Danek Duvall wrote:

> 
>   - 216: I'd put this down closer to where it's actually used.
> 
I had considered getting rid of the tuple on the return; I've now
done so and this is gone.

>   - 232: "set()"
ok
> 
>   - 233: seems like this loop might be something useful to have in img.
>     Perhaps an XXX comment for now?  Is the equivalent generator expression
>     any better here?
> 
>         pkgs.update((
>           p
>           for a in pargs
>           for p in all_pkgs
>           if a in str(p)
>       ))
> 
>     Up to you.
taken
> 
>   - 235: a in str(p)
> 
ok

>   - 236: pkgs.add(p) ?
> 
multivalued returns possible....

>   - 238: no need for this
ok
> 
> depend.py:
> 
>   - 65: Why do we need this?  Does get_version_installed() ever return a
>     version that isn't installed?  If so, that seems broken.
> 
I assume you mean the try?  It's needed since get_version_installed can 
toss an exception....

> driver.py:
> 
>   - 194: need line break before this.
fixed
> 
>   - 199: if the file doesn't exist, this'll throw an exception, right?
fixed
> 
>   - 211: check for multiple instances in name_to_major?
fixed
> 
>   - 220: four-space indent (continuation)
> 
fixed

>   - 223: no indent
fixed
> 
>   - 226: I think this ought to be okay on one line, if it fits.  Though it
>     might be worthwhile spelling out which aliases are missing, and which
>     aliases have been added.
> 

I've added more explicit error logging

> file.py:
> 
>   - 142: In directory.py, you simply put [e] in the second element of the
>     tuple.  I think you probably meant to do what you do here, but I'd
>     consider putting the exception in raw, since this would allow you to
>     display the exception in a more intelligent manner by the caller,
>     particularly in a GUI.
> 
> generic.py:
> 
>   - 341: "Returns True if the action is correctly installed in the given
>     image".  I don't think it's necessary to mention the need for child
>     classes to override the method.  It would be appropriate to throw a
>     NotImplementedError, if you wanted.
> 
>   - It strikes me that you could drop the first element of the return tuple
>     and just return the list of problems.  I think it might look a lot
>     simpler on both sides of the call.
> 
>   - 344: No need for the parens.
> 
>   - 353: space after comma
> 
> legacy.py:
> 
>   - 95: clean this up a bit.  If you can't think of anything to say, skip
>     the docstring.
> 
ok
>   - 100: What's the issue you see for upgrades?  It's probably worth being
>     explicit.  Though a better check would be to see if the pkginfo file
>     exists, since that's what needs to be there to get working what we want
>     to work, and it's not that difficult a check to make.
> 
noted that exact validation is a possibility
>   - 101: I think you need to explain what "forever" means somewhere.  Given
>     that you need to update the man page (pkg(1)), too, perhaps that's the
>     best place?
> 

What did you mean here?

>   - 106: I'd split after the "%" operator, and make the continuation indent
>     four spaces.
> 
fixed.

> link.py:
> 
>   - 95: delete

ok
> 
> unknown.py:
> 
>   - No need to add anything here.
> 

removed.

> image.py:
> 
>   - 606: "uid" instead of "nameuid"?
> 


> solaris.py:
> 
>   - 475: Is this call still necessary?
> 

yes; new additions handle only non-file types.

I've got a new webrev out containing these changes...

- Bart


-- 
Bart Smaalders                  Solaris Kernel Performance
[EMAIL PROTECTED]               http://blogs.sun.com/barts
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss

Reply via email to