[armedbear-devel] Status of 1.2.0
Hi Mark, all, My own testing didn't reveal any notable problems, nor did the ANSI test suite. There are a few regressions in cl-test-grid which indicate our changes to symbol interning. My proposal is to go forward with the 1.2 release and release a bugfix as soon as we identify and fix the problem which the cl-colors tests indicate. What's your position? -- Bye, Erik. http://efficito.com -- Hosted accounting and ERP. Robust and Flexible. No vendor lock-in.
Actually, the issue turned out to be rather easy to fix. Committed a fix to trunk now and running ANSI tests. If tests are successful, I'll backport to 1.2.x and start a testrun on my testgrid machine. (I may need help there though since Anton has always operated it so far.) Bye, Erik. On Sun, Jun 16, 2013 at 11:50 AM, Erik Huelsmann <ehuels@gmail.com> wrote:
Hi Mark, all,
My own testing didn't reveal any notable problems, nor did the ANSI test suite. There are a few regressions in cl-test-grid which indicate our changes to symbol interning.
My proposal is to go forward with the 1.2 release and release a bugfix as soon as we identify and fix the problem which the cl-colors tests indicate.
What's your position?
-- Bye,
Erik.
http://efficito.com -- Hosted accounting and ERP. Robust and Flexible. No vendor lock-in.
-- Bye, Erik. http://efficito.com -- Hosted accounting and ERP. Robust and Flexible. No vendor lock-in.
16.06.2013, 14:27, "Erik Huelsmann" <ehuels@gmail.com>:
If tests are successful, I'll backport to 1.2.x and start a testrun on my testgrid machine. (I may need help there though since Anton has always operated it so far.)
Let me know if you have any questions
Hi Anton, Just a second ago I started a testrun in a screen session. Could you verify it's actually running and running correctly? (I'm running a trunk version, because they differ only by a few (irrelevant) commits and I didn't know how to start 1.2.x) Bye, Erik. On Sun, Jun 16, 2013 at 12:41 PM, Anton Vodonosov <avodonosov@yandex.ru>wrote:
16.06.2013, 14:27, "Erik Huelsmann" <ehuels@gmail.com>:
If tests are successful, I'll backport to 1.2.x and start a testrun on my testgrid machine. (I may need help there though since Anton has always operated it so far.)
Let me know if you have any questions
-- Bye, Erik. http://efficito.com -- Hosted accounting and ERP. Robust and Flexible. No vendor lock-in.
I have inspected ~/cl-test-grid/work-dir/agent/logs/agent.log It is now running tests for abcl-1.3.0-dev-svn-14545-fasl42-linux-x86. 16.06.2013, 15:18, "Erik Huelsmann" <ehuels@gmail.com>:
Hi Anton, Just a second ago I started a testrun in a screen session. Could you verify it's actually running and running correctly? (I'm running a trunk version, because they differ only by a few (irrelevant) commits and I didn't know how to start 1.2.x)
Bye,
Erik.
On Sun, Jun 16, 2013 at 12:41 PM, Anton Vodonosov <avodonosov@yandex.ru> wrote:
16.06.2013, 14:27, "Erik Huelsmann" <ehuels@gmail.com>:
If tests are successful, I'll backport to 1.2.x and start a testrun on my testgrid machine. (I may need help there though since Anton has always operated it so far.)
Let me know if you have any questions
-- Bye, Erik.
http://efficito.com -- Hosted accounting and ERP. Robust and Flexible. No vendor lock-in.
That's what I wanted to start. Can we set up a separate 'run-agent' for 1.2.x? Or doesn't it work that way? Regards, Erik. sent from my phone On Jun 16, 2013 2:19 PM, "Anton Vodonosov" <avodonosov@yandex.ru> wrote:
I have inspected ~/cl-test-grid/work-dir/agent/logs/agent.log It is now running tests for abcl-1.3.0-dev-svn-14545-fasl42-linux-x86.
16.06.2013, 15:18, "Erik Huelsmann" <ehuels@gmail.com>:
Hi Anton, Just a second ago I started a testrun in a screen session. Could you verify it's actually running and running correctly? (I'm running a trunk version, because they differ only by a few (irrelevant) commits and I didn't know how to start 1.2.x)
Bye,
Erik.
On Sun, Jun 16, 2013 at 12:41 PM, Anton Vodonosov <avodonosov@yandex.ru> wrote:
16.06.2013, 14:27, "Erik Huelsmann" <ehuels@gmail.com>:
If tests are successful, I'll backport to 1.2.x and start a testrun on my testgrid machine. (I may need help there though since Anton has always operated it so far.)
Let me know if you have any questions
-- Bye, Erik.
http://efficito.com -- Hosted accounting and ERP. Robust and Flexible. No vendor lock-in.
Another instance can be started in parallel. Although if the machine is not multi-core, it is the same as executing the two tests sequentially. Where to get the 1.2.x version? 16.06.2013, 16:52, "Erik Huelsmann" <ehuels@gmail.com>:
That's what I wanted to start. Can we set up a separate 'run-agent' for 1.2.x? Or doesn't it work that way?
Regards,
Erik.
sent from my phone On Jun 16, 2013 2:19 PM, "Anton Vodonosov" <avodonosov@yandex.ru> wrote:
I have inspected ~/cl-test-grid/work-dir/agent/logs/agent.log It is now running tests for abcl-1.3.0-dev-svn-14545-fasl42-linux-x86.
