ChunkBrowser Question

Ian,

I fried an image today, but had a very recent backup.  The problem was that
I changed the name of a class, by let's say creative means, and left the
image in a confused state.  My first though was to copy the methods to a new
class and then delete the offending class, but it didn't work.  However, it
put a delete chunk in the change log - hence my question below.

It almost looked like I had repaired the damage, but some prodding proved
otherwise.  I made a backup of the bad image/change/sources,  restored the
recent backup, and used the CB in it to open the tail of the change log from
the broken image.

All seems well, but should I be worried that the class delete chunk that
refers to the bogus name is checked?  A quick glance at #identityAsClass
make me think that it's simply that a class by the offending name does not
exist.  If there's more to it, please say so.

Thanks for the ChunkBrowser!!!!

Bill

--
Wilhelm K. Schwab, Ph.D.
[hidden email]

Bill,

> All seems well, but should I be worried that the class delete chunk
> that refers to the bogus name is checked?

No, the tick is the icon that should be displayed.  The idea for the
comparison flag is that you get a tick if the chunk and the current
image match.  In this case the chunk said "delete this class" so, as the
class wasn't present in the current image, the tick was shown.  As there
is no way of determining whether a class has ever existed in the current
image then there is not a lot else that can be done.

Thanks for the question though - it made me notice a little bug related
to refreshing the source view.

--
Ian

Ian,

> > All seems well, but should I be worried that the class delete chunk
> > that refers to the bogus name is checked?
>
> No, the tick is the icon that should be displayed.  The idea for the
> comparison flag is that you get a tick if the chunk and the current
> image match.  In this case the chunk said "delete this class" so, as the
> class wasn't present in the current image, the tick was shown.

Sounds good.


> As there
> is no way of determining whether a class has ever existed in the current
> image then there is not a lot else that can be done.

It's a great solution.  Ordinarily, I wouldn't have thought twice about it,
but _any_ appearance of the damaged name was cause to get curious - I
wouldn't want to repeat the disaster w/o realizing it.


> Thanks for the question though - it made me notice a little bug related
> to refreshing the source view.

Thanks for the ChunkBrowser!

Have a good one,

Bill

--
Wilhelm K. Schwab, Ph.D.
[hidden email]