Primordial Ooze

Over my vacation to the Dominican Republic, I had some time to do some programming. You might be wondering why in the world I would be programming on my vacation. I have a few answers.

First, I like programming in serene environments. The sound of the beach, the lapping waves and the wind against palm trees is very soothing. I enjoy this almost as much as sitting by the pool or beach drinking a good beer. Second, my son was sick and he basically layed in bed all day like a log. I had to watch him part of the day while the rest of the family enjoyed the sun. Third, the raw files my camera was generating were eating up like 20MB each. Many of my shots were crap and I wanted the raw files gone.

What’s a raw file? A raw file is a proprietary file format stored by most medium to high-end cameras. A raw file stores a lot more color and pixel information than a regular jpg file. First, it’s uncompressed so when you zoom into little details you won’t find jpg compression artifacts. Second, it stores more colors. Typically 12–bits of color per channel. That’s 4096 shades of gray instead of 256. This can make a big difference in certain circumstances, especially when you are changing your image brightness, contrast, or doing color correction. Without this extra color information, an innocent image brightening can wipe out a lot of detail in that very white sand.

I use Google’s Picasa software to manage my photos.  When my camera records raw, I also told it to store a 3 mega-pixel jpg file. Picasa easily imports these and ignores the raw files. I have to manually copy the raw files onto my hard drive.

Picasa is awesome software and I use it to crop photos, delete bad ones, and basically review everything. However, any time I delete a jpg-based photo in Picasa, the raw file remains.

Fortunately the raw file has the same file name as the associated jpg file. It was easy to write a program that will find all *.sr2 (my raw extension) files that have no associated *.jpg file in a given directory tree. Vioa, enter Raw Version 0. Download Raw Version 0. I do have to come up with a better name.

The program will only work under Windows XP and you’ll need to download and install the Microsoft .NET Framework Version 2.0 Redistributable Package (x86).

Now I can manage my photos in Picasa and clean up the crappy raw files consuming gigabytes of space on my hard drive.

By Nick at 11:06 PM | Permalink | TrackBacks (0)

Here is a picture I took this morning with my son in front of a sunrise. I learned what the “SL” flash seems to do. This was quite an exciting discovery for me. It exposes the background for a bit and then issues the flash (or vice versa?) so a foreground subject doesn’t show up as a silhouette against a bright background.

UPDATE: I wrote this while still on vacation, but the Internet connection was pretty bad and the post never made it. I’m back so here it is! Bandwidth restored!

By Nick at 11:00 PM | Permalink | TrackBacks (0)

Words are hard to describe the experience of watching the start of a brand-spanking new day. Next challenge, getting the kids up to see this tomorrow!

By Nick at 11:12 AM | Permalink | TrackBacks (0)

I can’t begin to describe just how often the concept behind uBrowser is. A colleague sent me the link and when I read a little bit about the project, who’s writing it, and what the technology will ultimately accomplish, my jaw dropped.

The page says, “uBrowser is a simple Web Browser that illustrates one way of embedding the Mozilla® Gecko rendering engine into a standalone application using LibXUL. In this case, the contents of the page is grabbed as it's being rendered and displayed as a texture on some geometry using OpenGL™. You are able to interact with the page (mostly) normally and visit (almost) any site that works correctly with Firefox® 1.5.”

Furthermore,  the developer works for Linden Lab, makes of the popular online virtual reality Second Life. Second Life’s claim to fame is their ability to allow users to create their own worlds. You can build towers, buildings, kiosks, clothes, cars, etc.

Now, with the inclusion of the FireFox rendindering engine, in-game players/designers can include web content in their domains. The possibilities are almost limitless.

By Nick at 7:37 AM | Permalink | TrackBacks (0)

Dan Bricklin, of VisiCalc fame, has created a new piece of software called WikiCalc. Just like the name implies, this is a web-based Wiki, but instead of editing word-processor style documents, you edit spreadsheets.

