The app uses OpenCV to do some processing on live output from my video camera. Here are some screengrabs from the prototype so far:

I look younger without a nose.

The only problem I’ve had is this: it wouldn’t remember the window size and position of the desktop app on my Ubuntu machine. I installed CrashPlan as root as suggested by the installer (so that it can backup anything on the system). This put CrashPlan in /usr/local/crashplan, owned by root. Turns out that CrashPlan desktop wants to save window geometry in /usr/local/crashplan/conf. But it can’t because I run the desktop app as me, not as root.

So, the fix would be for CrashPlan to store user settings in the user’s home directory (for example ~/.crashplan/ui.properties). But the workaround until that happens is to manually create a properties file in the /usr/local/crashplan/conf folder that CrashPlan desktop has access to. Open a terminal and type:

cd /usr/local/crashplan/conf
USERNAME=`whoami`
sudo touch ui_$USERNAME.properties
sudo chown $USERNAME ui_$USERNAME.properties

That creates an empty properties file in the crashplan directory with your username that you have access to. You should be able to just copy that into a terminal. Double check the directory where CrashPlan is installed and note the backticks. My username is john so it creates a file named ui_john.properties. The next time you close CrashPlan, it will be able to write to that file so it will remember its window position and size.

Our goal is to establish long-term relationships with developers,
and we encourage you to contact us before terminating your agreement
to see if we can improve your Amazon Appstore Developer Program
experience.

If you would like to terminate your agreement, please send an email
with your developer name, contact information, and a list of your
applications to apps-notices@amazon.com with a CC via email to
contracts-legal@amazon.com. We will notify you of receipt of your
email and respond within 48 hours with confirmation of your request
to terminate your agreement.

So I did that. Felt a lot like filling out one of those stupid mail-in rebate forms. Two days later I received a notice that my agreement was terminated and that they’d “appreciate feedback on why you are terminating your agreement.” The email was signed: Winkie C., Amazon Appstore Account Team.

It’s Javascript. It understands a handful of verbs and nouns (and one magic word) but it isn’t a real Z-machine interpreter by a long stretch. Coding it reminded me of creating my first text adventure game in BASIC on an Apple II.

About 1/3rd of the Javascript is used to animate the grue. I wrote it so that I could instantiate multiple grues on the page but ended up with just one because I couldn’t be bothered to solve the problem of grues overlapping. I drew the grue—quite proud of that, actually. Turned out nicely. I was inspired by “G” from The Geek ABCs (headinjurytheater.com)—a very lovely grue.

There are plenty of bug tracking systems out there. Conventional wisdom says it’s better to use one of those than build one. I’ve tried a lot of them. But they’re all too complex, too cumbersome, too expensive, or too ugly. (And, yes, aesthetics count.)

So I created Caravela.

Caravela is a simple and powerful bug tracking system designed from the ground up to be fast and to stay out of your way.

For example, Caravela doesn’t require you to register. Caravela creates an account for you the instant you land on the home page. You can create a new bug tracker for your project in exactly one click. No joke. It is fully configured and ready to be used in one click. And like magic, when you close your browser and return, it remembers you.

Creating bugs is just as easy. There is exactly one required field: the bug description. There’s not even a field for bug title or summary. Caravela automatically generates a title for you based on the description.

There are exactly four statuses and four priorities. And you can change them from the listing on the search results page rather than having to click into each bug.

Everything that can be updated can be done inline without a page refresh. Comments, status changes, administering users, attachment uploads, you name it. Want to update the status? One click, done. Upload an attachment? Click to open the file browser, click to select the file, done. Add a tag? Type it, press enter, done.

But all of that ease of use doesn’t mean Caravela isn’t powerful. You can have public or private bug trackers (as many as you like); multiple team members with different roles; comments, attachments, and tags; full-text search with sorting and filtering controls; tag subscriptions; email notifications; and there’s a lot more still on the way.

