Warning re: DiskWarrior and SuperDuper’s clone-on-mount

Please don’t get me wrong: both of these are superior products, and I have and use and love both.

That said, SuperDuper (SD) offers a “clone-on-mount” (auto-clone) option that can wreak havoc when and if you decide you want to run Diskwarror (DW) on the destination drive.

That is, if you mount the destination drive, and SD starts cloning something to it, you may have problems if you try to run DW on that drive to repair it.

Here’s the issue:

Source drive S, and Destination drive D

User wants to have DW repair drive D.
Drive D is also the target of an automount which clones S to D.

Drive S has files A, B, C, D and E.
Destination drive D has files A, B, F and Q.

User inserts D
SD clone starts and User cancels clone.

DW looks at drive D, and sees A, B, F and Q.
DW unmounts the drive to work on it.
DW creates a replacement directory listing A, B, F and Q… remounts drive D, and then pauses. DW does NOT write the directory, because it’s waiting for the user to OK it.
Seeing that D has remounted,  the auto-clone kicks in and SD copies drive S to drive D.

here’s the crux of it:

after the SD clone, D now has files A, B, C, D and E. (not A, B, F and Q, which is what DW thinks it has)
User says to DW, “Yep, install the new directory.” and DW dutifully replaces the directory to say that D has files A, B, F and Q.

Drive D directory is now totally hosed.

There is basically no way around this except to start up SD and turn off all the auto-clone scripts and then quit SD before you run DW.. Now you can run DW without issues… just remember to re-enable the scripts in SD after DW is done.

On the death of the Finder

OK: the title is a bit over the top….. the finder is not dead… but it’s in its later years. I’m prompted to write this as the release of Lion is impending.

Many of us understand the Finder/Desktop metaphor as the “filing cabinet” it’s intended to emulate. Others (although this blows my mind) don’t understand filing/organizing at all. I suppose organization is a trait of some folks, while for others it’s literally beyond comprehension. My dad, for example, filed things in cabinets obsessively, but could not grasp the same concept on the Mac.

Apparently, that second group of people is the majority. For them, the Finder was a nightmare, and virtually everything in their computer was on their desktop.

For those of us who pride ourselves on our ability to organize and retrieve information, our directory (folder) structure are not only a source of pride, but something we eventually begin to obsess over. We become uncomfortable when something isn’t file properly and promptly. And rightly so, since when retrieving an item, that obsession has served us well over the years.

But for both the filing-obsessed and the filing-phobic, there is, indeed, a better way. The latter will have already embraced it. This blog-entry is for those filing-champions who start jerking and twitching at not knowing exactly where a document is filed.

The point of the Finder is, um, to “find” things… so that we can “use” them. The point of any filing system is to make it easier to retrieve a given item … hence the extra effort required up-front to do the filing in the first place.

Yes, it’s much easier and faster to simply throw everything you own into one large box… but picking out one given item (especially after memory has faded) can then take hours to do… or even become virtually impossible. So instead we take a bit of extra time to use multiple storage places, and assign a meaning that we personally find useful to each one. Then we file the item therein.

If those boxes are too large (the categories too broad) then it’s still time-consuming to sort thru them to find a given item… so we use sub-categories – (boxes within boxes).

My point is that this “do the work up-front filing system” is highly personal, and reasonably effective.

Enter the computer.  We simply carried over that old physical filing system technique to the desktop metaphor … and it works … that is… it works just as well as the physical one does… but with all the attendant shortcomings along with it. (Folders fail if they contain too much; the titles are not descriptive; you put something in the wrong place; you forget the name of the folder, et al.)

The computer however, can make the process more efficient, both up-front (just toss everything in one box) and on retrieval.

The difference is that we are just specifying the search criteria after the fact instead of during the filing process.

In the database world, we can find something if we remember almost anything at all about it. We can increase our comfort/familiarity level by using “smart folders” which look like our old familiar directory/folders, but are vastly superior.

We can drag a document to the “main box” and drop it (equivalent of putting everything into one box) or we can drag and drop it on a set of keywords (with exactly the same effort) and further enhance our search (just like sub-folders, but with less effort.)

Further, a database approach is not only faster on the input side, and the retrieval side; not only does it contain “sub-folders (keywords)” but it offers something no folder/directory setup can: items can be “in multiple places” at the same time.

You can have a smart folder for clients; one for projects; one for billing… and each one can reference a single document.

The whole point of this however is that we spend 95% of our time -using- a document, and where it is physically located is relevant only to retrieving it (so that we can use it.)

And, if you get jittery, you can simply use the contextual menu to select “reveal in finder” (yes: even in iTunes) and you’ll be taken directly there… you don’t even need to navigate the sub-directories manually.

Having been at this for 36 years, I too was a bit hesitant, even uncomfortable, at “giving up” my carefully constructed and personally meaningful directory structures for many things.

However, over the past several years, I’ve not only over-come those feelings, but actually embraced the database paradigm. I use DevonThink Office for text documents; LightRoom for photos; iTunes for music and so on.

When/if I need things organized into a project folder (for backup to an optical disc, for example) I simply choose the smart folder that contains it, and either select “export” or better yet “burn.”

Where the data is located is important only in the context of being able to retrieve it, and in that respect, databases are leagues ahead of the multiple shoebox approach.


Post Script:

I guess I should have pointed out that a directory/folder -is- a record in a database. There are no “places” on a hard drive that are “folders” that “contain” your files. Your files are likely scattered in pieces all over the place on your drive; it’s much more like trail-mix than a box of rasins and a box of nuts.

Directories (aka folders) exist only in the database that is called the “catalog” and “B-Tree” (among others.) When you click on a file in the Finder, that database is consulted, the constituent bits of your file are found, and presented.

The actual, technical different between iTunes/DevonThink et al, and the Finder is far less than one might think, and is mostly just the user interface.