DISQUS

Glad you like it.

There is a private API method call used that may or may not be supported in < 10.4 - look in WebView_Private.h - I think it might have been added to WebKit to support transparent backgrounds for Dashboard Widgets. I can't think of anything else that might limit CHyperAlert to 10.4 though.

Maybe a job for a Sandvox plugin?

Yeah Python (and by extension PyObjC) is great for things like this. But I had the code anyway and just needed to polish it up a tad and whip the demo up.

Well feel free to create your own multiple array object in Objective-C then just provide a serialiser/deserialiser methods for them. Not sure if I made my code extensible in this way - but should be easy to add in.

I think they deserve credit for shipping it, making it look very pretty and keeping the disk image below 10MB. I wonder where the line is between their GUI code and their back-end. It seems to use CoreData for storage (chat logs perhaps) which implies that the Cocoa (Dumbarton) side is more than just a mere GUI.

Which is fine. But not necessarily the optimum behavior. It is better in my opinion if a bug report is sent immediately. (A signal handler can catch the crash immediately, iirc the guys behind MacSQL have released source code showing how to do this).

I like QuickSilver too - but I really dislike the silly filename with greek characters in it.

Well the point is who the heck knows what pre-release software can do? It could accidentally delete your home directory when all it meant to do was delete an old cache file, etc. Heck even released software can have catastrophic bugs (remember the itunes installer that coul wipe out an entire volume?)

But it stands to reason there will be more bugs with beta software because it hasn't gone through as many test cycles as release software. (the beta test _is_ a test cycle). Honest developers (like Sandvox) are up front with that and make sure the users are aware that the software is incomplete.

As to users not knowing what beta means - well that basically means that developers should avoid public betas because the meaning of 'beta software' has been diluted by these long term betas (such as Google news). What are you suggesting? Developers should continue the confusion by releasing bug free public betas?

The whole Sandvox/SCR thing seems like a non-issue to me. A few vocal users have taken a mistake on the developer's part (i.e. not informing the user they were installing SCR) during the public beta of a product and a have magnified the problem out of all proportions. The software was beta, the correct response would have been to inform the developers of the bug and see if/how they plan to address the problem. Sure, if the developers showed no signs of addressing the problem then knock yourself out, blog about it.

Yep. I made the classic mistake of shipping the Debug build not the Release build. The zip file now contains the release build binary.

Yup - completely understand. I'll be upgrading to a different blogging engine soon and will make sure it supports easy copy & paste.

Hey Peter,

I only noticed the third side effect because Service Scrubber listed one of the apps I develop multiple times.

I think the worst side effect might be the need to reapply the preferences every time an application is installed. Perhaps you might want to cache the user's settings and just have a re-apply button? That way instead of a using having to scan the list to find what applications/services he wants to disable he could just click one button.

Also - I think it would be wiser to backup the settings to a separate file, that way the user could restore the settings without ServiceScrubber just by overwriting the original Info.plist with the backup (you could also use the unix find command to find all modified applications) if ever ServiceScrubber stopped working...

Personally I'd prefer a Haxie to solve this. It seems more elegant than modifying whole bunches of applications. While Haxies can reduce system stability I think most users know they what they are getting into. You can also license APE for free if you release your software as freeware.

Well actually the best solution would be for Apple to fix the solution themselves, without Service Scrubber (or Blacktree's prefpane that Mark Grimes mentions) the services menu is practically unusable. Thanks for releasing such a great piece of software to help deal with the situation.

Thanks for the comment. I removed the smart quotes (looks like ExpressionEngine likes to transform your quotes for you).

I'll try out the modification to the sed statement and if it works for me too I'll update the code. Thanks!

Hmmm. Not getting the build problem you're seeing. Sorry! I'll be updating the stable branch soon. Maybe the problem will magically go away ;-) Here's hoping.

Hey Ted,

I broke the palette. I found the breakage yesterday and its fixed in the trunk. I am updating the stable branch now.

