asdf merge requestshttps://gitlab.common-lisp.net/asdf/asdf/-/merge_requests2022-01-10T01:07:58Zhttps://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/199Better handling of missing packages in DEFINE-PACKAGE :IMPORT-FROM2022-01-10T01:07:58ZEric TimmonsBetter handling of missing packages in DEFINE-PACKAGE :IMPORT-FROMCloses #98 #101
Signal a style warning if we `:import-from` using a package that does not exist and an empty symbol list. Remains an error to `:import-from` a package that does not exist with a non-empty symbol list.
Add some tests to...Closes #98 #101
Signal a style warning if we `:import-from` using a package that does not exist and an empty symbol list. Remains an error to `:import-from` a package that does not exist with a non-empty symbol list.
Add some tests to try and ensure `:uiop/package` and `:uiop/package-local-nicknames` are not modified.3.3.6Robert GoldmanRobert Goldmanhttps://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/203Refine DEFINE-PACKAGE docstring.2022-01-17T21:08:35ZRobert GoldmanRefine DEFINE-PACKAGE docstring.Expand on the use of :UNINTERN keyword argument, based on discussion
around issue #85.
/cc @gefjonExpand on the use of :UNINTERN keyword argument, based on discussion
around issue #85.
/cc @gefjon3.3.6https://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/204Add allegro to CI2022-01-20T23:31:15ZEric TimmonsAdd allegro to CII finally got the Allegro Express Docker image built (had some technical issues related to it being 32-bit only that were only recently resolved).
Some TODOs before this is ready for merging:
- [x] Make sure tests actually pass.
- [x] ...I finally got the Allegro Express Docker image built (had some technical issues related to it being 32-bit only that were only recently resolved).
Some TODOs before this is ready for merging:
- [x] Make sure tests actually pass.
- [x] Confirm with Franz that they're happy with the image. I reached out to them before doing it to explain my plans and got their approval, but I think it'll be nice to just touch bases again.
- [x] Confirm that using it for ASDF CI purposes is fine with the Express license. My read is that it's fine so long as the ASDF development is non-commercial. But I'll leave that decision up to @rgoldman.https://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/201Adds FAQ item on customizing paths for sources and binaries, fixes typo2022-02-14T20:13:00ZNathan RingoAdds FAQ item on customizing paths for sources and binaries, fixes typoIt's my first time writing Texinfo, so if there's some additional linter or something to run, I can do so!
This adds a short FAQ item describing the small build script I use for building binaries as part of a larger (mostly non-CL) proj...It's my first time writing Texinfo, so if there's some additional linter or something to run, I can do so!
This adds a short FAQ item describing the small build script I use for building binaries as part of a larger (mostly non-CL) project. It hopefully conforms to best practices; if not, I'll change it such that it does.3.3.6https://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/205Expand UIOP:DEFINE-PACKAGE docstring.2022-02-23T17:22:00ZRobert GoldmanExpand UIOP:DEFINE-PACKAGE docstring.Specify behavior of DEFINE-PACKAGE without specified USE.Specify behavior of DEFINE-PACKAGE without specified USE.3.3.6https://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/211Add clasp support for uiop/launch-program2022-06-28T21:27:38ZChristian SchafmeisterAdd clasp support for uiop/launch-programClasp now supports uiop/launch-program.
Copied source code from ecl into clasp to support uiop/launch-program
so API's between clasp and ecl are identical.Clasp now supports uiop/launch-program.
Copied source code from ecl into clasp to support uiop/launch-program
so API's between clasp and ecl are identical.https://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/212Test for issue 124.2022-07-26T20:58:47ZRobert GoldmanTest for issue 124.If `*package*` is set to a package that does not import `defpackage`,
then loading the test system will fail.
This test, which is set as an expected failure right now, demonstrates the existence of #124If `*package*` is set to a package that does not import `defpackage`,
then loading the test system will fail.
