The result is that, over the last few months I’ve gotten several days’ worth of work on Hopville done. Lots of small changes and new features have made it to the site, mostly unannounced.  While I plan to continue working on new features for the near future (because now I realize that Hopville is the site that will make me rich and famous and handsome and smart), I wanted to take this time to summarize the stuff that has already made it to the site in 2011:

• Customizable yeast attenuation. Each strain will continue to have a default attenuation value, but brewers can adjust the attenuation percentage on their recipes in order to better match their own situation or experience.
• Hopville now includes a feature allowing you to follow other brewers.  Now folks can easily keep tabs on each other’s brewing activity.
• Malt additions can now be marked as Late Boil Additions. Ingredients marked in this way will not affect the calculated gravity of the boil, which means they also won’t affect IBU calculations in formulas sensitive to boil gravity. Many brewers who use extract in their beers requested this feature – late boil additions are a great way to maximize hop utilization when brewing with extracts.
• Another common request was to allow for “each” units for miscellaneous ingredients so that, for instance, you don’t have to calculate or guess a specific weight or volume for something unit-based, like a Whirlfloc tablet.
• Brewers who measure the color of their finished beer can now store their result as the measured SRM/EBC.
• Added a page to highlight Brewing Statistics. What’s there now is a first draft – as time goes on I hope to find lots of interesting information to pull out of Hopville’s database and display here in fancy graphs and charts and things.
• Added a “share” button to easily link any recipes to a post on one of several social media sites.
• Created the official Facebook Page for Hopville.com.
• Added a new category for recipes, Extract with Specialty Grains. Formerly all Extract recipes were sorted and filtered equally, whether or not the recipe included grains. Now folks looking for one type of extract recipe or the other can find them more easily.
• Bug fixes improved sundry items: top navigation, large volume batches, recipe cloning, metrics mode, BeerXML syntax, recipe sorting, recipe “interestingness” score, direct heat mash rests, partial mash categorization, lovibond range…

Most importantly (in the grand scheme of things), now you can Support Hopville with a simple donation via PayPal. I’m hoping to create a positive feedback loop where Hopville’s fans provide significant enough financial support to keep me from getting distracted by other projects. By staying focused, the site’s improvements will come at a much faster rate, hopefully feeding back into increased financial support, meaning the site could become viable as a part-time job for me instead of the hobby site it is now.  Paying members are encouraged to participate directly in this feedback loop by voting for their favorite future features on another new part of the site, the “Future Features” page.

Current recipe count: 50,205

Current recipe count: 14,960

Thinks are calmer now as I settle into both new roles, and I continue to make progress on the site.  I’m now aiming to roll out the new stuff by the end of March. You can mark my words, or cross your fingers, or knock on wood – not really sure how to instruct you at this point.  But anyway…what is the new stuff?  I’ve boiled this release down to two major updates: better site navigation and improved Beer Calculus performance.

YOU SHOULD STOP READING NOW IF YOU’RE NOT A GEEK.

The performance improvements are what made this release a challenging one.  I rewrote all the front-end parts of Beer Calculus to make them easier to maintain and to improve the way the front-end communicates to the back-end.  Current users won’t be shocked by the changes, but their ongoing use of the site will be every-so-subtly more fun and rewarding.  Like buying a newer, noticeably-faster-at-first computer.  And on my end, the code is already cleaner and easier to manage, which means I hope to build new stuff at a faster pace than before.  I’m definitely looking forward to getting this release out the door – when a new launch (including lots of small UI improvements) is overdue by a whole month the existing site becomes quite an eyesore.

THIS IS YOUR SECOND WARNING, NON-NERD.  STOP READING!

Since this blog post announces that there is nothing new to see, I figured I’d use this opportunity to describe those behind-the-scenes changes that nobody will notice or want to hear about on launch day anyway.  Here’s what I’ve been working on:

• All Beer Calculus HTML will be rendered via HAML templates instead of ERB.  Net effect? The HTML will be more structurally and semantically sound, making the pages easier for browsers to render, easier for search engines to crawl, and easier for humans (especially me) to read.

• Beer Calculus code goes RESTful.  Net effect?  Lighter pages and simpler, more AJAX-centric page requests result in a zippier interface that’s more responsive and fun to use.  Ever notice the the scary message your browser sends you every time you try a page reload on Beer Calculus?  Yeah…that won’t happen anymore.

