I really like the way how Google Chrome deals with OpenSearch. When you browse and OpenSearch-aware site, Google Chrome fetches the OSDD file and make it immediately available as a shortcut in the address bar. For example, if you open a couple of pages in this site then you type simone you should see the following behavior.

Google Chrome suggests an available OpenSearch provider

If you hit tab, the autocomplete list disappears and you enter the “OpenSearch mode”. You can now enter the search query and confirm to view the search results.

Google Chrome integrates OpenSearch provider in the address bar

Creating an OpenSearch OSDD is as simple as writing less than 10 lines of XML code. Here’s mine. For more information visit the OpenSearch website.

PS. If you are a Ruby developer, here’s my Builder template. Replace TITLE, DESCRIPTION and PATH_TO_SEARCH with your own variables.

Related posts

1. How to run Google Chrome OS on Mac OS X with VMWare Fusion
2. Google Chrome OS User Agent
3. Google Site performance tool says Google is not serving Gzipped resources

Redmine 0.9.0 offers tons of improvements but one among all forced my decision to upgrade without waiting for a final release: Rails 2.3.5. Redmine 0.8.x is built on Rails 2.1.0 and there seems to be some kind of weird Rails-Redmine-Passenger-Ruby EE memory leaks.

With the new release, Redmine has changed the most part of its URLs. The development team basically migrated all routes to a more RESTful-oriented architecture. They did a wonderful job, unfortunately they probably chose the worst way to ensure backward-compatibility: duplication.

All the old routes are still available along with the new ones and this can cause several duplicate content issues. To patch the problem, I configured some Apache redirection rules in order to permanently redirect the old URLs to the new ones.

The gist doesn’t pretend to solve all issues. It’s just a quick and (quite) dirty fix.

From Redmine side, I would have setup a Rails Metal to redirect all the old URLs to the new ones. This is not as efficient as relying on the web server, but this is compatible with any user environment.

Feel free to fork, clone and update the gist. After all, this is a Git repository!

Related posts

1. Configuring a Git repository with Redmine
2. Apache .htaccess query string redirects
3. Generic Apache configuration snippets

Likewise, people tends to pay more attention when an expiration is expressed in words. For this reason, the next RoboDomain EAP will include a small improvement on how dates are formatted. When applicable, RoboDomain will use a relative time — in 2 hours, in 3 days, in 1 month — instead of the raw timestamp value.

RoboDomain with Relative date

Under the hood, we used the awesome timeago jQuery plugin with a couple of customizations.

This feature, along with the usability improvements and many other changes, will be available with the next EAP release.

Related posts

1. RoboDomain usability improvements
2. Welcome RoboDomain

In these two weeks we concentrated most of our effort to improve the application usability. In our mind, RoboDomain is a super powerful and helpful toy, but this is mainly because we already have a global overview of the project.

Looking at our server logs, we noticed some of our existing users are not taking advantage of RoboDomain features mainly because they don’t fully understand their capabilities. For this reason, in these weeks we worked hard to add some friendly messages all over the application.

First, we fixed the welcome page that was definitely the most unfriendly landing page I have ever seen in my life. We added a brief introduction to RoboDomain and some quick-links for getting started with your brand new account.

We also wanted to remember our users they are using an Early Access Preview of the product and some features might be incomplete or not available. Of course, EAP stage is not a synonymous for bugs or low-quality.

User feedback are really important for us and we decided to make the process of submitting new ideas even more easy. From any page, you can now access RoboDomain Uservoice forum clicking on the feedback tab on the left side of the page and immediately post a new idea or submit a bug report.

These changes are scheduled for the next EAP which should be available in a couple of days.

Related posts

1. Welcome RoboDomain
2. Understanding RoboDomain “Transactions” feature
3. Relative time… when the time does matter

Transactions represent the history of your domain including purchases, upgrades, hosting transfers and any other operation on your domain. One of the main RoboDomain goals is to collect and organize all the information about a domain in a single place to let you easily search and query them.

