Commit d68c77b5 authored by Edward Z. Yang's avatar Edward Z. Yang

[docs/backpack] Get lint to stop complaining

Signed-off-by: default avatarEdward Z. Yang <ezyang@cs.stanford.edu>
parent 7a15a68d
......@@ -484,7 +484,7 @@ in an applicative module system! No really!''} \\
Flattening can be seen as one of four different representations of packages
at the OS/library level. While it promotes maximal sharing of identical
modules, it is perhaps too fine-grained for most purposes.
\emph{TODO: Describe the alternatives.}
\emph{ToDo: Describe the alternatives.}
\paragraph{Package slicing} Instead of changing the package database,
we automatically
......@@ -953,14 +953,14 @@ rely on that implementation. However, the Renamer does other, harder things
that we don't need, so ideally we could factor out the import/export
resolution from the Renamer for use in shaping.
Unfortunately the Renamer's import resolution analyzes .hi files, but for
Unfortunately the Renamer's import resolution analyzes \verb|.hi| files, but for
local modules, which haven't yet been typechecked, we don't have those.
Instead, we could use a new file format, .hsi files, to store the shape of
Instead, we could use a new file format, \verb|.hsi| files, to store the shape of
a locally defined module. (Defined packages are bundled with their shapes,
so included modules have .hsi files as well.) (What about the logical
so included modules have \verb|.hsi| files as well.) (What about the logical
vs.~physical distinction in file names?) If we refactor the import/export
resolution code, could we rewrite it to generically operate on both
.hi files and .hsi files?
\verb|.hi| files and \verb|.hsi| files?
Alternatively, rather than storing shapes on a per-source basis, we could
store (in memory) the entire package shape. Similarly, included packages
......@@ -1109,7 +1109,7 @@ being the same.)
The logical shape context changed to say that \verb|A| and
\verb|B| now map to the same physical module identity. But does the existing
recompilation avoidance mechanism say that \verb|M| should be re-typechecked?
It's unclear. The .hi file for \verb|M| records that it imported \verb|A| and
It's unclear. The \verb|.hi| file for \verb|M| records that it imported \verb|A| and
\verb|B| with particular ABIs, but does it also know about the physical module
identities (or rather, original module names) of these modules?
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment