Sure, you can set a global option to keep all, or remove after 7 days – but that’s global. However, those settings should be different per feed. Some casts are purely news, and can be removed after a day or two. Others, you may want to keep until you heard them, or even forever.

I’ve been searching forever for a solution that allows me to do that, and never found anything. I was actually getting close to doing something myself. My mistake was to search for an actual podcatcher application for OSX. Instead, what I wanted was a little external tool: CastAway. It allows custom retention settings (and much more) for each individual feed.

Yes, the option flood can seem a little bit daunting, and the UI might need some work – but for a $7 piece of shareware, it’s great. And, most importantly, it does the job it promises to do, and it does it well. Finally, I’ll get the podcast flood under control.

With OSX Leopard, iCalFix has ceased to be useful. The core functionality is now part of iCal, iCalFix is incompatible with Leopard, and Apple is phasing out the method I used to actually modify iCal. As a result, many people have asked me how to remove iCalFix from their machines.

While I’ve (usually – apologies to those of you I missed!) responded via e-mail, I never got around to writing this up on my blog. Partially because I was too busy, and partially because it does hurt a bit to bury a project. Well, no use in postponing the inevitable. There are two ways to remove iCalFix. No matter which one you choose, you must reboot your Mac after you’re done – otherwise iCalFix will linger around until you finally do reboot.

1. Via Terminal.app

Before I share the code with you, please be careful. If you don’t know what this line does, or how mistyping it could delete all your data, do not use Terminal.app. For all the other brave souls out there, this line will do the trick:

sudo rm -rf /Library/InputManagers/iCalFix

2. Via the Finder

If the shell is not your natural home, you can always use Finder to get rid of iCalFix. Open your main hard drive, find the “Library” folder. In there, find “InputManagers”. In there, find “iCalFix” and drag it into the trash can. Empty the trash can after rebooting, and that’s it.

Obviously, this means I should focus on a different project. I have a couple of ideas – let’s see how things work out.

At the very least, you need one external hard drive to store your data – a backup on the same drive will do you almost no good, because in most cases data loss affects entire drives. Yes, storing on the local drive will cover you against accidental deletes of single files, but that’s not a sufficient solution. You do want a bootable backup of your entire hard drive – SuperDuper is one of the best software solutions for this.

Even better if you have two hard drives – this way, you can store one hard drive off site and update it from time to time. Especially here in quake- and fire-prone California, it’s good to know you have your data in a physically different location.

And if you run Time Machine, you want to add yet another hard drive – it’s not a bootable backup, and booting from your installation CD with a subsequent restore via Time Machine is not something I’d look forward to.

That’s a ton of hard drives – where can you get them for an affordable price? Well, if you’re like me, you’ll have a couple old hard drives lying around. All you need to do is buy an external enclosure – I picked the Rocketfish Enclosure because it’s red, shiny, and cheap – and put in your old IDE hard drive.

Alternatively, big hard drives are cheap these days – I just bought a 500 GB Iomega Drive for about $140. That boils down to less than 30 cents per Gigabyte.

Once you’ve got those and a regular backup schedule – at least once a week! – you have some basic protection in place. Next time, I’ll take a look at online backup solutions. But believe me, you want at least one local, bootable backup.

Previous Posts:

• It’s 3 AM – do you know where your data is?

Tags: OSX, backup

That, in my book, is a bad idea. I’ll invoke the mythical UI Grandma (“Would your Grandma know what this means?”). It’s simply confusing. People have just gotten used to the idea that the blue button is a news feed. Throwing a choice of format at them is a bad idea. (Especially since it doesn’t really matter. RSS 0.92, RSS 2.0, Atom 0.3 – nobody on the user side cares as long as they get their feed. If this is what brought you here, just pick one of the feeds at random.)

Off to bugreporter we go, Radar #5570712

Boy, I’ve spent a lot of time there since Leopard came out.

But what’s the verdict in terms of actually using it?

It’s a rather mixed one. On the UI front, I’m rather disappointed. There’s a lot of rather ugly stuff going on. (Well, ugly for Apple. Still beats any other OS out there.) The only explanation I have is that they’re trying to go for the “upscale executive” look. Which is a bummer – OS X used to be fun when you look at it, now it looks more like work. Still, it does look fairly good. And I’m sure there’ll be plenty of 3rd party apps that fix any shortcomings in no time.