I tried it out on my web server and found the setup and use a bit clunky but this is early alpha software so that’s to be expected. It did work. Dan uses AJAX techniques so the calculations happen on the fly. I’m not using mod_perl so there is a significant delay, only a second or two, every time you enter data in a cell. Obviously there is a round trip to the server going on here. For simple data entry, I’m wondering if he can avoid this round tip and instead insert the data locally via JavaScript?

In any case, it’s a cool piece of software with a lot of promise. I’d like to see the UI cleaned up a bit as this thing matures. For instance, unlike regular wikis, the process of creating a new spreadsheet is very deliberate. Pages get “published” and not automatically created as you create WikiWords. That, or I could be missing something.

In addition, it would be great if pages could get tagged like Num Sum spreadsheets. iRows is another one. None of these solutions seems to get it completely right in my head but I still have to hand it to all of these folks. Some heavy lifting was involved in all of these implementations!

By Nick at 7:31 AM | Permalink | TrackBacks (0)

I created a new category called workflow where I’ll tag my entries having to do with my personal workflow. This includes software I use, how I use it, etc.

I re-tagged my original post on how I switched to the Google Desktop Search at home. well, now I’ve switched at work. What lured me?

• Faster searches – MSN Search would frequently hang and report that it wasn’t running when I issued a search. It would also hang every time I rebooted, forcing me to shut it down.
• Integration with all of my machines. I can now search my archive mail on my laptop, even though the 5 years of Outlook archive files resides on my desktop.
• Encryption of your indexed data. This slows down things, but I don’t think I’ll notice it.

When Vista rolls out, my use of Firefox and Google Desktop Search may be threatened. Until then…

By Nick at 7:46 AM | Permalink | TrackBacks (0)

The folks over at pbwiki have been hard at work enhancing their free wiki service, pbwiki. Their tagline is, “PBwiki makes creating a wiki as easy as making a peanut butter sandwich.”

In fact, they are currently running an offer to double the space on my own pbwiki if I advertise blog about their service.

Why don’t you give them a try? There’s no maintenance to your wiki, it’s free, it’s constantly updated and upgraded, it has a nice clean style, works in lots of languages, and has the option to upgrade to a premium account that can store a whopping 1GB of data (you start out with 10).

I did notice a few problems which I’m sure they’ll fix. Since my Wiki is new and no “recent changes were made” the rss feed link choked on this rare kind of wiki while the atom feed worked fine.

Warning:  arsort() expects parameter 1 to be array, null given in /xfs/pbwiki/user/rss2.php on line 69
Warning:  Variable passed to each() is not an array or object in /xfs/pbwiki/user/rss2.php on line 73–

Once I made a change to the wiki, the rss feed fixed itself.

In addition, their password logic has a flaw. I entered a 64–character strong password to manage the account. I usually do this from the passwords generated at Steve Gibson’s Ultra High Security Password Generator. The site accepted the password, but didn’t allow me to log in with it. There was obviously some kind of failure there.

So, I cropped off a smaller chunk of characters from the beginning, tried again, and the password reset took and I was able to log in.

By Nick at 7:23 AM | Permalink | TrackBacks (0)

… at home anyway. For some reason, I was getting increasingly unhappy with the performance of MSN Desktop Search at home. I was always torn between the merits of the Google and Microsoft offerings. The feature of MSN Desktop Search which pushed me over the edge was their neat “macro” feature. I could define little keywords and map them to arbitrary web links. Thus, a definition like the one below would substitute whatever I type in place of the $w. Consider this macro definition:

@wiki,http://www.primordia.com/cgi-bin/moin.cgi/$w

If I then typed:

wiki FrontPage

My browser would launch and I’d go here:

http://www.primordia.com/cgi-bin/moin.cgi/FrontPage

First, I find myself not using that feature like I used to. Second, the only other feature which inspired me to choose MSN Desktop search was the “richness” of the results window. Since the results window was basically Windows Explorer and Microsoft wrote a custom extension to implement such features as “Open Containing Folder” and “Thumbnail View”.