This test, which is set as an expected failure right now, demonstrates the existence of #124https://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/213Bind package to cl user2022-07-26T23:18:47ZRobert GoldmanBind package to cl userBind the value of `*package*` to `common-lisp-user` package when compiling or loading lisp code. This binding can be overridden, but avoids anomalies that happen when ASDF is executed in an odd `*package*` context. See #124Bind the value of `*package*` to `common-lisp-user` package when compiling or loading lisp code. This binding can be overridden, but avoids anomalies that happen when ASDF is executed in an odd `*package*` context. See #124Robert GoldmanRobert Goldmanhttps://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/208Modify tests for ABCL 1.9 release.2022-08-02T14:09:14ZRobert GoldmanModify tests for ABCL 1.9 release.Update for tests that can now be expected to pass instead of failing.
Also fix issue #115Update for tests that can now be expected to pass instead of failing.
Also fix issue #1153.3.6Robert GoldmanRobert Goldmanhttps://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/221Resolve "WITH-TEMPORARY-FILE gives warnings when pathname variable is not used"2023-08-14T14:58:32ZRobert GoldmanResolve "WITH-TEMPORARY-FILE gives warnings when pathname variable is not used"Closes #142Closes #142Robert GoldmanRobert Goldmanhttps://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/222Resolve "uiop: Remember single/multi-threaded in the implementation-identifier."2023-08-29T22:07:27ZRobert GoldmanResolve "uiop: Remember single/multi-threaded in the implementation-identifier."Closes #143
I don't think it's important that we have a complex version specifier so much as that we avoid folding together sets of inconsistent FASL files, correct?Closes #143
I don't think it's important that we have a complex version specifier so much as that we avoid folding together sets of inconsistent FASL files, correct?https://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/223Resolve "(SETF (UIOP:GETENV VAR) NIL) should unset environment variable VAR"2023-10-29T18:38:42ZRobert GoldmanResolve "(SETF (UIOP:GETENV VAR) NIL) should unset environment variable VAR"Closes #146Closes #146Robert GoldmanRobert Goldmanhttps://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/224Fix to with-optimization-settings.2023-12-14T01:06:19ZRobert GoldmanFix to with-optimization-settings.Thanks to Yehouda Harpaz. Fixes error in backquote syntax.
Fixes #148Thanks to Yehouda Harpaz. Fixes error in backquote syntax.
Fixes #148Robert GoldmanRobert Goldmanhttps://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/219Resolve "uiop:dumplisp compiles with warning on Allegro 11"2024-01-04T19:03:30ZRobert GoldmanResolve "uiop:dumplisp compiles with warning on Allegro 11"Closes #141Closes #141Robert GoldmanRobert Goldmanhttps://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/225Restore correct uiop/dump-image on older ACLs.2024-01-08T15:49:22ZRobert GoldmanRestore correct uiop/dump-image on older ACLs.The Allegro API `sys:resize-areas` changed in Allegro 11, and in trying
to adapt to that, I broke function on earlier versions. Now check for
Allegro version and compile appropriately.
I hope this finally fixes #141The Allegro API `sys:resize-areas` changed in Allegro 11, and in trying
to adapt to that, I broke function on earlier versions. Now check for
Allegro version and compile appropriately.
I hope this finally fixes #141Robert GoldmanRobert Goldmanhttps://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/226p-i-s: defpackage: handle SBCL exts, cerror on unknown option2024-01-10T15:55:19ZPhoebe Goldmanp-i-s: defpackage: handle SBCL exts, cerror on unknown optionPer issue 147, this commit makes two (related) changes:
- When encountering an unknown `defpackage` option, signal a continuable error
of class `package-inferred-system-unknown-defpackage-option-error`,
rather than previously signal...Per issue 147, this commit makes two (related) changes:
- When encountering an unknown `defpackage` option, signal a continuable error
of class `package-inferred-system-unknown-defpackage-option-error`,
rather than previously signalling an `ecase` fallthrough error.