Let me know how you get on with my code.

Phil:

See my posting here: /index.php/weblog/mod_spotlight_apple/

I've also worked on a hack for safari that allowed you to search spotlight natively in safari using a URL scheme of "x-spotlight".

Working on it. See updated post.

Daniel:

True. I probably understated the difficulty of it a bit. I already have a code to do a lot of what it would do (displaying thumbnails from spotlight) so it might be a bit easier for me.

But not solving the problem yourself because Apple already has a widget that they're sharing between there apps in the hopes they'll share it with 3rd parties isn't going to help much. I'd love Apple to prove me wrong and provide a hook for it - but hell they don't even provide APIs to create all the different buttons they have in the iApps.

I never considered that this would be a "global disk browser" (heck we have the Open dialog for that - and it has spotlight built-in). I believe something that works almost identically to Apple's browser but works with every application's media (instead of just the iApps) is easily accomplished.

For extra points you could use iTunes/iPhotos XML files - at least until Apple provides (if they ever do) spotlight metadata for this information.

And I dont believe every developer should write their own. Something like this should definitely be OpenSource and shared between developers.

Well in a perfect world the iApp media would have spotlight metadata attached to show what folder it had been organised into, what film the photo was imported with and so on. AFAIK that isn't present.

I think limiting a media browser to media that has been organised only by Apple's iApps is self limiting. I think it would be best to ignore the Apple Media browser (lets face it, the chances of Apple providing an API is pretty low) and create your own (feel free to release it as an Open Source component).

But the trouble with Apple's browser is that it _only_ works with media in iPhoto, iTunes, etc. If someone came along and wrote a better iPhoto or iTunes (it could happen!). Then Apple's browser would not pick media stored by the new applications.

The real problem is that Apple's browser only browses its own (semi) private data

Writing the browser itself is relatively trivial. Spotlight is a really simple API and providing a UI to it via Cocoa bindings wouldn't take too long at all - even with thumbnailing and other media previewing.

Python does have regular expressions. They're not built into the language, but are in the standard Python library.

Secondly - only the first version of my code is looping multiple times. The second version with the generator expressions (and don't forget you need Python 2.4) is really only looping once. For example:

<pre>theWords = (theLine.rstrip('\r\n') for theLine in theFile.readlines())
### Remove single letters...
theWords = (theWord for theWord in theWords if len(theWord) > 1)
### Make the words upper case...
theWords = (theWord.upper() for theWord in theWord)
print theWords</pre>

No loop is occurring until the print statement is reached. The (xxx for xxx) lines are just creating iterator objects which are like stages of a loop.

I take that back - your perl is SUPER fast:

<pre>real 0m0.877s
user 0m0.435s
sys 0m0.057s</pre>

Congrats.

Well - I guess my timing was a little off...

I just ran it again:

<pre>real 0m8.366s
user 0m7.315s
sys 0m0.211s</pre>

That's quite a difference! (This is on the same Powerbook - with the only difference major difference I can think of is that File Vault is turned off now - not that I see how file vault can make much of a difference)

I just changed all the list comprehensions to generator expressions and tested it:

<pre>real 0m5.095s
user 0m2.683s
sys 0m0.177s</pre>

With output piped to /dev/null it is a touch faster

<pre>real 0m3.057s
user 0m2.584s
sys 0m0.114s

That's about the same ballpark as your perl script.

I was never claiming my code was super efficient or I wasn't looping too many times... ;-)</pre>

Hey Tom,

It has been tested on one other (non iSight) webcam. But that driver was firewire and didn't need any 3rd party drivers. So YMMV.

You can use subversion to checkout the source code (" co http://http://toxic-public.googlecode.com///tox...) or you can wait until I upload an tarball archive (hopefully tonight).

Thanks Mike.

I'm hoping there's a Mac developer group here in the upstate - but it looks like it might just be me and the guy who makes x-plane...

Hope to see you at WWDC next year though!