Dear Anton,
could you try what is currently in the "test" branch of ASDF as a release candidate for 3.2.1 ? And if you have time, could you try what is in the "plan" branch as a release candidate for 3.3.0 ?
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org Puritanism: The haunting fear that someone, somewhere, may be happy. ― H.L. Mencken
Thanks a lot!
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org Nostalgia isn’t what it used to be.
On Fri, Mar 10, 2017 at 10:03 PM, Anton Vodonosov avodonosov@yandex.ru wrote:
Faré, hello. Sorry for the delay,
I've started the tests for 3.2.1 RC from the test branch (commit 4d66692).
I'm packing for travel, so there might be delays in publishing results due to time / internet availability.
Best regards,
- Anton
09.03.2017, 04:35, "Faré" fahree@gmail.com:
Dear Anton,
could you try what is currently in the "test" branch of ASDF as a release candidate for 3.2.1 ? And if you have time, could you try what is in the "plan" branch as a release candidate for 3.3.0 ?
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org Puritanism: The haunting fear that someone, somewhere, may be happy. ― H.L. Mencken
I've finally build the reports.
Sorry for the delay, it's to a large part caused by amazon simpledb glitches - it took several days to extract data.
The lisps completed so far:
abcl-1.4.0-fasl42-linux-x86 ccl-1.10-r16196-f96-linux-x86 ccl-1.11-r16635-f96-linux-x86 ccl-1.9-r15756-f96-linux-x86 clisp-2.49-unix-x86 cmu-snapshot-2016-12__21b_unicode_-linux-x86 ecl-16.1.2-unknown-linux-x86-bytecode ecl-16.1.2-unknown-linux-x86-lisp-to-c sbcl-1.1.16-linux-x86 sbcl-1.2.6-linux-x86 sbcl-1.3.12-linux-x86
The full diff: https://common-lisp.net/project/cl-test-grid/asdf/asdf-diff-57.html
But you must be mostly interested to see the subset where tests fail on new ASDF: https://common-lisp.net/project/cl-test-grid/asdf/asdf-diff-58.html
Most of these failures are "OPERATION instances must only be created through MAKE-OPERATION"
After filtering out these errors, this is what remains: https://common-lisp.net/project/cl-test-grid/asdf/asdf-diff-59.html
SBCL 1.1.10 continues to fall into debugger as we discussed in the past (http://www.mail-archive.com/asdf-devel@common-lisp.net/msg05509.html)
Best regards, - Anton
OK, so regarding the calls to make-instance, eco needs a fix: https://github.com/eudoxia0/eco/issues/1 and cl-protobufs needed a fix (now committed, pushed).
Many of the errors are due to an old iolib. The bug was fixed in December, but somehow quicklisp is following iolib releases rather than the master branch. Stelian, should you make a new release? Or should quicklisp follow master?
Otherwise, it looks mostly good. There is some breakage such as lisp-executable on ecl, but I'll let Mark address that.
Anton, no need to test my 3.3.0 candidate anymore at the moment: I found a blocking bug in it that I must address first.
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org Reisner's Rule of Conceptual Inertia: If you think big enough, you'll never have to do it.
On Fri, Mar 17, 2017 at 3:36 PM, Anton Vodonosov avodonosov@yandex.ru wrote:
I've finally build the reports.
Sorry for the delay, it's to a large part caused by amazon simpledb glitches - it took several days to extract data.
The lisps completed so far:
abcl-1.4.0-fasl42-linux-x86 ccl-1.10-r16196-f96-linux-x86 ccl-1.11-r16635-f96-linux-x86 ccl-1.9-r15756-f96-linux-x86 clisp-2.49-unix-x86 cmu-snapshot-2016-12__21b_unicode_-linux-x86 ecl-16.1.2-unknown-linux-x86-bytecode ecl-16.1.2-unknown-linux-x86-lisp-to-c sbcl-1.1.16-linux-x86 sbcl-1.2.6-linux-x86 sbcl-1.3.12-linux-x86
The full diff: https://common-lisp.net/project/cl-test-grid/asdf/asdf-diff-57.html
But you must be mostly interested to see the subset where tests fail on new ASDF: https://common-lisp.net/project/cl-test-grid/asdf/asdf-diff-58.html
Most of these failures are "OPERATION instances must only be created through MAKE-OPERATION"
After filtering out these errors, this is what remains: https://common-lisp.net/project/cl-test-grid/asdf/asdf-diff-59.html
SBCL 1.1.10 continues to fall into debugger as we discussed in the past (http://www.mail-archive.com/asdf-devel@common-lisp.net/msg05509.html)
Best regards,
- Anton
Dear Anton,
I fixed the blocker issue in the 3.3.0 candidate "plan" branch (rebased). Can you test it now?
