When Apple said they were introducing an "App Store" to OSX a-la the iPhone, iPad, etc. there was much ranting and foaming at the mouth...Many geeks proclaimed the death of Mac OS X as we know it. "Apple is iPhone-izing the Mac!" they cried.

They're largely wrong, at least in all the important ways.

Here are the salient points:

• Discovery - people are *far* more likely to purchase things that are available from a single, readily accessible source. Having a veritable floatilla of awesome software that's discoverable right from the Apple menu of every Mac means that if you're a publisher, people are FAR more likely to try your software. I myself found a free copy of Autodesk Sketchpad in the app store that I am very much enjoying. I might end up buying the commercial product as a result - possibly through the app store! (See caveats below for why I might choose the shrink wrapped version.)

• Ease of installation - just like on the iPhone and iPad, you click "Buy" or "Install" and *BOOM* it's on your machine, happily bouncing in your dock. For non technical people, this is a godsend. They do not want to know about .dmg files, dragging apps to the Applications folder, or the like. Yes, I know, this is VASTLY easier than the folks in the Windows world have it with their highly complex installers, but even still.

• One Stop Shopping for updates - Sure, lots of devs these days are bundling some kind of update mechanism into their apps, but this is once again - seamless and easy to use for non technical people.

Of course, every silver lining has a cloud attached, so there are a few down sides:

• Inflexible upgrade path - Apple has imposed rules around publishers app store offerings, I don't know all the details, but I do know that publishers are not happy. They want total control of how they monetize the sale of upgrades for their customers, and to a certain extent I can't blame them. Over all I think the net winner here is the consumer, so we'll see how many publishers sign up and grudgingly cope with the new rules.

• Restrictive rules about permitted content - Nothing sexy, nothing violent, no Java or Rosetta, just like the IOS app store.

From my perspective, people who object to the app store on OSX are on shaky ground, because unlike the IOS world where the app store is the ONLY way to get apps, the app store on OSX isn't. You can still install apps the old fashioned way, and Apple has given no indocation that they intend to do this now or at any time in the future. I don't think it's in their best interest to do this, and I think people who are predicting imminent doom really need to get their collective head checked.

Then again, geeks do have a propensity for hysterics.

To sum up, I think everyone (publishers, consumers, and the wider Mac community) has everything to gain and nothing to lose from the app store, and I for one welcome the addition.

I hate to write a negative blog post about a tech publisher's ebooks program, but I'm afraid they've driven me to it...

Not that I don't sympathize, writing a book is from what I understand Really Hard, but as a consumer if I've signed up to buy something, I expect a completed product to come out the other end. To be fair, Manning does offer either a refund or an exchange plus a bonus free e-book, but most of the truly hot topics that I found interesting were MEAP books again, so the cycle repeats.

Secondly, Manning uses the Yahoo Store for their ebook delivery. Wow, what a nightmare. You get n days to download your book once they send you the email, and then POOOF your purchase goes off into the ether. Compare and contrast with O'Reilly and Associates and the Pragmatic Programmers (two publishing houses I can't recommend enthusiastically enough.) where you can re-download your purchases anytime, for *life*.

If you buy early access titles from either of these publishers, you are not chained to your email for fear that the ticking time bomb of your download notice will expire when you're not looking.

In general, I think Manning really needs to take a long, hard look at its store front user experience. I'm not saying I'll never buy a title from them again, but I am saying it's a HECK of a lot less likely that I will, and I'll always choose pragprog or O'Reilly if I have a choice..

So, I finally did it. After a nearly non stop marathon reading of the last three books in the series (I wanted to read all the books before I saw the movies, and I'm very glad I did this. As always, the books have way more depth.), I've come to the following conclusions...

I happen to quite like the series, but in order to enjoy it as I have I found it necessary to keep my fanboy hat firmly in its place the entire time.

I could poke a thousand holes in the series, but what it all comes down to is this - it's a totally whimsical universe, devoid of the kind of carefully construed logical frameworks that hard core fantasy and scifi fans clamor for.