Speed-wise, Leopard, and especially Spotlight, took a great leap forward. Spotlight was way faster than under Tiger even while it was still indexing.

In terms of functionality, it’s an interesting beast. No single feature is overwhelming enough that you’d say “I MUST HAVE THIS”. But there are tons of neat little usability enhancements that will make your life much easier. Be it the new text-highlighting search in Safari, the ability to restore previous Safari sessions, Data Detectors in Mail – all very good additions. Oh yes, WebClips too. It’s a really nice feature, but not the one thing to sell them all. So, overall, it’s worth the money, but not too exciting. And no, not even Time Machine is that exciting. Then again, it’s backup software – it’s a marvel that it’s interesting at all.

But the real enhancements in Leopard are on the Developer end of things. The new frameworks (Core Animation, FSEvents, etc..) are mind-blowing. The tools have gotten even better. Instruments (ex-XRay) is the best profiling tool I’ve ever seen. In fact, the name “profiling tool” doesn’t even do it justice. Objective C 2.0 is a solid step forward. The scripting support for both Python and Ruby is fantastic.

In short, if you’re a developer, you must have Leopard. If you’re an end-user, it might not be the worst choice to wait a week or two before you install it. It will be inevitable, though – I’m fairly certain all the developers out there would love to go Leopard-only with new software.

That’s scary, considering that most of my life resides on some form of hard drive by now. So I decided to take a step back and look at what I want from my data.

1. I don’t want to physically lose it. That means it needs to be stored in at least two different locations. (Two different online services is fine – my blog lives at dreamhost and on Amazon’s S3) Currently, I’m using SuperDuper to back up to an external hard drive, but it is getting a bit unwieldy as more and more of my data starts to live online.

2. I don’t want to be at the mercy of a single proprietary format. I don’t want to be tied into any given application – I like testing out new things, and I’m not fond of what’s commonly called “The Microsoft Tax” – the need to constantly update your office applications. At the very least, any software I use should have the ability to bulk export into an open format.

3. I need a local backup of online data – I never want to be cut off just because I forgot to update my credit card information, or some provider goes out of business.

4. Data needs to be revisioned. I’m not only interested in the current state – if I accidentally modified something a few weeks ago, I want to be able to undo that change. Maybe Leopard and TimeMachine will solve this problem – we’ll see.

5. Some of my data I want synchronized across several locations. An online calendar is no good if it’s only online, and an offline one doesn’t help me if I’m not near my laptop.

6. Some data needs a bidirectional synchronization with other peoples data – calendar information most prominent among them. I do want to share my calendar with my wife, and she needs to be able to modify it independently of mine. For that problem, there’s BusySync – but there’s more data that wants to be shared.

In the coming weeks, I’ll be taking a look at a lot of my different data sources, re-evaluating them in accordance with those guidelines. As I find solutions, I will be sharing them in this space. In the meantime, I’d love to hear from you what your data is doing when you don’t look.

Ah well. The sad part here is that I spent a lot of time working around another Leopard issue that I’d heard about – no more input managers in Leopard. And, since it seemed interesting at the time, I was trying to figure out how to best get my little application to still work with iCal. That was – apart from the learning experience – a waste of time.

And, as a learning experience, what, exactly, did I learn?

• I need to be a member of ADC, Apple’s developer program. If I had been, I would’ve seen seed editions of Leopard and would have known about this beforehand.

• Corollary to the above – I need to charge for at least some of my software. ADC doesn’t pay itself. I’m sorry about that, because I liked the donation ware approach. At the same time, it’s depressing when you get a hit from e.g. lifehacker, 2000 people download your app, and not a single one of them donates. (At this point, a huge “Thank you!” to all of you who did donate!)

  There’ll still be the odd small app that’s just not worth charging for, but do expect me to branch out into bigger projects.

• Don’t waste time on things that seem “interesting”. I could’ve concentrated on a lot of other things instead of trying to work around a future problem that I couldn’t even verify first hand.