16.06.2013, 15:18, "Erik Huelsmann" <ehuels@gmail.com>:
Hi Anton, Just a second ago I started a testrun in a screen session. Could you verify it's actually running and running correctly? (I'm running a trunk version, because they differ only by a few (irrelevant) commits and I didn't know how to start 1.2.x)
Bye,
Erik.
On Sun, Jun 16, 2013 at 12:41 PM, Anton Vodonosov <avodonosov@yandex.ru> wrote:
16.06.2013, 14:27, "Erik Huelsmann" <ehuels@gmail.com>:
If tests are successful, I'll backport to 1.2.x and start a testrun on my testgrid machine. (I may need help there though since Anton has always operated it so far.)
Let me know if you have any questions
-- Bye, Erik.
http://efficito.com -- Hosted accounting and ERP. Robust and Flexible. No vendor lock-in.
Have you seen the abcl-1.3.0-dev-svn-14545-fasl42-linux-x86 results already? I've published the diff report here: http://common-lisp.net/project/cl-test-grid/abcl/abcl-diff13.html What deserves attention is exscribe crash. Reproduce like this: rm -rf ~/.cache/common-lisp/ java -jar lisps/abcl/dist/abcl.jar --eval '(ql:quickload :exscribe)' > exscribe.load.log 2>&1 less exscribe.load.log The compiler crashes due to java.lang.Error: "ABCL Debug.assertTrue() assertion failed!"
Exscribe issue fixed last night - and greatly simplified the crashing function in the process! Bye, Eeik sent from my phone On Jun 17, 2013 8:15 PM, "Anton Vodonosov" <avodonosov@yandex.ru> wrote:
Have you seen the abcl-1.3.0-dev-svn-14545-fasl42-linux-x86 results already?
I've published the diff report here: http://common-lisp.net/project/cl-test-grid/abcl/abcl-diff13.html
What deserves attention is exscribe crash.
Reproduce like this: rm -rf ~/.cache/common-lisp/ java -jar lisps/abcl/dist/abcl.jar --eval '(ql:quickload :exscribe)' > exscribe.load.log 2>&1 less exscribe.load.log
The compiler crashes due to java.lang.Error: "ABCL Debug.assertTrue() assertion failed!"
Diff for the revision 14556 (and by the way, new quicklisp 2013-06-15): http://common-lisp.net/project/cl-test-grid/abcl/abcl-diff14.html As you see, there are some regressions in this revision. Some of them are caused by the same error "The assertion (EQ (CAR (SECOND JVM::FORM)) (QUOTE LAMBDA)) failed." Also, there are some crashes (tinaa, metatilities, etc). Reproduced tinaa crash with rm -rf ~/.cache/common-lisp/ java -jar lisps/abcl/dist/abcl.jar --eval '(ql:quickload :tinaa)' > tinaa.load.log 2>&1 less tinaa.load.log reveals "java.lang.Error: ABCL Debug.assertTrue() assertion failed!" Best regards, - Anton 18.06.2013, 09:40, "Erik Huelsmann" <ehuels@gmail.com>:
Exscribe issue fixed last night - and greatly simplified the crashing function in the process!
Bye,
Eeik
sent from my phone On Jun 17, 2013 8:15 PM, "Anton Vodonosov" <avodonosov@yandex.ru> wrote:
Have you seen the abcl-1.3.0-dev-svn-14545-fasl42-linux-x86 results already?
I've published the diff report here: http://common-lisp.net/project/cl-test-grid/abcl/abcl-diff13.html
What deserves attention is exscribe crash.
Reproduce like this: rm -rf ~/.cache/common-lisp/ java -jar lisps/abcl/dist/abcl.jar --eval '(ql:quickload :exscribe)' > exscribe.load.log 2>&1 less exscribe.load.log
The compiler crashes due to java.lang.Error: "ABCL Debug.assertTrue() assertion failed!"
Hi Anton, On Fri, Jun 21, 2013 at 11:42 AM, Anton Vodonosov <avodonosov@yandex.ru>wrote:
Diff for the revision 14556 (and by the way, new quicklisp 2013-06-15): http://common-lisp.net/project/cl-test-grid/abcl/abcl-diff14.html
As you see, there are some regressions in this revision. Some of them are caused by the same error "The assertion (EQ (CAR (SECOND JVM::FORM)) (QUOTE LAMBDA)) failed."
These I still have to look at.
Also, there are some crashes (tinaa, metatilities, etc). Reproduced tinaa crash with rm -rf ~/.cache/common-lisp/ java -jar lisps/abcl/dist/abcl.jar --eval '(ql:quickload :tinaa)' > tinaa.load.log 2>&1 less tinaa.load.log reveals "java.lang.Error: ABCL Debug.assertTrue() assertion failed!"
But these are now fixed on trunk. I'll consider backporting in a bit. Bye, Erik.
23.06.2013, 07:42, "Anton Vodonosov" <avodonosov@yandex.ru>:
22.06.2013, 23:35, "Erik Huelsmann" <ehuels@gmail.com>:
But these are now fixed on trunk. I'll consider backporting in a bit.
What do you mean by backporting? Is there a release branch for 1.2.x?
Indeed, there is. I have checked out and build svn repo http://abcl.org/svn/branches/1.2.x/ at cl-test-grid.cloud.efficito.com machine, in the directory ~lisps/abcl-branch-1.2.x We can test when necessary. If you want to run tests yourself, you can update and rebuild this branch, and then run ~/cl-test-grid/run-agent-1.2.x.sh Best regards, - Anton
participants (2)
-
Anton Vodonosov -
Erik Huelsmann