Let's take a look at magic just as a for instance. In the Potter-verse, spells are formulaic incantations that produce ambiguous results. Compare and contrast this with something like Jim Butcher's Dresden files, where things like the laws of physics come heavily into play.

In the end, the Harry Potter series isn't primarily about magic, although wizards and witches fill the pages of the book, it's about good (Harry) in the never ending struggle against evil, and the terrible price that those who stand up against evil must pay.

In my humble opinion though, it's a heck of a ride. Very entertaining from start to finish, and in the end analysis couldn't all of us do with a bit of an escape into whimsy once in a while anyway?

While at first blush the series appears to be young adult reading, good only for kids too young to be exposed to the evils of the world, and the first books are indeed very light hearted and very much fairy tale like in nature, but as the series progresses it becomes much, MUCH darker. Evil and its outcome is depicted in much sharper contrast, and death looms around every corner.

I would encourage any hard core fantasy fan who read the first book, said "meh" and walked away to give it another try. Look at it from a different perspective, and just try to enjoy the story for what it is. In the end I predict you will not be disappointed.

April 2012 Update: Nowadays we useCheffrom Opscode solutions to manage our crontabs, and just about everything else in our enterprise infrastructure. It rocks :)

Time and time again over the years I've dealt with the same problem - who took a random pot shot at some critical user's crontab file and deleted things without asking?

Then you look at the crontab for the user in question and realize that the lines invoking the script that used to be there have either been deleted or commented out. What the? Who did this, and why?

Git to the rescue! With the help of a simple Bourne Shell script, you can keep your crontab managed so you'll not just be able to see who changed what and when, but if you have the Git hook installed to send mail on commits, you can be notified of those changes in real time. Pretty cool eh? :)

Since crontab has no built in security precautions other than requiring you to BE the user whose crontab you're submitting, we can't lock people not using this script out, but if you tell everyone that changes they make outside of the script may be summarily ignored and overwritten (and put something to that effect in the comment block at the top of your crontab) you should be in good shape. The script will compare what's in Git with what's currently installed in cron, and if there are discrepancies it will give you a chance to cleanly exit and resolve them, or allow you to ignore them and roll forward with editing and submitting what's in Git.

Here's the script. It assumes you're running as your regular user and have sudo privs to the user whose crontab you wish to edit. It also assumes you've created a Git repository called system_cron.git

To set it up, just edit reponame and gitrepo to appropriate values for your site and copy the script to somewhere folks can access it in their PATH.

To use it, just invoke it with the user whose crontab you want to edit - for example:edit_crontab.sh build

#!/bin/sh

export tmpdir="/tmp/crontab_$$"

if [ $# -lt 1 ]; then
    echo "Usage: $0 "
    exit 1;
fi

if [ -z "$EDITOR" ]; then
    echo "No editor found.  Using vim."
    export EDITOR="/usr/bin/vim"
fi

crontab_user=$1

crontab_file="`uname -n`-$crontab_user.crontab"
echo "crontab_file=$crontab_file"

git clone /home/git/system_cron.git $tmpdir
cd $tmpdir

sudo -u $crontab_user crontab -l > currcrontab_$$

if [ $? -ne 0 ]; then
    echo "sudo to user $crontab_user failed! Do you have sudo privs?"
    exit 1;
fi

diff=`diff currcrontab_$$ $crontab_file`

if [ $? -ne 0 ]; then
    echo "Currently running crontab for $crontab_user differs from Git!"
    echo "Here are the differences:"
    echo $diff
    echo
    echo -n "Continue editing / submitting what's in Git? (Y/n): "
    read yesorno
    if [ "$yesorno" != "Y" ]; then
        echo "Very good.  Exiting."
        exit 1;
    fi
fi

$EDITOR $crontab_file

echo "Here are your changes:"
git diff --exit-code $crontab_file
if [ $? -eq 0 ]; then
    echo "No changes made.  Not submitting anything."
    exit 1;
fi

echo -n "Submit these changes to Git and crontab? (Y/n): "
read yesorno
if [ "$yesorno" != "Y" ]; then
    echo "Your changes are in $tmpdir/system_cron/$crontab_file."
    echo "Please clean up this directory when you're done with it."
    exit 1;
else
    git commit $crontab_file
    if [ $? -ne 0 ]; then
        echo "There was a problem committing your crontab to git!"
        exit 1;
    fi
    git push origin master
    if [ $? -ne 0 ]; then
        echo "There was a problem pushing your crontab to git!"
        exit 1;
    fi