If you think it’s something you could use, I’d love it if you gave Caravela a try. And if you like it, I hope you’ll tell your friends (and me!).

That totally made my day.

I usually don’t hear from people except when there’s a problem. Bug reports are probably the most common way for developers to interact with users. That’s natural and I don’t fault anybody for it. But it can wear you down, even if it’s only one complaint for every hundred downloads. But that one thank you… magic.

I think writing software—or just about any act of creation, really—is accompanied by the hope that others will find that what you’ve created is useful/beautiful/good. That’s true for commercial software, too, but I think it’s especially true for free and open source software.

So. Do you have some favorite app that you use regularly? That makes your life better or your work easier? That makes you happy? Go to the developer’s web site. Send them a thank you note. It’ll make their day.

So I’m evaluating Bibble 5 Pro (version 5.1f). I had to process over 1,000 images this weekend and LightZone was killing me. I really like LightZone. But, damn… it is slow. It does have batch processing capabilities but they’re not particularly robust (there’s no way to apply some adjustments but not others, for example).

Bibble 5 Pro

Did I mention that Bibble 5 is fast? Being able to flip through a group of images as fast as I can click is awesome. Plus it has a keyboard shortcut for just about everything. Awesome. Plus, the output queues. Awesome. And it has selective editing with layers built-in. Awesome.

On top of all of that, Bibble 5 has built-in asset management (DAM) that supports multiple catalogs, searching, keywords, the works.

I’ve been trying to find a decent DAM application for Linux for a long time. I used F-Spot for a while but stopped using it once I found out it was modifying the EXIF on my original files. They’ve fixed that recently but… once bitten. Shotwell looks very promising so far. It is a basic DAM app. Not as featureful as F-Spot but also not as bloated. It feels much lighter and more responsive.

F-Spot and Shotwell and Lightroom and Bibble all use a database to make searching and browsing your catalog speedy. That means you have to import your photos. What that does is load all of the metadata about each photo and stores that in the database. Then you can add things like keywords that are also saved in the database. The database is important. I’m convinced that this is the way to go for the best combination of features and speed for any DAM app. Apps like gThumb don’t have a database and don’t require you to import your photos. But gThumb is also noticeably slower when searching or doing batch tagging operations.

The database concept is fine with one condition: I must be able to get my data out of the database. I’m creating a photo archive that I intend to use for decades. My archive already spans 14 years and over 40,000 photos. Companies die. Formats change. So, the long-term archiving strategy must be one of upgrading your data to the next format. That includes both hardware and data.

Which is all to say that open formats and the ability to export your data are important. Keywording all of my photos doesn’t do me any good if a few years from now the company goes bust and there is no way to get my data out of their database. What am I supposed to do then? Keep that antique computer running in the closet just so I can look at photos from the turn of the century? Or what if a better app comes along and I just want to switch?

Bibble and Lightroom both support an export format called XMP. It’s basically a little XML file that contains all of a photo’s metadata. XMP isn’t exactly open (Adobe owns it). But it’s open enough. It’s XML and XML files are just text files. You can double click one and open it and read it and it’s pretty easy to write computer programs to parse and work with the data. So one way to export your data is to just export the XMP files.

I created a catalog in Bibble for that project I mentioned earlier. And I poked around in the database folder it created. And I was excited to find out that Bibble uses an open format for its database: SQLite! SQLite is a public domain database engine. That happens to be the same database engine that Lightroom and F-Spot and Shotwell all use, too.

That opens up a lot of doors as far as interoperability and portability goes. For example, it would be possible, using standard non-proprietary tools and technology, to export all of your keyword data from the Bibble 5 database and import it directly into Shotwell’s database. Or create an application that synchronizes all of your keywords between Bibble and F-Spot.

You can browse the Bibble databases by using a SQLite client. On Ubuntu Linux you can install a command line client with: sudo apt-get install sqlite3.