And, to avoid known issues, can you override iolib and cl-protobufs with recent git checkouts?
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org No one has the right to a position; everyone has the right to positions being well filled. — Ernest Renan
Started.
Meanwhile, the reports for 3.2.0 RC were updated with some more lisps. Now it's
abcl-1.4.0-fasl42-linux-x86 ccl-1.10-r16196-f96-linux-x86 ccl-1.11-r16635-f96-linux-x86 ccl-1.8-r15286m-f95-linux-x86 ccl-1.9-r15756-f96-linux-x86 clisp-2.49-unix-x86 cmu-snapshot-2014-01__20e_unicode_-linux-x86 cmu-snapshot-2014-05-dirty__20e_unicode_-linux-x86 cmu-snapshot-2014-12___20f_unicode_-linux-x86 cmu-snapshot-2016-12__21b_unicode_-linux-x86 ecl-13.5.1-unknown-linux-i686-bytecode ecl-13.5.1-unknown-linux-i686-lisp-to-c ecl-16.0.0-98fc12d3-linux-i686-bytecode ecl-16.0.0-98fc12d3-linux-i686-lisp-to-c ecl-16.1.2-unknown-linux-x86-bytecode ecl-16.1.2-unknown-linux-x86-lisp-to-c sbcl-1.1.16-linux-x86 sbcl-1.2.6-linux-x86 sbcl-1.3.12-linux-x86
Best regards, - Anton
19.03.2017, 16:27, "Faré" fahree@gmail.com:
Dear Anton,
I fixed the blocker issue in the 3.3.0 candidate "plan" branch (rebased). Can you test it now?
And, to avoid known issues, can you override iolib and cl-protobufs with recent git checkouts?
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org No one has the right to a position; everyone has the right to positions being well filled. — Ernest Renan
On Sun, Mar 19, 2017 at 3:09 PM, Anton Vodonosov avodonosov@yandex.ru wrote:
Started.
Thanks.
Unhappily, iolib requires a patch to work with ASDF 3.3: https://github.com/sionescu/iolib/pull/42
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org Solipsism is a lonely place. Psychopaths crave love, but can't get no satisfaction: even elected by millions, it's still non-people voting for them.
On Mon, Mar 20, 2017 at 1:11 PM, Anton Vodonosov avodonosov@yandex.ru wrote:
Running with updated iolib.
Meanwhile, here is what has completed for the previous setup:
ccl-1.10-r16196-f96-linux-x86 cmu-snapshot-2016-12__21b_unicode_-linux-x86 sbcl-1.3.12-linux-x86
https://common-lisp.net/project/cl-test-grid/asdf/asdf-diff-60.html
Some errors are rather strange.
Looking at the last four failures, one I cannot reproduce, and three are genuine incompatibilities introduced by ASDF 3.3, that I need to work on and/or get people to fix their code.
ASDF 3.3 is definitely not ready to be merged into master, but that's great progress, and I'll investigate further to resolve the remaining incompatibilities.
trivial-arguments --> https://github.com/Shinmera/trivial-arguments/issues/3 weblocks --> https://groups.google.com/forum/#!topic/weblocks/MYt3IT9ao3E yaclanapht --> works for me zenekindarl --> https://github.com/KeenS/zenekindarl/issues/20
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org Monad=Computation with effects. M α=computation that returns α. bind=chain computations. return=computation with no effect. Monad Transformer=mixin that adds some effects to computations.
