mps

mps copied to clipboard

Documentation for mps_arena_postmortem does not indicate a requirement or a use case

1

The documentation for mps_arena_postmortem describes what it does, it does not describe why it was implemented, why you would want to use it, or indicate what sort of requirements it...

thejayps

proc.review.edit assumes everything from proc.review.log is clear to the proc.review.role.editor

1

As a proc.review.role.editor I have found logging information to be unclear subsequent to proc.review.log, and during proc.review.edit. It is possible that aspects of logging comments met with rule.generic.clarity for me...

thejayps

Arena extens/contraction callbacks documentation does not clearly describe when callbacks could be called

1

Raise https://github.com/Ravenbrook/mps/pull/213#discussion_r1190094507 Also see #224: what does "And check for other stuff (e.g. format methods) where we should mention that." mean?

thejayps

Explaining the purpose of roles can help leaders to understand the flexibility available to them in review planning, eg that a role could be split among 2 people. https://github.com/Ravenbrook/mps/blob/586336a7021c2276a3f798ad10e6e4ab5f438226/procedure/review.rst#52-review-planning

thejayps

In the case of https://github.com/Ravenbrook/mps/pull/213/files where the change doesn't include key parts of the product document, the review planning procedure does not prompt the leader to identify this part of...

thejayps

[mps_root_create_table](https://www.ravenbrook.com/project/mps/master/manual/html/topic/deprecated.html#c.mps_root_create_table) "Deprecated This function is equivalent to: mps_root_create_area(root_o, arena, rank, mode, base, base + count, mps_scan_area, NULL, 0)" The above is incorrect as there is one too many input arguments

thejayps

We don't have a management process for limiting [work in progress](https://en.wiktionary.org/wiki/work_in_progress), and so the number of branches and pull requests tends to grow without bound. Work in progress has planning...

rptb1

This issue is specific to internal Ravenbrook systems. Documentation can be found (if one knows where to look) under /html folders within the internal "infosys" for custom MPS versions. In...

thejayps

The shield is not checked in cool builds

2

Change d82da0e1401d5541a9fbba78ca2bd5b03cb0d6c1 refactored the shield into its own structure, ShieldStruct, which makes sense in terms of abstractions. [Various checks that were part of GlobalCheck](https://github.com/Ravenbrook/mps/commit/d82da0e1401d5541a9fbba78ca2bd5b03cb0d6c1#diff-8a34b9342ebc6fbd4d47e3d94dcc3e65871b738830a2a56f257a2c673bb213abL157-L189) were [moved to the new ShieldCheck](https://github.com/Ravenbrook/mps/commit/d82da0e1401d5541a9fbba78ca2bd5b03cb0d6c1#diff-32e19141eb9cbf244e80f8ad6859333c93af022817ccd8b252aad8c7e2cf8922R45-R83)...

rptb1

Tests assert with SHIELD_DEBUG set

1

Running tests with SHIELD_DEBUG enabled with `make -f lii6ll.gmk VARIETY=cool CFLAGS='-DSHIELD_DEBUG' clean testci` at current master (https://github.com/Ravenbrook/mps/commit/179341b6cc2d9ab623ea53d10914d66c3668d47d) produces multiple assertion failures in various tests, all of the form: ``` shield.c:644:...

rptb1