Then you can go into your Bibble catalog folder and type sqlite3 base to open the main database. The meta file contains the metadata database. So then you can do queries like, say, getting all of the keywords for a specific photo:

select mb.value from Master m join Version v on v.masterID=m.id join Container c on c.master=m.id join meta.Browse mb on mb.childID=c.childID join ContainerInfo ci on ci.id=c.id join meta.MetaData mm on mm.id=mb.id where m.id=2732 and mm.value=’Keywords’;

Awesome! That plus XMP export basically negate any fear I have of lock-in and makes it much easier for me to just go with Bibble for asset management.

Except…

If you have a modern camera, you probably shoot a video every once in a while. It annoys me a little that none of these apps handles video. I don’t expect to be able to edit them. But it would be nice if videos could be imported into the asset database and tagged and searched and then I could click and watch them. Shotwell might add video support this year. Bibble and Lightroom probably never will. (I know—they’re photo apps.) I think I’ll either have to ignore my videos for now and wait for a better solution or use a separate system for cataloging my video.

So, that’s where I’m at. Bibble allowed me to cull and process those 1,000 photos with ease. It did crash a couple of times but I never lost any work. Overall, it’s a very good product… way ahead of anything else available for Linux right now in this category. I’m going to experiment with building some catalogs and see how I like the asset management compared to doing the same thing in Shotwell. Video… I’ve still got to figure that out. Share if you’ve got a good system.

I’d love to hear what you think and if you have any solutions to some of the issues I’m having or if you use Bibble’s catalog features.

stream by aspheric.lens, on Flickr

The word “lifestreaming” feels overly broad and I do a mental eye roll every time I say it, but I’ll use it for now for the sake of expediency.

So I’ve been using Flickr and Delicious forever and Twitter for a while and Google Reader and all of these services are great and let me share things in different ways. But what’s always bothered me is that these Shared Things were not only siloed off from each other, they were on sites I didn’t control. Yes, they all make it very easy to republish your stuff. I was doing that for a while with sidebar widgets for each of those services (I still have them on my Asides page, actually). But it still all felt very unconnected.

What I really wanted was to hook into the sharing functionality on those sites so that when I pushed the Share button in Google Reader, say, it would also send the item to my personal blog. That would allow me to aggregate my photos and bookmarks and tweets in one place and keep a permanent record of all of those things that is in my control.

I looked into several services (thinking I might let someone do the aggregating for me and then download the result), standalone software and plugins for WordPress. One of the better and easy to use solutions in the WordPress plugin category is Lifestream by David Cramer. It does almost everything I need except that I didn’t like the way the resulting lifestream ended up being something separate from the actual blog. I wanted my blog itself to become the lifestream.

In the standalone software category, PubWich looks very promising. But, again, I wanted something with seamless integration with WordPress.

So I ended up writing a plugin over the weekend. What it does:

• Periodically scans and downloads feeds from sources you define
• Makes a new blog post for every new item in each feed
• Blog posts can be created as new WordPress 3.0 Post Types or posted to specific categories
• Defines filters so that you can hook into each individual feed as it is being imported to tweak the content
• Saves a history of every feed

Then based on that, I created a WordPress theme that styles posts depending on the source. That’s how I display the post icons and the “View original” link with the domain of the source website, for example. Because every item becomes a blog post they inherit everything WordPress provides: they go into the feed, comments, permalinks, etc. That’s cool! (In a very nerdy way.)

So far, after one day of use, I’m very happy with it. I intend to release the plugin after I do some more work on it. Let me know if you’re interested in helping me test it.

These are the icons I created for kidsweatherreport.com. I don’t know why I didn’t think of this sooner, but it just struck me today to package them up and distribute the source files under a Creative Commons license (cc-by). The license allows you to do anything you like with the images including modifying them and using them in commercial works provided you attribute me when you do so (see included license for details).

The source file is an SVG image. It’s a vector image so it’s best to edit it using a vector illustration application (like Inkscape or Illustrator) but it can be opened in Photoshop and the like. I created it using Inkscape. (Inkscape is awesome.)

Download FD Weather Icons