Karen,

Thanks...

You are correct,  Dave has pointed it out also, I am incorrectly writing 
the "END BOOTADM" at the end of the file. I'll fix that.

Huge Thanks!

Joe


Karen Tung wrote:
> Hi Joe,
> 
> Please see my responses below.
> 
> Joseph J VLcek wrote:
>> Karen,
>>
>> Thank you for your feedback. I appreciate it.
>>
>> Karen Tung wrote:
>>> Hi Joe,
>>>
>>> I am having trouble following the code as it is written, even with the
>>> comments.  At least, the comments need to be improved.  From my
>>> understanding of the problem that need to be solved, and reading the
>>> existing code many times, I *think* the code is doing the following.
>>>
>>> --------------
>>> Find the first entry in the menu.lst file that boots the installed 
>>> rpool,
>>> and enable Happy Face Boot for that entry.  The original content of
>>> the entry will be saved to create an additional entry that doesn't
>>> have Happy Face Boot enabled.  Comments and white space from the
>>> original entry will not be included in the additional entry.  If more 
>>> than
>>> one entry exist in the original menu.lst file, only the first entry will
>>> be modified to enable Happy Face Boot.
>>> --------------
>>
>> That is correct. I can add similar text as a comment.
>>
>>
>>>
>>> Other comments I have about the code.
>>>
>>> 1) You don't need to have 2 "try".  You can just have 1 try with with 
>>> multiple "except"
>>> and the "finally"
>>
>> I check with Jack Schwartz on this. I turns out for python 2.4 the 
>> extra level is needed.
>>
> I just wrote a little test program to check.  You are right that it does
> need the extra level.  I am still surprised that we need the extra level.
> I guess I was thinking about Java.
>>
>>>
>>> 2) You seem to use "END BOOTADM" as the indication of end of the 
>>> entry.  I don't think
>>> that's a good idea, since that's just a comment.  I think one entry 
>>> should go from one
>>> title line to the next title line.
>>
>> "END BOOTADM" is not being used as the entry indication. The comment 
>> line containing "END BOOTADM" is saved and then rewritten to the end 
>> of the file.
>>
> end of the file?  I guess that's where I got confused.  I assume that 
> you just write it to the end of
> the happy face boot entry.  Why write it to end of the file?
>> The title line is being used to indicate the next entry.
>>
>>
>>>
>>> 3) The way it is coded now, the "text boot" entry will be the last 
>>> one in the menu.lst
>>> instead of immediately following the line that we have enabled Happy 
>>> Face Boot.
>>> Is that the desired behavior?  I would think that it makes more sense 
>>> to have it
>>> immediately follow the Happy Face boot line.
>>
>> I purposely put the "text boot" entry at the bottom.
>>
>>
>> Currently there is only one entry. I had initially coded this fix to 
>> rely on that but I felt it would be safer to make the code a little 
>> more robust.
>>
>> In the near future the extra "text boot" entry will not be needed 
>> because support to interrupt the graphical boot using keyboard input 
>> will provide visibility to the boot messages.
>>
> OK, this makes sense.
> 
> Thanks,
> 
> --Karen
> 
>>
>>>
>>> Thanks,
>>>
>>> --Karen
>>
>>
>> Huge thanks Karen!
>>
>> Joe
>>
>>
>>>
>>>
>>> Joseph J VLcek wrote:
>>>> Hello,
>>>>
>>>> Can two people please do a code review for bug:
>>>>
>>>> 4673 Grub text only mode is required
>>>> http://defect.opensolaris.org/bz/show_bug.cgi?id=4673
>>>>
>>>>
>>>> The webrev is available at:
>>>> http://cr.opensolaris.org/~joev/bug4673/
>>>>
>>>> This fix involves a rewrite of ICT enable_happy_face_boot
>>>>
>>>> I also fixed some begin and end comment block issues that were 
>>>> causing vim to display some of the code in the color of a comment.
>>>>
>>>> * The modules affected and tested:
>>>>
>>>> ict.py
>>>>
>>>> * Testing done on logic
>>>>
>>>> I wrote a stand alone wrapper around the changes so they could be 
>>>> exercised outside of an installation environment.
>>>>
>>>> I tested:
>>>>
>>>>    - a menu.lst file with multiple entries
>>>>    - a menu.lst file with a single entry
>>>>    - a menu.lst file with embedded comments
>>>>    - a non-existing menu.lst file, reports an error
>>>>    - a read only menu.lst files, reports an error
>>>>
>>>> * Testing done for GUI Install
>>>>
>>>> I booted a 101 live Image and used mount -F lofs to applied the 
>>>> updated version of ict.py
>>>>
>>>> After the installation I performed reboots selecting the default 
>>>> before the timeout, I let the timeout select the default and I 
>>>> selected the text boot entry
>>>>
>>>> * Results:
>>>>
>>>> The installation completed successfully and the menu.lst file 
>>>> contained the default graphical boot entry and a second, text boot 
>>>> entry and both entries provide the expected behavior.
>>>>
>>>> * Testing done for AI
>>>>
>>>> No AI testing was performed.
>>>>
>>>>
>>>> Thank you,
>>>> Joe
>>>>
>>>> _______________________________________________
>>>> caiman-discuss mailing list
>>>> caiman-discuss at opensolaris.org
>>>> http://mail.opensolaris.org/mailman/listinfo/caiman-discuss
>>>>   
>>>
>>
> 


Reply via email to