A transaction is composed by 4 different resources, each own with its own properties.

1. The operation type. The first domain is likely to be a purchase operation, followed by recurring renewal operations
2. The hosting associated to the operation (expect this feature to change in the upcoming release)
3. The order
4. The payment associated to the order

Let’s see a real example.

You own example.com domain and you just added the record into your awesome RoboDomain account. It should looks like this.

In the tabbed navigation menu there’s a link called Transaction. Because haven’t recorded any transaction yet, the page should look like the following one.

Let’s start recording the domain purchase. Click on the “Record new transaction” button and the transaction wizard will guide you through the application process.

Select the registration type, the date you started the operation and the date the operation finished, if different. Then, fill in the details about the hosting provider.

The wizard is composed by 4 steps. You can quit at any step and come back later to fill the missing information. As soon as you finish the first step, a new transaction page will be available.

The design is quite raw at the moment since we’re still in the process of improving the wizard workflow, but it provides a clean overview of all your transaction details.

Transactions are displayed in reverse chronological order in the domain transactions section.

The latest domain transactions are also displayed in the domain page. An overview of the latest transactions for all domains is displayed in your account dashboard.

Transactions are useful for the domains you actually own, so were’s working on a way to disable this feature when not requires.

Also, there are tons of improvements already scheduled in our TODO list to reduce the amount of time required to fill the transaction details and provide some additional capabilities. Just to mention a few: notifications, cross-domain references and integration with public registry data.

As usual, don’t forget to tell us what do you think about this feature. You can submit bugs or suggest new features using the RoboDomain uservoice forum.

Related posts

1. Welcome RoboDomain
2. RoboDomain usability improvements
3. Understanding Ruby and Rails: Benchmarking your Ruby scripts

I logged the user agent browsing my website with a Google Chrome developer preview. And here’s the Google Chrome Browser user agent on my Mac OS X.

As you can see, the browser token is the same because both Google Chrome OS and my Mac are running the same browser version. Google Chrome OS identifies itself with the CrOS i686 9.10.0 token, as you can see in the first user agent.

I’m wondering how long it takes before analytics software will start tracking Google OS users.

Related posts

1. Safari 4 User Agent
2. How to run Google Chrome OS on Mac OS X with VMWare Fusion
3. Search SimoneCarletti.com from Google Chrome address bar

The answer is yes. Google Chrome OS actually works inside a virtual machine.

If you are running VMWare or Virtual box, setting up a Google Chrome OS instance is really straightforward. Thanks to the guys at gdgt, you can download a Google Chrome OS image ready to be used with VMWare or VirtualBox. Also, Paul Miller from Engadget published an awesome screencast with a step-by-step tutorial to run Chrome OS as a virtual machine.

Here’s a screenshot of my Google Chrome OS virtual machine.

Google Chrome OS on Mac OS X using VMWare Fusion

Related posts

1. Google Chrome OS User Agent
2. VMware Fusion “The network bridge device on /dev/vmnet0 is not running” error
3. Search SimoneCarletti.com from Google Chrome address bar

If you are interested in you don’t have too much time to take a decision. It’s a time-limited offer and it expires tomorrow.

I have been using JIRA and Confluence in a couple of project. As an open source project, the Zend Framework benefits of a free license of the full suite. It’s a shame this offer doesn’t include Crucible, I’m really looking forward to try it.

Altought I’m a really happy Redmine user, I don’t rule out the possibility of purchasing a license. It could be useful, soon or later…

No related posts.

A quick online search revealed me multiple solutions but the most of them were targeting VMware 1.0. The error above in WMvare 2.0 or greater can be easily fixed typing the following command in the Terminal (thanks Aaron).

After a few seconds your virtual machine should automatically reconnect throught your macosx network bridge.