OK, so with ~120 failures, obviously, ASDF 3.3 needs to be made more compatible, using style-warnings where there are currently warnings, providing additional backward-compatibility functions, etc.
staple => https://github.com/Shinmera/staple/issues/3 moptilities => https://github.com/gwkkwg/moptilities/pull/2 (also fixes tinaa, ...) transparent-wrap => inherited from trivial-arguments stump-touchy-mode-line => otherwise fixed
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org Anarchism is founded on the observation that since few men are wise enough to rule themselves, even fewer are wise enough to rule others. — Edward Abbey
On Tue, Mar 21, 2017 at 9:50 AM, Faré fahree@gmail.com wrote:
On Mon, Mar 20, 2017 at 1:11 PM, Anton Vodonosov avodonosov@yandex.ru wrote:
Running with updated iolib.
Meanwhile, here is what has completed for the previous setup:
ccl-1.10-r16196-f96-linux-x86 cmu-snapshot-2016-12__21b_unicode_-linux-x86 sbcl-1.3.12-linux-x86
https://common-lisp.net/project/cl-test-grid/asdf/asdf-diff-60.html
Some errors are rather strange.
Looking at the last four failures, one I cannot reproduce, and three are genuine incompatibilities introduced by ASDF 3.3, that I need to work on and/or get people to fix their code.
ASDF 3.3 is definitely not ready to be merged into master, but that's great progress, and I'll investigate further to resolve the remaining incompatibilities.
trivial-arguments --> https://github.com/Shinmera/trivial-arguments/issues/3 weblocks --> https://groups.google.com/forum/#!topic/weblocks/MYt3IT9ao3E yaclanapht --> works for me zenekindarl --> https://github.com/KeenS/zenekindarl/issues/20
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org Monad=Computation with effects. M α=computation that returns α. bind=chain computations. return=computation with no effect. Monad Transformer=mixin that adds some effects to computations.
Failures on new version only: https://common-lisp.net/project/cl-test-grid/asdf/asdf-diff-62.html
it's very interesting, babel on ccl fails with this:
QUICKLISP-CLIENT:SYSTEM-NOT-FOUND : System "hu.dwim.stefil-api-impl" not found
what is this -api-impl thing? rgrep "api" babel/ and hu.dwim.stefil/ gives nothing. rgrep -- "api-impl" quicklisp-* yields also nothing.
where is this coming from?
I made another round of fixes based on Anton's test results. There are more systems for me to test.
Common incompatibilities ------------------------------- * people using deprecated functions, notably SYSTEM-DEFINITION-PATHNAME or RUN-SHELL-COMMAND. See docstrings of these functions for better replacements. * people using clear-system from perform. That's a big no-no, unhappily, it has been promoted by caveman and prove. Bad! Please undo this thing, it confuses ASDF to no end. ASDF will know that tests need to be re-run if you modified anything. If not, you can force a re-run with (load-system :my-prove-system-test :force t).
Notable things ------------------ * Attila, the -api-impl failures are part of the cl-test-grid infrastructure, not of the underlying systems tested by quicklisp, so don't worry about it. All the hu.dwim.* systems build for me after iolib and moptilities are updated. Some fail by lack of libgvc.so.4, and some fail tests, but that's not my issue. I also had issues with updating some repos with darcs, but could get a new copy from empty.
