Raw insanity

Raw digital camera formats were created unique to each manufacturer to drive highly technical photographers insane trying to find raw 'development' nirvana. About every six months I select a couple of my recent photographs and run them through all the available raw processor's I can get demo's for or own. I then 'pixel peep' the results which means examining the results in great detail at very large blow-ups, often to the point where one can see the pixels.

Since each of these raw developers are software applications they have vastly different ideas from their engineer creators as to how any of us want to process our photo's. Then each engineer or software team has their own idea's as to what we understand and how they can deliver controls which will help us get the optimal results.

I can have two strategies for running each program's adjustments. I can take the programs 'default' settings or I can try to make each program produce an optimal result by adjusting the numerous and not always named the same thing controls.

I am then left with multiple 'versions' of the same image. They are very hard to compare, because despite my best attempts to make them at least have the same level of brightness, contrast and dynamic range they don't match. And then there is the process of 'comparing'. Absolutely the best software for doing comparing is for Windows, the FastStone image viewer which allows one to compare up to 4 photo's. Your screen is split into even sections for each photo and then you can scroll around and blow up the photo's, each action is synced in each window, so you are always looking at the same thing in each of 4 different windows. FastStone is only one of two software packages I have found, Windows, Mac or Linux that does this with up to 4 photo's. The other is also Windows based and is a asset management system called IdImager.

Back to the comparison strategies. There are arguments for either of them depending upon what kind of photographer you are. If you take 100's of shots at a time, which digital cameras really encourage, then doing the comparisons using 'default' settings might be a good strategy. However if you either take a few photo's or are very good at pruning your 100's into a few, then trying for 'optimal' results might be a good comparison approach. I do both, of course, since I can't make my mind up :)

If you are looking for some 'winner' from me you are not going to get it. I find a different winner each time I do the tests and sometimes a different winner with each photo. So maybe it doesn't really matter. But this is exactly the kind of endless technical manipulation inside a GUI, which changes parameters sometimes understandably and sometimes not, but always with noticeable results, that is obsessive but ultimately leads to madness.

Oh well, go to go now and check out the latest 'gui' for dcraw....

The Empire Strikes Back, part 10

Or at least it feels that way. My previous post on our CIFS breakthrough would the analogous, in movie plot terms, to Star Wars, A New Hope. This week, it's the empire strikes back as we still can't use our system. This is because of a, so far, unknown idmapper problem.

Of course, the empire here is our Active Directory. The reason I say part 10 is (to continue the movie metaphor's) we seem to be stuck in a plot that mixes Groundhog Day with the first two (chronologically released) Star War's movies.

We start out as a small band of highly technically equipped rebels trying to restore the Republic by integrating all OS's into a common's of functional equality. We struggle with small set backs and then a big revelation is made to us that puts us into despair. And then it starts all over again. We never do get to a play out of the "Return of the Jedi".

The "force" of course is computer technology, very powerful and fairly mysterious. Full of young masters who can manipulate it without quite knowing why and old masters who think the why is embedded in the past events they were part of.

Ok, I have stretched this analogy about as far it can go without becoming too ridiculous. One good thing did come out of this latest round and that is we found more people documenting what they are doing with OpenSolaris, CIFS and ZFS here:
http://tinyurl.com/b3uy7o and here: http://tinyurl.com/ak8mxg .

This is promising as it shows the opensolaris ecosystem is growing. We still don't have anything like an ubuntu or Centos for it, but those took a long time to develop in the linux ecosystem. It's a matter of numbers, how many people are using this stuff and how many of them will contribute back to the community. In this regard SUN has to tread a fine line by keeping the licensing open enough to attract a fair portion of open source advocates and putting back their own developments, while still finding a way to package software up such that customers are willing to pay for it, ala RedHat.

a last-second, half-court prayer...and drained it! Nothin' but net!

These are the words our technical guru used to describe Nexenta's delivery of a fix for CIFS bug: CR 6764696. Only one other detail remained - that was doing the Domain join using a newly created ID with full rights over a pre-created computer container for the server. Why the OU admin ID, which creates this new id can't do the join is a mystery that maybe Nexenta will reveal.

So, congratulations and deep thanks go to Nexenta for this fix.

I didn't explain the last-second allusion though. And it was this. I had given our technical team one week more to look for solutions among the SAMBA and SUN CIFS server options. This morning, in about one hour, we are having a meeting (called by my boss) which is scheduled to discuss what we would do given failure......the options here were ugly.