Thanks Drew and Darren!! I'll leave them in and add a comment at the beginning of the imports.
John Sent from my iPad On Mar 5, 2012, at 1:31 AM, Darren Kenny <[email protected]> wrote: > On 03/03/2012 01:27, Drew Fisher wrote: >> John, >> >> 7041360 is, I think why you can't remove those imports. >> >> From the CR: >> >> "If we change code to load the checkpoints with the 'dotted' version of >> the path, the imp.find_module() will only >> work if we've already imported that module explicitly in to our code." >> >> I know that Darren can confirm if this is the reason those imports are >> there. Don't remove them yet, please. > > This is exactly why those imports are there - if you run with the engine in > debug mode it isn't possible to catch any exceptions from these modules > using specific names due to the above bug. > > Now, we have changed things so that we're no longer running the engine in > debug mode so things will probably run without them now, but I feel it's > worth having these around should we have any need to since later it would > be very confusing. > > Possible a comment is worth while here instead of removing - they will be > imported anyway later by the engine... > > Thanks, > > Darren. > > >> >> -Drew >> >> On 3/2/12 6:21 PM, John Fischer wrote: >>> Drew, >>> >>> I have removed them. I am now working on creating a new image >>> to test. Now if only the servers we reachable. >>> >>> John >>> >>> Sent from my iPad >>> >>> On Mar 2, 2012, at 4:39 PM, Drew Fisher<[email protected]> wrote: >>> >>>> I don't think we can remove those. I remember that this has something to >>>> do with the engine and how the checkpoints are registered. >>>> >>>> Darren: does this ring a bell to you? >>>> >>>> -Drew >>>> >>>> On 3/2/12 12:33 PM, Mary Ding wrote: >>>>> John: >>>>> >>>>> There were also a lot of unused imports in auto_install.py according to >>>>> pylint and it will be nice if they can be fixed. >>>>> >>>>> >>>>> http://indiana-build.us.oracle.com/job/slim_code_cleanliness/Pylint_Audit/? >>>>> >>>>> Mainly it is W0611 and the following imports are unused: >>>>> >>>>> Unused import boot >>>>> Unused import device_config >>>>> Unused import boot_archive >>>>> Unused import instantiation >>>>> Unused import initialize_smf >>>>> Unused import transfer_files >>>>> Unused import apply_sysconfig >>>>> Unused import INSTALL >>>>> Unused import update_dumpadm >>>>> Unused import DataObject >>>>> Unused import discovery >>>>> Unused import varshared >>>>> Unused import setup_swap >>>>> Unused import create_snapshot >>>>> Unused import TargetSelectionZone >>>>> >>>>> >>>>> >>>>> On 03/02/12 11:28, Mary Ding wrote: >>>>>> John: >>>>>> >>>>>> My comments are just nits: >>>>>> >>>>>> In auto_install.py, if you use pylint 0.23, it will complain about the >>>>>> following lines: >>>>>> >>>>>> import os >>>>>> import os.path >>>>>> >>>>>> W0404 Reimport 'os.path' (imported line 10) >>>>>> >>>>>> It will be good if you can fix it. >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> On 03/01/12 17:37, John Fischer wrote: >>>>>>> All, >>>>>>> >>>>>>> Can I get a code review for: >>>>>>> >>>>>>> 7097012 AI should output to /dev/console as well as >>>>>>> /system/volatile/install_log >>>>>>> http://monaco.us.oracle.com/detail.jsf?cr=7097012 >>>>>>> >>>>>>> The webrev is located at: >>>>>>> >>>>>>> >>>>>>> https://cr.opensolaris.org/action/browse/caiman/johnfisc/7097012-AI-output-console/ >>>>>>> >>>>>>> The basic fix is to have all checkpoints produce output. This is done >>>>>>> by passing in >>>>>>> a kwargs show_output to engine.register_checkpoint(). In addition to >>>>>>> this fix I have >>>>>>> cleaned up the output for AI changing several things from debug to info >>>>>>> and visa versa. >>>>>>> Furthermore, there was some duplicate output during the early stages >>>>>>> due to fractional >>>>>>> percents. This was fixed by saving the previous non-time progress and >>>>>>> comparing it >>>>>>> with the current non-time progress. >>>>>>> >>>>>>> I have built an AI iso image using distro_const, setup and installed a >>>>>>> new host. Ethan >>>>>>> checked on a zones setup. I also ran the test suite. All testing was >>>>>>> as expected. The >>>>>>> code is pep8 and pylint clean, improved or unchanged. >>>>>>> >>>>>>> Thanks, >>>>>>> >>>>>>> John >>>>>>> _______________________________________________ >>>>>>> caiman-discuss mailing list >>>>>>> [email protected] >>>>>>> http://mail.opensolaris.org/mailman/listinfo/caiman-discuss >>>>>> _______________________________________________ >>>>>> caiman-discuss mailing list >>>>>> [email protected] >>>>>> http://mail.opensolaris.org/mailman/listinfo/caiman-discuss >>>>> _______________________________________________ >>>>> caiman-discuss mailing list >>>>> [email protected] >>>>> http://mail.opensolaris.org/mailman/listinfo/caiman-discuss _______________________________________________ caiman-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/caiman-discuss