* I sent a fix to a fork, but the original may be unmaintained: arnesi https://github.com/vseloved/arnesi/pull/1
Fixes I made ----------------
I directly pushed fixes to these: asdf-dependency-grovel asdf-finalizers
I submitted fixes to the following: babel https://github.com/cl-babel/babel/pull/30 bknr https://github.com/hanshuebner/bknr-web/pull/4 caveman https://github.com/fukamachi/caveman/pull/95 caveman2-widgets https://github.com/ritschmaster/caveman2-widgets/pull/8 caveman2-widgets-bootstrap https://github.com/ritschmaster/caveman2-widgets-bootstrap/pull/3 check-it https://github.com/DalekBaldwin/check-it/pull/20 cl-vectors https://github.com/fjolliton/cl-vectors/pull/4 madeira-port https://github.com/nikodemus/madeira-port/pull/6 (fixes sb-cga) prove https://github.com/fukamachi/prove/pull/44 s-dot (send diff by email to martin loetzsch) scriptl-examples https://github.com/rpav/ScriptL/pull/5
Fixed by my modifying ASDF to not complain about calling require in a lisp file: series
Cannot reproduce (could have been fixed by some of my recent ASDF tweaks): arrow-macros arrow-macros-test cffi-tests cl-amqp.test prepl projectured.* retrospectiff screamer-tests shelly-test stump-touchy-mode-line
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org Brain, n.: The apparatus with which we think that we think. — Ambrose Bierce, "The Devil's Dictionary"
On Sun, Mar 26, 2017 at 3:47 PM, Attila Lendvai attila@lendvai.name wrote:
Failures on new version only: https://common-lisp.net/project/cl-test-grid/asdf/asdf-diff-62.html
it's very interesting, babel on ccl fails with this:
QUICKLISP-CLIENT:SYSTEM-NOT-FOUND : System "hu.dwim.stefil-api-impl" not found
what is this -api-impl thing? rgrep "api" babel/ and hu.dwim.stefil/ gives nothing. rgrep -- "api-impl" quicklisp-* yields also nothing.
where is this coming from?
-- • attila lendvai • PGP: 963F 5D5F 45C7 DFCD 0A39 -- “Once the fabric of a just society is undone, it takes generations to weave it back together.” — Deepak Chopra
There's a weird bug in ASDF 3.3 that causes UIOP to be invalidated when used in defsystem-depends-on, so that CFFI to be always recompiled. Working on it.
But still going through the previous list of failures.
More fixes: cl-autorepo https://github.com/billstclair/cl-autorepo/pull/1 (run-shell-command) cl-blapack https://github.com/blindglobe/cl-blapack/pull/8 (system-definition-pathname) trivial-shell https://github.com/gwkkwg/trivial-shell/pull/11 (bad package sb-threads) cl-bson https://github.com/EuAndreh/cl-BSON/pull/1 (clear-system in perform) cl-cuda https://github.com/takagi/cl-cuda/pull/88 (missing mark-operation-done, clear-system) cl-gists https://github.com/Rudolph-Miller/cl-gists/pull/5 (clear-system in perform) cl-python https://github.com/metawilm/cl-python/pull/21 (lots of small issues) cl-secure-read https://github.com/mabragor/cl-secure-read/pull/2 (using operate instead of dependencies, what more with :force t) cl-tasukete https://github.com/gos-k/cl-tasukete/pull/3 (clear-system in perform) cl-textmagic https://github.com/trumae/cl-textmagic/pull/2 (clear-system in perform) cl-weather-jp https://github.com/fukamachi/cl-weather-jp/pull/1 f2cl (massive refactoring from asdf1 to asdf3, sent by mail to Raymond Toy)
More that now build for me: cl-bunny cl-custom-hash-table cl-gpu cl-iconv cl-indeterminism cl-inotify cl-ledger cl-libssh2 (though sent https://github.com/alxchk/cl-libssh2/pull/6) cl-yaclyaml
Can you tell I'm procrastinating on important things?
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org I have not yet begun to procrastinate
On Mon, Mar 27, 2017 at 4:26 AM, Faré fahree@gmail.com wrote:
I made another round of fixes based on Anton's test results. There are more systems for me to test.
Common incompatibilities
- people using deprecated functions, notably
SYSTEM-DEFINITION-PATHNAME or RUN-SHELL-COMMAND. See docstrings of these functions for better replacements.
- people using clear-system from perform. That's a big no-no,
unhappily, it has been promoted by caveman and prove. Bad! Please undo this thing, it confuses ASDF to no end. ASDF will know that tests need to be re-run if you modified anything. If not, you can force a re-run with (load-system :my-prove-system-test :force t).
Notable things
- Attila, the -api-impl failures are part of the cl-test-grid
infrastructure, not of the underlying systems tested by quicklisp, so don't worry about it. All the hu.dwim.* systems build for me after iolib and moptilities are updated. Some fail by lack of libgvc.so.4, and some fail tests, but that's not my issue. I also had issues with updating some repos with darcs, but could get a new copy from empty.