And what will happen to iCalFix? Well, I’ve got a current version that supports Growl. I’ll take out unfinished features, polish the rough edges, and release it on Friday. From there on out, it depends on you guys: How many of you plan on staying pre-Leopard? Are there features I could add that would make you want to pay? (And if you’ve donated already, that’d be an automatic free license to 1.0, should there be a 1.0)

As for other projects, I’m exploring. Acorn seems to be nicely shaping up as a platform for all sorts of interesting graphics ideas, so I might spend some time there. I’ve got a couple of other ideas floating around, so I’ll explore those too – and if there’s something you desperately need for your Mac, you can always let me know…

The first thing I tried out was simply running a filter kernel. Strangely enough, Apple’s documentation in that area is rather lacking – the assumption is always that you want to build the whole filter infrastructure with your kernel. Which I don’t – I want to get things going as soon as possible.

Well, after a bit of fiddling, here’s the necessary Python code:

kernelCode = """
kernel vec4 test(sampler src)
{
    return sample(src,samplerCoord(src))*0.5;
}
"""

def main(image):

    kernels = CIKernel.kernelsWithString_( kernelCode )
    kernel = kernels.objectAtIndex_(0)
    sampler = CISampler.samplerWithImage_( image )
    filter = CIFilter.alloc().init()

    return filter.apply_arguments_options_(kernel,[sampler,None],None)  

As an additional caveat, if you use Acorn versions older than 1.0.3, you need to attach a crop filter so the resulting image has extents. (Gus has added a fix for that in later releases)

extent = image.extent()

image = filter.apply_arguments_options_(kernel,[sampler,None],None)

# This part can be skipped with the latest Acorn - Gus added a fix. (Thanks, Gus!)
theCropVector = CIVector.vectorWithX_Y_Z_W_(0.0, 0.0, extent[1][0], extent[1][1])
cropFilter = CIFilter.filterWithName_("CICrop")
cropFilter.setValue_forKey_(theCropVector, "inputRectangle")
cropFilter.setValue_forKey_(image, "inputImage")
return cropFilter.valueForKey_("outputImage")

This gives you tremendous power, since it allows you to create ad-hoc filter effects – and all you have to do is change the python code on disc. Acorn will automatically use the latest version.

The only drawback here is that it’s hard to control filter parameters in a plug-in. You could write lots of U/I code – and you could do that in Python, yes – but Acorn already has a pretty fancy interface to filters. So instead, we’d like to register a core image filter via Plugins.

Can we do that? Technically, yes. PyObjC lets us subclass Objective C classes. Unfortunately, Python plugins get executed after the list of Core Image Filters is built – so our own filter will never show up in the filter menu. (The same is true for Objective-C Acorn plugins)

So – at least until Gus changes that – I’ll publish my filter kernels as Image Unit in Objective C. It’s a pain to create, compared to just wielding Python, though. (Technically, I could probably use the CGenericFilter in Jonathan Wights ToxicMedia framework. Haven’t gotten around to it yet. Please let me know if you try.)

I’d bug Gus about that, but I’ll wait till I can actually afford a license. I’ve soaked up quite a bit of his time already, and all on the trial version. Thank you for all the support, Gus – and I’ll be back

And because I’ve had so much fun developing filter kernels, here’s a Core Image plugin with the spotlight and reflection effect I used further up. Just drop it into ~/Library/Graphics/Image Units, and you have those two filters in the ‘Stylize’ category.

Update: Full source for Acorn Python plugin

Just recently, I stumbled upon an iCal replacement called Even-t. It’s way more powerful than iCal and offers a host of features. It is a tad bit pricey, but if you have extremely busy calendars, I’d definitely give it a look.

My wife and I are both rather busy, and unfortunately in different jobs. That means we spend a lot of time syncing up our calendars, making sure we know each others appointments, etc. And still – things slip through the cracks. We both keep our own iCal calendar, but there’s no easy way to share them.

That is, until BusySync entered the scene. This little app auto-discovers other iCal calendars around you and allows you to easily share them – both ways. I can add to my wife’s calendar, she can add to mine, and we can see each others schedule. I’ve been testing it for a couple of days now, and it’s been simply awesome. I couldn’t imagine living without it any more.

So go head over to the BusySync download page and get your trial copy. They’re in beta till the end of the month, and then will start selling for $49.95. Yes, that’s not cheap – but it is worth every cent.