Related posts

1. How to run Google Chrome OS on Mac OS X with VMWare Fusion
2. Google Chrome OS User Agent

December 30th, 2008. This post was published long time ago on the Italian version of my blog then moved here after the English blog has been opened. Please note that the following post can contain outdated information and (probably) multiple typos.

Firefox 2.0 is out… uhm? what are you saying? what a great news?… oh yeah, but this isn’t the main topic of this post.

As someone of you probably remember, on June I made a few tests with Firefox 2.0 Bon Echo 3 focused on how the new Firefox handles feeds. Firefox 2 gives you full control over Web feeds, showing you a preview and letting you choose how you want to subscribe.

At the end of the tests I wrote an article showing how is possible to add a new content handler, aka web aggregator, in Firefox 2.0… it was digged too! (Note. The article is no longer available).
Since Firefox 2.0 was not yet released at the time the article was written, I didn’t know that there was an easier way to add an content handler without tweaking by hand Firefox registry.

Using registerContentHandler() javascript function is possible to add with just one click a new feed handler… how? Let me show you.

Adding a new feed handler to Firefox 2.0 with registerContentHandler()

registerContentHandler() is a JavaScript function that allows you to register a new content handler in Firefox 2.0.
It accepts 3 required params:

mimeType

the content mime type.
For a feed must be application/vnd.mozilla.maybe.feed.

uri

the content handler URI.
%s can be used as a placeholder for the resource URI.

title

the content handler Title

Once again, let’s use Rmail as example.
Rmail add-new-subscription URI is (always the same) http://www.r-mail.org/bm.aspx?rss=%s while content handler title is Rmail.
Thus, the final JavaScript should looks like the following.

You can put/call this JavaScript function where everywhere in a HTML page, for example as a link (as Netvibes did)…

Example 1.

Look & feel

Click here and add Rmail

Code

… you can put it into a form button…

Example 2.

Look & feel

Code

or include it into a JavaScript function and call it anywhere in the page.

Example 3.

Code

More power to registerContentHandler()

This tutorial might be closed here, but the real reason why I wrote this post is to show how is possible to add more power to registerContentHandler() with a few JavaScript lines.

The idea comes from this post where Randy shows a button to add Rmail as a new content handler.
In a comment pat asks which is the best technique between Randy’s magic button (see example 2 and example 3) and my tutorial (Note. The tutorial is no longer available).

This is Randy’s answer.

Pat,
Simone’s technique is actually superior as it also includes the email address. I’m working on making that possible too!
Randy

Don’t worry Randy, I will give you an help! :D

The only difference between my code and Randy’s code is that I used the following Rmail subscription string (please note the mailto param)

instead of

Thus, what we should do is find a way to allow user to type its own email address and append it to the content handler URI.

My first idea was to create a new input text field. It’s better than using other techniques such as JavaScript input boxes since the field can be easily filled with a server side programming language in a more complex site, can be validated and you can re-use in thousand ways.
While I was writing the code, I decided to do a little more. Why not creating a script that allows developers to add custom arguments to a content handler URI instead of a single one? Rmail just needs a mailto param but other readers may need mail and user…

This is what I wrote!

The following script allows you to register a new content handler in Firefox 2.0 just calling a simple function.
In addition, you can create as many form field as you wish calling them name="args". For each field, the script will collect its value and if the field isn’t empty, will appends the value to the content handler URI.

Let’s see an example. Again, I will use Rmail and I will show you how it’s easy to append your personal email to the URI.

Example 4.

First you must include the script.
Don’t forget to customize chUri and chTitle variables, as written at the top of the javaScript file.

Then add an input text field for each argument you wish to append to the URI. You can also use hidden inputs to pass hidden values. Don’t forget to call them args or the script will not work.

Now call the subWith() function. You can use a link, a button or a common JavaScript event.

Enjoy the confirmation dialog! :D

Related posts

1. FeedBurner Stats 2 Feed