- I sent a fix to a fork, but the original may be unmaintained:
arnesi https://github.com/vseloved/arnesi/pull/1
Fixes I made
I directly pushed fixes to these: asdf-dependency-grovel asdf-finalizers
I submitted fixes to the following: babel https://github.com/cl-babel/babel/pull/30 bknr https://github.com/hanshuebner/bknr-web/pull/4 caveman https://github.com/fukamachi/caveman/pull/95 caveman2-widgets https://github.com/ritschmaster/caveman2-widgets/pull/8 caveman2-widgets-bootstrap https://github.com/ritschmaster/caveman2-widgets-bootstrap/pull/3 check-it https://github.com/DalekBaldwin/check-it/pull/20 cl-vectors https://github.com/fjolliton/cl-vectors/pull/4 madeira-port https://github.com/nikodemus/madeira-port/pull/6 (fixes sb-cga) prove https://github.com/fukamachi/prove/pull/44 s-dot (send diff by email to martin loetzsch) scriptl-examples https://github.com/rpav/ScriptL/pull/5
Fixed by my modifying ASDF to not complain about calling require in a lisp file: series
Cannot reproduce (could have been fixed by some of my recent ASDF tweaks): arrow-macros arrow-macros-test cffi-tests cl-amqp.test prepl projectured.* retrospectiff screamer-tests shelly-test stump-touchy-mode-line
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org Brain, n.: The apparatus with which we think that we think. — Ambrose Bierce, "The Devil's Dictionary"
On Sun, Mar 26, 2017 at 3:47 PM, Attila Lendvai attila@lendvai.name wrote:
Failures on new version only: https://common-lisp.net/project/cl-test-grid/asdf/asdf-diff-62.html
it's very interesting, babel on ccl fails with this:
QUICKLISP-CLIENT:SYSTEM-NOT-FOUND : System "hu.dwim.stefil-api-impl" not found
what is this -api-impl thing? rgrep "api" babel/ and hu.dwim.stefil/ gives nothing. rgrep -- "api-impl" quicklisp-* yields also nothing.
where is this coming from?
-- • attila lendvai • PGP: 963F 5D5F 45C7 DFCD 0A39 -- “Once the fabric of a just society is undone, it takes generations to weave it back together.” — Deepak Chopra
More progress. Almost there. Hopefully we can get things fixed in the next Quicklisp release or two; and then reevaluate the compatibility of ASDF 3.3 (that I have also improved).
Non-essential fixes to: cl-containers, metatilities-base, metatilities, moptilities, sent to gwking folio2 https://github.com/mikelevins/folio2/pull/2
Work for me: cserial-port datafly-test foreign-array glkit (sb-cga) hemlock.base (iolib) hemlock.clx (iolib) hemlock.tty (iolib) lambdalite linedit
Fixes to: cxml-rng (send by mail to david lichteblau) jwacs-tests https://github.com/chumsley/jwacs/pull/2 (system-definition-pathname) lol-re https://github.com/mabragor/lol-re/pull/4 (load-system in perform) defmacro-enhance https://github.com/mabragor/defmacro-enhance/pull/5 (load-system in perform) dexador https://github.com/fukamachi/dexador/pull/32 (clear-system in perform) metacopy, metacopy-with-contextl email to Attila and Gary (load-op in perform)
Plain broken: lisp-binary -- it's doing lower-level stuff than I care to debug, not obviously asdf fault. lisp-executable -- does forbidden stuff with ASDF internals that I'm not going to support; ASDF has program-op that is supported instead.
I still haven't inspected the spurious rebuilds of uiop that happened in some contexts. It might be due to my alternating between quicklisp-enabled and quicklisp-disabled contexts, which might somehow create cycles?
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org Success is getting what you want. Happiness is wanting what you get. — Dale Carnegie
On Mon, Mar 27, 2017 at 4:08 PM, Faré fahree@gmail.com wrote:
There's a weird bug in ASDF 3.3 that causes UIOP to be invalidated when used in defsystem-depends-on, so that CFFI to be always recompiled. Working on it.
But still going through the previous list of failures.
