober, server) | GPL-3.0-or-later | The value here is the platform-API research; copyleft keeps derivative apps free |
docs/ (the specs) |
CC-BY-4.0 | A wire protocol and a measurement format should be implementable by anyone, without license anxiety |
Full texts: LICENSE (GPLv3) and docs/LICENSE (CC BY 4.0).
Sources carry SPDX-License-Identifier headers.
If you want to build a compatible server or client, the protocol and schema docs are deliberately permissive — go ahead.
Building
See echolot-prober/README.md. Short version, from echolot-prober/:
echo "sdk.dir=/path/to/Android/sdk" > local.properties
./gradlew :app:assembleDebug