mu*index
reading
plain text
theme

The ecosystem

We publish the share, a ratio over the games we measured survives the ones we cannot reach.

Oldest handshake here: confirmed 4w ago.

Codebases

Of the 886 games listed, 610 told us what they run, and every share below is over those 610. A codebase we could not read is left out of the denominator

81 of 610 (13.3%) run a codebase no other listed game runs — one game each, which is assumed to be just that game. They are inside the denominator above:

Lineages

The same games, grouped by the tradition their server descends from — our reading of the codebase, not anything a game published. No game reports "MUSH": MSSP has no such value, and most of the MUSH world publishes no MSSP at all, so this is the only way the question can be asked. derived

107 of those games run codebases we do not place in any lineage — several say as much themselves, publishing FAMILY Custom. They are inside the denominator above and in nobody's share.

Protocols

Read every measured figure below as a floor. Nothing else here is requested, and a server may support a protocol without ever offering it.

MSSP is the one row below that is not a floor: we ask every server for it by name, so the games that did not offer it were asked and declined.

Protocol adoption. 'Measured' is what a server offered in a completed handshake; 'declared' is what its MSSP claims.
protocol measured — of 885 games whose handshake we completed declared — of 309 games whose MSSP report we hold
TLS not measured — never observed 53 of 309 (17.2%)
UTF-8 108 of 885 (12.2%) · 777 games neither offered nor asked 90 of 309 (29.1%)
GMCP 37 of 885 (4.2%) · 848 games neither offered nor asked 69 of 309 (22.3%)
MXP 122 of 885 (13.8%) · 763 games neither offered nor asked 96 of 309 (31.1%)
ANSI not measured — never observed 200 of 309 (64.7%)
ATCP not measured — never observed 12 of 309 (3.9%)
CHARSET 114 of 885 (12.9%) · 771 games neither offered nor asked 51 of 309 (16.5%)
EOR 45 of 885 (5.1%) · 840 games neither offered nor asked 1 of 309 (0.3%)
MCCP 216 of 885 (24.4%) · 669 games neither offered nor asked 80 of 309 (25.9%)
MCCP3 1 of 885 (0.1%) · 884 games neither offered nor asked 0 of 309 (0.0%)
MCP not measured — never observed 22 of 309 (7.1%)
MSDP 96 of 885 (10.8%) · 789 games neither offered nor asked 57 of 309 (18.4%)
MSP not measured — never observed 72 of 309 (23.3%)
MSSP 319 of 885 (36.0%) · 566 games declined when asked not asked — every report here is the answer
NAWS not measured — never observed 1 of 309 (0.3%)
PUEBLO not measured — never observed 18 of 309 (5.8%)
TTYPE not measured — never observed 1 of 309 (0.3%)
VT100 not measured — never observed 36 of 309 (11.7%)
XTERM 256 COLORS not measured — never observed 134 of 309 (43.4%)

Adoption over time

Each point is a share over the games we had measured that day, so this line moves for two reasons: a game changing what it offers, and the set of games we can measure changing around it. Only the first is adoption. The transition count below is the part that is purely games changing their minds.

Measured share of each protocol, oldest reading first
protocol then now
TLS not measured Aug 2026 not measured
UTF-8 12.0% Aug 2026 12.2%
GMCP 2.6% Aug 2026 4.2%
MXP not measured Aug 2026 13.8%

407 capability changes recorded so far — the material a curve is drawn from.