More fixes: cl-autorepo https://github.com/billstclair/cl-autorepo/pull/1 (run-shell-command) cl-blapack https://github.com/blindglobe/cl-blapack/pull/8 (system-definition-pathname) trivial-shell https://github.com/gwkkwg/trivial-shell/pull/11 (bad package sb-threads) cl-bson https://github.com/EuAndreh/cl-BSON/pull/1 (clear-system in perform) cl-cuda https://github.com/takagi/cl-cuda/pull/88 (missing mark-operation-done, clear-system) cl-gists https://github.com/Rudolph-Miller/cl-gists/pull/5 (clear-system in perform) cl-python https://github.com/metawilm/cl-python/pull/21 (lots of small issues) cl-secure-read https://github.com/mabragor/cl-secure-read/pull/2 (using operate instead of dependencies, what more with :force t) cl-tasukete https://github.com/gos-k/cl-tasukete/pull/3 (clear-system in perform) cl-textmagic https://github.com/trumae/cl-textmagic/pull/2 (clear-system in perform) cl-weather-jp https://github.com/fukamachi/cl-weather-jp/pull/1 f2cl (massive refactoring from asdf1 to asdf3, sent by mail to Raymond Toy)
More that now build for me: cl-bunny cl-custom-hash-table cl-gpu cl-iconv cl-indeterminism cl-inotify cl-ledger cl-libssh2 (though sent https://github.com/alxchk/cl-libssh2/pull/6) cl-yaclyaml
Can you tell I'm procrastinating on important things?
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org I have not yet begun to procrastinate
On Mon, Mar 27, 2017 at 4:26 AM, Faré fahree@gmail.com wrote:
I made another round of fixes based on Anton's test results. There are more systems for me to test.
Common incompatibilities
- people using deprecated functions, notably
SYSTEM-DEFINITION-PATHNAME or RUN-SHELL-COMMAND. See docstrings of these functions for better replacements.
- people using clear-system from perform. That's a big no-no,
unhappily, it has been promoted by caveman and prove. Bad! Please undo this thing, it confuses ASDF to no end. ASDF will know that tests need to be re-run if you modified anything. If not, you can force a re-run with (load-system :my-prove-system-test :force t).
Notable things
- Attila, the -api-impl failures are part of the cl-test-grid
infrastructure, not of the underlying systems tested by quicklisp, so don't worry about it. All the hu.dwim.* systems build for me after iolib and moptilities are updated. Some fail by lack of libgvc.so.4, and some fail tests, but that's not my issue. I also had issues with updating some repos with darcs, but could get a new copy from empty.
- I sent a fix to a fork, but the original may be unmaintained:
arnesi https://github.com/vseloved/arnesi/pull/1
Fixes I made
I directly pushed fixes to these: asdf-dependency-grovel asdf-finalizers
I submitted fixes to the following: babel https://github.com/cl-babel/babel/pull/30 bknr https://github.com/hanshuebner/bknr-web/pull/4 caveman https://github.com/fukamachi/caveman/pull/95 caveman2-widgets https://github.com/ritschmaster/caveman2-widgets/pull/8 caveman2-widgets-bootstrap https://github.com/ritschmaster/caveman2-widgets-bootstrap/pull/3 check-it https://github.com/DalekBaldwin/check-it/pull/20 cl-vectors https://github.com/fjolliton/cl-vectors/pull/4 madeira-port https://github.com/nikodemus/madeira-port/pull/6 (fixes sb-cga) prove https://github.com/fukamachi/prove/pull/44 s-dot (send diff by email to martin loetzsch) scriptl-examples https://github.com/rpav/ScriptL/pull/5
Fixed by my modifying ASDF to not complain about calling require in a lisp file: series
Cannot reproduce (could have been fixed by some of my recent ASDF tweaks): arrow-macros arrow-macros-test cffi-tests cl-amqp.test prepl projectured.* retrospectiff screamer-tests shelly-test stump-touchy-mode-line
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org Brain, n.: The apparatus with which we think that we think. — Ambrose Bierce, "The Devil's Dictionary"
On Sun, Mar 26, 2017 at 3:47 PM, Attila Lendvai attila@lendvai.name wrote:
Failures on new version only: https://common-lisp.net/project/cl-test-grid/asdf/asdf-diff-62.html
it's very interesting, babel on ccl fails with this:
QUICKLISP-CLIENT:SYSTEM-NOT-FOUND : System "hu.dwim.stefil-api-impl" not found
what is this -api-impl thing? rgrep "api" babel/ and hu.dwim.stefil/ gives nothing. rgrep -- "api-impl" quicklisp-* yields also nothing.
where is this coming from?