- Infer dependencies from the two nonstandard options
defined by SBCL's package locks extension, `:lock` and `:implement`.
`:lock` introduces no dependencies,
while `:implement` introduces dependencies as if it was `:use`.
This commit also extends the `package-inferred-system-test.script`
to exercise SBCL's package locks extension.
Assigned to @rgoldman for review, since he assigned the issue to me.
CC @Gleefre, who opened the issue.
Other implementations' extensions:
| Implementation | Extensions to DEFPACKAGE |
| ------ | ------ |
| CMUCL | None |
| Lispworks | `:local-nicknames` (already handled), `:add-use-defaults` [1](#fn:1) |
| Allegro | `:implementation-packages` (pretty sure this should be ignored because the implementation packages are likely forward references), `:alternate-name` (ignore), `:local-nicknames` (already handled), `:flat` (ignore), hierarchical packages [2](#fn:2) |
| clasp | None mentioned in docs |
| ECL | `:lock` |
| CCL | None mentioned in docs |
| clisp | None mentioned in docs |
| MKCL | Don't see any docs at all. Maybe shares the `lock` API with ECL? |
# Notes
<a id="fn:1">See [Lispworks docs](https://www.lispworks.com/documentation/lw80/lw/lw-common-lisp-12.htm)</a1>. As far as I can tell, you just ignore `:add-use-defaults` on Lispworks, and probably should raise either a warning or a continuable error on any other implementation.
<a id="fn:2">See [Franz's Allegro docs](https://franz.com/support/documentation/11.0/implementation.html#63-cldefpackage-implementation). [Franz's hierarchical packages](https://franz.com/support/documentation/11.0/packages.html#hier-packs-1).
Hierarchical packages impose a different package name parsing: there are _relative_ names. I think we can safely ignore this issue for now. If they don't work, it's not this issue that keeps them from working.
</a>Phoebe GoldmanPhoebe Goldmanhttps://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/206Iss1082024-01-26T23:50:55ZRobert GoldmanIss108Address #108 by checking for presence of Allegro patch fixing the misbehavior.
Don't have any great ideas about how to test this, but it should be feasible. Complex, though.Address #108 by checking for presence of Allegro patch fixing the misbehavior.
Don't have any great ideas about how to test this, but it should be feasible. Complex, though.https://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/227Another revision of UIOP:DUMP-IMAGE for Allegro.2024-01-27T00:29:43ZRobert GoldmanAnother revision of UIOP:DUMP-IMAGE for Allegro.Based on further discussions with Duane Rettig and Kevin Layer at Franz.Based on further discussions with Duane Rettig and Kevin Layer at Franz.Robert GoldmanRobert Goldmanhttps://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/228Resolve "On Allegro 11 test-deferred-warnings fails"2024-01-27T18:53:38ZRobert GoldmanResolve "On Allegro 11 test-deferred-warnings fails"Closes #152
Variables consulted by `UIOP:reify-deferred-warnings` were originally list-valued, but are now hash-valued, causing errors in this function.Closes #152
Variables consulted by `UIOP:reify-deferred-warnings` were originally list-valued, but are now hash-valued, causing errors in this function.Robert GoldmanRobert Goldmanhttps://gitlab.common-lisp.net/asdf/asdf/-/merge_requests/207Doc fixes2024-01-27T18:57:25ZRobert GoldmanDoc fixesFix a number of issues in the discussion of how to create a custom operation, identified by Zaque and @gefjon
Discussion of `component-depends-on` and, to a lesser extent `input-files` were out of date.
Also, in passing, fixed a docst...Fix a number of issues in the discussion of how to create a custom operation, identified by Zaque and @gefjon
Discussion of `component-depends-on` and, to a lesser extent `input-files` were out of date.
Also, in passing, fixed a docstring typo.