• JavaScript moves to the jQuery framework.  I came late to the jQuery party but there’s no stopping me now.  Net effect? Smoother and more responsive effects – web interfaces getting closer and closer to behaving as well as desktop software, without having to resort to Flash.  I still thought I’d have to resort to Flash (meaning, rewrite the calculator all over again in ActionScript) as of about six months ago, so it’s big news that JavaScript has matured to this inflection point.  Lots of cooler interface stuff to follow.

Current recipe count: 13,687

Luckily for Hopville, now that I am about to start work again, I’m learning lots of knew stuff and I’m completely energized about getting back into web development.  On a recent trip back east, I had some productive time on the plane – including my first-ever launch of website features from a chair in the sky (since my Delta flights were equipped with WiFi).  Here’s what’s new:

• Recipe pages now have dedicated sections where the brewer can add tasting notes and taste ratings.
• Brewers pitching multiple yeasts in a single batch can now list unlimited yeast strains and bacteria cultures on a recipe page.  Previously only one yeast was allowed, so brewers pitching multiple cultures had to list the supplemental strains in the recipe comments.  Now the extra cultures become a part of the main recipe.
• Measured OG and FG can be recorded on recipe pages – when present, the display of the measured values takes priority over the values estimates by Beer Calculus.
• Brewers can easily export their entire Hopville recipe catalog as BeerXML from their recipe page.  This is a nice shortcut for folks who want to backup or export their entire list of recipes at once, rather than one at a time from each recipe page.
• New and slightly more enticing homepage…though it’s still a work in progress, design-wise.  Much improved version launching in a few weeks.
• Featured recipes! Recipes are now given an “interestingness” score which allows Hopville to feature lists of the juiciest recipes on the site.  Now that there are over 10,000 recipes on Hopville, this kind of highlighting will become an important signal-vs-noise aid, making it convenient to find the best recipes quickly.

Over the next month I’ll be continuing to do some major upgrades in preparation for a larger rollout of changes I’m planning to launch on February 25th.  It’s nice to be having fun with code again…and Hopville stands to reap the benefits.

Current recipe count: 11,134

• Added BJCP style validation to Beer Calculus.  As you work on recipes in the calculator, you’ll see warnings whenever the calculations seem to put your recipe out of bounds for the chosen beer style.  These warnings can be ignored entirely, but at least the information is now available where it would be most useful.
• Added a random recipe link to the recipe navigation menu.  Self-explanatory.  And fun!  Try it!
• Moved Beer Calculus so that it lives under the hopville.com domain at beercalculus.hopville.com.  This shouldn’t have any effect on users, but it’ll help with Hopville’s “Google juice”, keep me from having to jump through hoops to maintain state between two domains, and make it more apparent that Hopville is there to support Beer Calculus.  Since the calculator preceded Hopville by so many years, lots of folks don’t even know they can save their BC recipes yet!

Current recipe count: 2026

The location of Hopville, OR according to Google Maps

So, I had to go there, and I had to turn left at Wigrich road, and I had to go that 75 ft.  How could I not?  Knowing in advance that “Hopville” would amount to two Willamette Valley crop fields divided by a road, I was pleasantly surprised to discover that one of those two fields is for hops.  Trellised fields were pretty common in the area, as I discovered with a little more exploration.  I drove around for a while looking for a sign (literally, a sign) that I was in Hopville, hoping to snap some photos to use on Hopville.com.  (Ever notice that the site has no logo?  Well, one day soon, I hope that it will, and a snapshot with the word “Hopville” might’ve helped a bit.)  But here’s the thing: I never found any sign of Hopville.  But I did find, at the dead end of Wigrich Road (a right turn on Wigrich being the dead end direction), a very good sign:

Yep, I stumpled upon the very hop yard where the estimable Oregon brewery Rogue grows and processes its own hops.

I researched this a bit online after I’d returned home.  Rogue never mentions Hopville in any of their literature, so I am not sure if anyone besides Google believes the place exists. But, I’m just sayin’, here’s a wider view of the Hopville map:

Rogue brands their hops using a town a few miles northwest of this area, a town called Independence.  Apparently it was once the “hop capital of the world”.  From Rogue’s press release:

In May of 2008, Rogue entered into a strategic alliance with heritage hop growers the Coleman family. Rogue planted 22 acres of hops and will add 20 more this fall on the former John I. Haas Alluvial Hop Farm just south of Independence. The land, at the end of Wigrich Road, is part of a historic hop farm called the Wigrich Ranch, that in the 1920s was the largest hop yard under a single trellis in the world.

