Half of this file is a log; say where it should go
This commit is contained in:
parent
c05edef985
commit
10fcde8ce9
19
NEXT.md
19
NEXT.md
@ -63,6 +63,25 @@ merged; 3 and 4 are still running.
|
||||
`Emit.redefinition` already takes and is already tested for. The second also unblocks source interleaving in the
|
||||
disassembly buffer.
|
||||
|
||||
### Queued — split this file
|
||||
|
||||
About half of it is a build log. Of 1,682 lines, roughly 850 narrate work that is finished: lines 727-1507, from "The
|
||||
reload primitive" through the two break-loop sections, plus "What milestone 4 added" and "Where build time goes".
|
||||
|
||||
Do not simply delete them. That narrative is where the load-bearing reasons live — why nothing is ever `dlclose`d, why
|
||||
a call bound at link time cannot be made to notice a redefinition, why the renderer is a compile-time walk over a type
|
||||
rather than a function in the runtime. Losing those means deriving them again.
|
||||
|
||||
- **`NEXT.md` keeps** what is forward-looking: Start here, In flight, the Queued sections, Managed classes, Open,
|
||||
Blocked and unfinished with its six sublists, Sharp edges, Macros, Watch for, Untracked on purpose.
|
||||
- **`BUILT.md` takes** how the parts that exist actually work: the reload primitive, cells, the agent, the session,
|
||||
the daemon, the Emacs client, `C-x C-e`, the REPL buffer, conditions and the break loop — plus "The layout, which is
|
||||
the whole backend design" and "Why the FFI goes through a C shim", which are rationale rather than plans.
|
||||
- **Delete outright:** "Next — the REPL is the priority". Every item in it is struck through and marked Done, and it
|
||||
says what "Start here" already says.
|
||||
|
||||
`overview.md` is not the home for any of it; it is 54 lines and marks itself superseded.
|
||||
|
||||
### Queued — the runtime under a sanitizer
|
||||
|
||||
Nothing has ever run under ASan or UBSan on the test path. `grep -i 'sanitize\|asan\|valgrind'` over `lib/ bin/ test/
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user