-- • attila lendvai • PGP: 963F 5D5F 45C7 DFCD 0A39 -- “Once the fabric of a just society is undone, it takes generations to weave it back together.” — Deepak Chopra
This series of incompatibilities is all resolved. Last bits:
WFM: mathkit modularize-hooks modularize-interfaces oclcl-test (iolib) opticl opticl-core opticl-doc parser-combinators-tests periods-series physical-dimension science-data nst (org-sampler)
Fixed: org-sampler email (run-shell-command) macrodynamics https://github.com/DalekBaldwin/macrodynamics/pull/1 (system-definition-pathname)
Plain broken: mnst-relay ? funky test in nst
Now to wait for all the fixed issues to propagate to upstream then Quicklisp.
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org National Independence: where individuals are Property of the Nation, and their shackles are MADE IN <THIS COUNTRY>. — Faré
On Wed, Mar 29, 2017 at 8:51 PM, Faré fahree@gmail.com wrote:
More progress. Almost there. Hopefully we can get things fixed in the next Quicklisp release or two; and then reevaluate the compatibility of ASDF 3.3 (that I have also improved).
Non-essential fixes to: cl-containers, metatilities-base, metatilities, moptilities, sent to gwking folio2 https://github.com/mikelevins/folio2/pull/2
Work for me: cserial-port datafly-test foreign-array glkit (sb-cga) hemlock.base (iolib) hemlock.clx (iolib) hemlock.tty (iolib) lambdalite linedit
Fixes to: cxml-rng (send by mail to david lichteblau) jwacs-tests https://github.com/chumsley/jwacs/pull/2 (system-definition-pathname) lol-re https://github.com/mabragor/lol-re/pull/4 (load-system in perform) defmacro-enhance https://github.com/mabragor/defmacro-enhance/pull/5 (load-system in perform) dexador https://github.com/fukamachi/dexador/pull/32 (clear-system in perform) metacopy, metacopy-with-contextl email to Attila and Gary (load-op in perform)
Plain broken: lisp-binary -- it's doing lower-level stuff than I care to debug, not obviously asdf fault. lisp-executable -- does forbidden stuff with ASDF internals that I'm not going to support; ASDF has program-op that is supported instead.
I still haven't inspected the spurious rebuilds of uiop that happened in some contexts. It might be due to my alternating between quicklisp-enabled and quicklisp-disabled contexts, which might somehow create cycles?
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org Success is getting what you want. Happiness is wanting what you get. — Dale Carnegie
On Mon, Mar 27, 2017 at 4:08 PM, Faré fahree@gmail.com wrote:
There's a weird bug in ASDF 3.3 that causes UIOP to be invalidated when used in defsystem-depends-on, so that CFFI to be always recompiled. Working on it.
But still going through the previous list of failures.
More fixes: cl-autorepo https://github.com/billstclair/cl-autorepo/pull/1 (run-shell-command) cl-blapack https://github.com/blindglobe/cl-blapack/pull/8 (system-definition-pathname) trivial-shell https://github.com/gwkkwg/trivial-shell/pull/11 (bad package sb-threads) cl-bson https://github.com/EuAndreh/cl-BSON/pull/1 (clear-system in perform) cl-cuda https://github.com/takagi/cl-cuda/pull/88 (missing mark-operation-done, clear-system) cl-gists https://github.com/Rudolph-Miller/cl-gists/pull/5 (clear-system in perform) cl-python https://github.com/metawilm/cl-python/pull/21 (lots of small issues) cl-secure-read https://github.com/mabragor/cl-secure-read/pull/2 (using operate instead of dependencies, what more with :force t) cl-tasukete https://github.com/gos-k/cl-tasukete/pull/3 (clear-system in perform) cl-textmagic https://github.com/trumae/cl-textmagic/pull/2 (clear-system in perform) cl-weather-jp https://github.com/fukamachi/cl-weather-jp/pull/1 f2cl (massive refactoring from asdf1 to asdf3, sent by mail to Raymond Toy)
More that now build for me: cl-bunny cl-custom-hash-table cl-gpu cl-iconv cl-indeterminism cl-inotify cl-ledger cl-libssh2 (though sent https://github.com/alxchk/cl-libssh2/pull/6) cl-yaclyaml
Can you tell I'm procrastinating on important things?