Quite a coincidence considering that I was only trying to find Hopville; apparently Rogue was too.  If I didn’t have Maritime Pacific as my neighborhood brewery in Seattle, I think Rogue would have to be Hopville’s official microbrewery.  Maybe I should return during growing season so I can see hops on those trellises and see if Rogue wants to sponsor their “neighborhood website”.  Hmm….

Current Hopville recipe count: 1554. Hey, ain’t that a beer?

In other developments:

• I finally opened up the ingredients database to the masses. Previously I was the only person who could add new ingredients into the database, so folks had to drop me an email request or list their unlisted ingredients inside the recipe notes.  Now anybody can add anything right into the recipe on their own.
• To accommodate an open ingredient database, I added inline ingredient search to Beer Calculus.  Now, rather than having to read through an alphabetical select box to find the malt, hop, yeast, or miscellaneous ingredient you need, you can switch on a search box and find the selection by typing a keyword or two.  And if the ingredient you’re searching for can’t be found, it’s easy to add it.
• Previously all IBU results were calculated using an ages-old custom formula that combined parts of the “Daniels method” with the “Garetz method”.  I’ve now updated the IBU number to be customizable to the popular formulas (Tinseth, Garetz, Rager), but I also left in the old “Hopville method” so that existing recipes keep their existing IBU value.  By default, new recipes added to Hopville will calculate IBUs by averaging the three commonly used formulas, but brewers with a formula preference can customize to use that instead.
• Fixed some rounding errors introduced by the recent migration to metrics-based recipe storage, and some other bugs I introduced in the broad sweep of new changes.

Current recipe count: 675

/*
HOPVILLE COLORS
F9F9FF off white
4a443c main text
B34700 link text
FF6600 action link text
eeeeee tiger body
*/

Then you have an easy hex color reference right at the top of your main stylesheet. But “easy” becomes relative. On Hopville, I ended up describing shades of green as “medium”, “medium light”, “lightest”, and “extra-lightest”. Extra-lightest green? Which one was that again? And as the site’s color scheme has evolved, the legend has gotten out of date.  Some of the colors on the site aren’t in the legend any more, some of the colors in the legend aren’t in use anymore, and so on.
It just occurred to me that this could be automated, that I could scan all the stylesheets with a script, pull out the color attributes, and generate a color swatch page on the fly.  So when I want “that one shade of green”, I can go look it up and cut and paste the hex value, without having to dig through stylesheets or maintain a color palette legend.  Rails made it easy to generate it on the fly with an erb template (ignoring MVC for simplicity’s sake) and not even have to set up a cronjob.

Dynamic color palette browser created from stylesheets in Rails.

The code:

<%
stylesheet_dir = RAILS_ROOT + '/public/stylesheets'
files = Dir.open(stylesheet_dir)
css_files = []
colors_in_stylesheets = Hash.new {}

files.each do |file|
  if file.match(/\.css$/)
     File.open(stylesheet_dir + '/' + file).each do |line|
        if line.match(/\#([0-9a-f]{3,6})\;/i)
          unless colors_in_stylesheets[file]
            colors_in_stylesheets[file] = {}
          end
          colors_in_stylesheets[file][$1.downcase] = 1
        end
      end
  end
end

block_size = 100
%>

<style type="text/css">
div.swatch-block {
  margin:10px;
  float:left;
  width:<%= block_size %>px;
  height:<%= block_size %>px;
  text-align:center;
  border:1px dotted #ccc;
}
div.swatch-block span {
  background-color:#fff;
  color:#999;
  border:1px dotted #ccc;
  padding:3px;
}
h2 { clear:left; }
</style>

<h1>Color Swatch</h1>
<% colors_in_stylesheets.keys.sort{|a,b|
   colors_in_stylesheets[b].size <=> colors_in_stylesheets[a].size
   }.each do |stylesheet| %>
  <h2><%= stylesheet %></h2>
  <% colors_in_stylesheets[stylesheet].sort.each do |color, garbage| %>
    <div class="swatch-block" style="background-color:#<%= color %>;">
    <span><%= color %></span>
    </div>
  <% end %>
<% end %>

This works great for me since I always use hex colors, but I imagine it could be tweaked to find other color definitions as well.