Some forks and siblings do try to backport e.g. security fixes (and also features) from Mozilla's mozilla-* and comm-* repositories.
What I can imagine totally happening is that, if they are going full GenAI, they'll somehow touch the entire codebase with it. They've done reformattings and similar changes in the past. Browsing through the mercurial history becomes more fun, and, more importantly, patches won't apply cleanly as the code went through such modifications. So, now? I can totally see GenAI touching all the files for reformatting so that you can't just backport the reformatting commits.
(And this without considering the whole mess that is JS and HTML features being defined in unversioned, non-stable and backwards-incompatible "living standards". (fetch() comes to mind as an example of this, the defaults were changed, and sites relying on the new defaults broke where fetch() was available with the old defaults...; I didn't check well enough to know whether postMessage()'s new features are an example of this or not (it did break reCAPTCHA in SeaMonkey < 2.53.23)))
njsg
npub1la…qh4f5
2026-03-07 09:12:24 UTC
in reply to nevent1q…4ula
Author Public Key
npub1laxpel7l93kdellz978n2z5xfrknhkqalck62eta84wy7xe654zswqh4f5Seen on
wss://relay.momostr.pinkPublished at
2026-03-07 09:12:24 UTCEvent JSON
{
"id": "3f9b5863c9bb6d8f8388785dd10d7814078529cf8fbcec0088e0465b151dadcd",
"pubkey": "ff4c1cffdf2c6cdcffe22f8f350a8648ed3bd81dfe2da5657d3d5c4f1b3aa545",
"created_at": 1772874744,
"kind": 1,
"tags": [
[
"p",
"0a4c13cc1a718ca63b19615ea4cf25156f43aa76c31f751e4f1cd12e3ae06f60"
],
[
"proxy",
"https://mementomori.social/@njsg/116187119231422622",
"web"
],
[
"e",
"f49f4029f7a99062ff2fd28a7a43296f9d5d83108c7bdf11d51412d5345ec0bb",
"",
"root",
"0a4c13cc1a718ca63b19615ea4cf25156f43aa76c31f751e4f1cd12e3ae06f60"
],
[
"proxy",
"https://mementomori.social/users/njsg/statuses/116187119231422622",
"activitypub"
],
[
"L",
"pink.momostr"
],
[
"l",
"pink.momostr.activitypub:https://mementomori.social/users/njsg/statuses/116187119231422622",
"pink.momostr"
],
[
"-"
]
],
"content": "Some forks and siblings do try to backport e.g. security fixes (and also features) from Mozilla's mozilla-* and comm-* repositories.\n\nWhat I can imagine totally happening is that, if they are going full GenAI, they'll somehow touch the entire codebase with it. They've done reformattings and similar changes in the past. Browsing through the mercurial history becomes more fun, and, more importantly, patches won't apply cleanly as the code went through such modifications. So, now? I can totally see GenAI touching all the files for reformatting so that you can't just backport the reformatting commits.\n\n(And this without considering the whole mess that is JS and HTML features being defined in unversioned, non-stable and backwards-incompatible \"living standards\". (fetch() comes to mind as an example of this, the defaults were changed, and sites relying on the new defaults broke where fetch() was available with the old defaults...; I didn't check well enough to know whether postMessage()'s new features are an example of this or not (it did break reCAPTCHA in SeaMonkey \u003c 2.53.23)))",
"sig": "dcf4a04caf6770f7b4132e8e8e2d8d680fc8234360d23ba38dbdd40ee8e70c99f7d9acd2fa3c9ac4ec087373e25767dd28d90005c482491b246090a7ac5e7b37"
}