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org I have not yet begun to procrastinate
On Mon, Mar 27, 2017 at 4:26 AM, Faré fahree@gmail.com wrote:
I made another round of fixes based on Anton's test results. There are more systems for me to test.
Common incompatibilities
- people using deprecated functions, notably
SYSTEM-DEFINITION-PATHNAME or RUN-SHELL-COMMAND. See docstrings of these functions for better replacements.
- people using clear-system from perform. That's a big no-no,
unhappily, it has been promoted by caveman and prove. Bad! Please undo this thing, it confuses ASDF to no end. ASDF will know that tests need to be re-run if you modified anything. If not, you can force a re-run with (load-system :my-prove-system-test :force t).
Notable things
- Attila, the -api-impl failures are part of the cl-test-grid
infrastructure, not of the underlying systems tested by quicklisp, so don't worry about it. All the hu.dwim.* systems build for me after iolib and moptilities are updated. Some fail by lack of libgvc.so.4, and some fail tests, but that's not my issue. I also had issues with updating some repos with darcs, but could get a new copy from empty.
- I sent a fix to a fork, but the original may be unmaintained:
arnesi https://github.com/vseloved/arnesi/pull/1
Fixes I made
I directly pushed fixes to these: asdf-dependency-grovel asdf-finalizers
I submitted fixes to the following: babel https://github.com/cl-babel/babel/pull/30 bknr https://github.com/hanshuebner/bknr-web/pull/4 caveman https://github.com/fukamachi/caveman/pull/95 caveman2-widgets https://github.com/ritschmaster/caveman2-widgets/pull/8 caveman2-widgets-bootstrap https://github.com/ritschmaster/caveman2-widgets-bootstrap/pull/3 check-it https://github.com/DalekBaldwin/check-it/pull/20 cl-vectors https://github.com/fjolliton/cl-vectors/pull/4 madeira-port https://github.com/nikodemus/madeira-port/pull/6 (fixes sb-cga) prove https://github.com/fukamachi/prove/pull/44 s-dot (send diff by email to martin loetzsch) scriptl-examples https://github.com/rpav/ScriptL/pull/5
Fixed by my modifying ASDF to not complain about calling require in a lisp file: series
Cannot reproduce (could have been fixed by some of my recent ASDF tweaks): arrow-macros arrow-macros-test cffi-tests cl-amqp.test prepl projectured.* retrospectiff screamer-tests shelly-test stump-touchy-mode-line
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org Brain, n.: The apparatus with which we think that we think. — Ambrose Bierce, "The Devil's Dictionary"
On Sun, Mar 26, 2017 at 3:47 PM, Attila Lendvai attila@lendvai.name wrote:
Failures on new version only: https://common-lisp.net/project/cl-test-grid/asdf/asdf-diff-62.html
it's very interesting, babel on ccl fails with this:
QUICKLISP-CLIENT:SYSTEM-NOT-FOUND : System "hu.dwim.stefil-api-impl" not found
what is this -api-impl thing? rgrep "api" babel/ and hu.dwim.stefil/ gives nothing. rgrep -- "api-impl" quicklisp-* yields also nothing.
where is this coming from?
-- • attila lendvai • PGP: 963F 5D5F 45C7 DFCD 0A39 -- “Once the fabric of a just society is undone, it takes generations to weave it back together.” — Deepak Chopra
I still haven't inspected the spurious rebuilds of uiop that happened in some contexts.
FWIW, i also see those rebuilds every now and then if i have explicit dependency on :uiop or :asdf (which i've gotten rid of, so i don't see this anymore).
-
Anton Vodonosov -
Attila Lendvai -
Faré