    # if we made it this far. We're all good.  Install that puppy!

    echo "Installing your crontab."
    sudo -u $crontab_user crontab $crontab_file
    if [ $? -ne 0 ]; then
        echo "ERROR! Your changes were NOT installed! Something went wrong."
        exit 1;
    fi
fi

echo "Cleaning up tmp directory..."
#rm -rf $tmpdir

As I mentioned here previously, my last big job search after I got laid off from MIT was decidedly Not Fun...

I got turned down by 2 or 3 prospective employers because of this, which is something that hasn't happened to me for a very long time.

So, when I took said job, I steeled myself for the worst and called my old school, Worcester State College.

The result was not at all what I was expecting. I still only need two courses to get my bachelors degree!

I got all the paperwork filled out, jumped through all the necessary hoops and found out that I (still) need a science, and a science with a lab in order to graduate.

That information in hand, I went casting around for a science course I could take. Not easy, many of them required calculus, and those that didn't hab lab periods smack dab in the middle of my work day.

I found one course, Astronomy, that only claimed to require College Algebra and was taught Tuesday and Thursday evenings. So, I leapt.

BIG mistake.

The course started a week and a half ago. Since that time, I have poured ~30 hours and a number of sleepless nights into it, and unfortunately I have to admit that there is eactly zero chance of my passing this course. No amount of effort, lack of sleep, or heroics short of quitting my job (Freaking unlikely since I'm INCREDIBLY happy where I am now. I escaped from Windows enterprise hell and am working for an amazing company, Blue State Digital.) will save me.

I had been expecting the math to kick my ass, but in the end analyais, I could handle that just fine! There wasn't a single bit of algebraic manipulation I couldn't get past with a little work and some Google thrown in for good measure.

It was the formulas! Crazy, INSANE formulas crawling across the page and eating my poor addled brain like some kind of scene out of Dawn of the Dead.

F_ev / F_em = ( G M_v M_e / R_ev² ) / ( G M_m M_e / R_em²)

I have to withdraw from the battlefiend this time - I've lost the skirmish, but I am bound and determined to win the war!

I'll take some time to recover, take a fresh look at the problem, and talk to my academic advisor to see if I can apply some 'life credit' or something, or at the very least not bludgeon myself senseless over my lack of higher math, so I can finish up the silly degree and get on with my life!

I received a piece of E-mail from Amazon a few days back apologizing about the fact that my pre-order of the latest Jim Butcher Dresden Files novel had been canceled, offering me a really cheap price on the print edition.

A couple of days and a great deal of "WTF?" later (That's What The French, this being a family friendly blog and all :) I trotted on over to Jim's website and found an explanation.

Essentially, Amazon and Penguin are bickering over the sale price of the book.

I would *happily* pay full hard cover list for the Kindle edition of the newest Dresden novel.  Jim Butcher is my favorite urban fantasy author, and when these books come out I get them the very first *instant* I can and then become some sort of literary crack fiend, obsessively trying to find excuses to crawl off into a corner and page flip.

Amazon has finally come out with the details around their Kindle Development Kit (KDK) you can read the details here.

However, I'm a bit disappointed. I realize that the Kindle is meant as a light weight device, but I have always found J2ME's I/O model to be rather cripplingly baroque in so many ways.

Looking over Amazon's APIDoc however, it does seem like they've built their own infrastructure atop J2ME, so perhaps they've made it easier and more pleasant to code for.

I'm torn between wanting to get in on the ground floor and not wanting to invest the time and energy in learning a technology (J2ME) which I feel has gone way past its expiry date.

Anyway, we'll see but I'll definitely be keeping a close eye on this one.

For the last year or so I've been working at a 100% Windows based enterprise development shop.   They ship a fairly large and complex application for health insurance providers, which draws on many of the core technologies in the Microsoft enterprise stable.