Despite the fact that the Google Desktop Search uses a browser, in my case Firefox, it manages to pull off some of these platform-dependent features you wouldn’t expect a browser to handle. The implementation is simple, I believe. The browser talks to a local web server that’s embedded into the Google Desktop Search engine. For specific “open folder” requests, the platform-specific web server goes and opens t he window. Firefox doesn’t need that capability.

I am further inspired to use the Google offering because I find myself using Google Talk and my Gmail account more and more. The ever-growing feature list of the Google sidebar are starting to reel me in.

By Nick at 1:40 PM | Permalink | Comments (1) | TrackBacks (0)

A colleague of mine sent me the link to a fantastic movable type plugin called CloudCreator. Once installed, it transformed my category list into a tag cloud, ala del.icio.us and flickr. You can see it on my sidebar. I had to tweak  the default look a bit as you’ll see below.

Using this simple default markup,

  <div class="blogs-module module">
    <h2 class="module-header">Tags</h2>
    <div class="module-content">
      <MTCatCloud all="1">
        <span style="<$MTCatCloudCSS$>"><$MTCatCloudName$></span>
      </MTCatCloud>
    </div>
  </div>

CloudCreator creates a tag cloud like so:

The problem was that the tag labels weren’t links. So, I hacked up the markup as best I knew how as follows:

  <div class="blogs-module module">
    <h2 class="module-header">Tags</h2>
    <div class="module-content">
      <MTCatCloud all="1">
        <span style="<$MTCatCloudCSS$>"><a href="http://www.primordia.com/blog/archives/<$MTCatCloudName$>/"><$MTCatCloudName$>>
      </MTCatCloud>
    </div>
  </div>

This yielded a kind of ugly-looking underlined-link scenario as follows:

Modifying the markup one last time as follows:

  <div class="blogs-module module">
    <h2 class="module-header">Tags</h2>
    <div class="module-content">
      <MTCatCloud all="1">
        <span style="<$MTCatCloudCSS$>"<a style=";text-decoration: none" href="http://www.primordia.com/blog/archives/<$MTCatCloudName$>/"><$MTCatCloudName$>>
      </MTCatCloud>
    </div>
  </div>

Yielded the final look. The default stylesheet for my site “highlights” links in a lighter color when you hover over them. In this case, my mouse was over “general”:

Also, as a consequence of adding this capability I have added a bunch more categories, er tags, to my blog:

science, illustration, photography, web2.0, and tagging.

At some point I’ll go through all of my entries and re-assign tags for a better, more tagged, collection of posts. That process will likely be very tediuous so I hope there’s a good tool out there to help me do it.

Oh, and say goodbye to this old clunker!

By Nick at 7:50 AM | Permalink | TrackBacks (0)

So many games, so little time.

These games are all brilliant! A friend of mine said the developer works out of Fresno and does all of the games for Adult Swim. He’s brilliant!

Brilliant!

The games are all flash-based, beautifuly rendered, and the spites scroll like butter. Ha, that’s an Apple Keynote reference if you missed that.

Go try out the free games and while you’re at it, buy some of his stuff over at cafepress.

By Nick at 10:40 PM | Permalink | TrackBacks (0)

This post that I saw over at boing-boing is a must-read.

The wasp slips her stinger through the roach's exoskeleton and directly into its brain… The zombie roach crawls where its master leads, which turns out to be the wasp's burrow… Now the wasp turns to the roach once more and lays an egg on its underside… The larva grows inside the roach, devouring the organs of its host, for about eight days. It is then ready to weave itself a cocoon--which it makes within the roach as well.

Whoa!

By Nick at 7:40 AM | Permalink | TrackBacks (0)

This king sized snowman was made during the winter of 1999 in Bethel, Maine. It's 113 feet tall (34.5 m), weights approximately 9 million pounds and has entire fir trees for arms!

Whoa! Chck out the details over at the Earth Science Picture of the Day website. Subscribed.

By Nick at 11:24 AM | Permalink | TrackBacks (0)