odyniec.net

For some time now, I’ve been using Twitter on a daily basis. I’m far from being a hardcore twitterer (or whatever the term is), but I do spend a couple minutes every day reading new stuff that appears on my feed.

When I browse tweets on my desktop machine, I usually don’t use any apps — I just visit the Twitter website and don’t need anything more fancy than that. However, the list of people that I follow has grown over time, and I recently started missing one useful feature that the Twitter website used to have — easy access to lists. I have created a few private lists and put most of the followed people on them, and being able to quickly open a list to read just the tweets of people on that list was really convenient.

Back in the old days (that is, like a year ago), Twitter used to display lists in the sidebar, next to the timeline. But, then they were removed, and instead buried several clicks away in the interface. Probably because the vast majority of Twitter audience simply wasn’t using lists, Twitter had no interest in making that feature easily accessible. (Some people were speculating that lists will be completely removed at some point — but I hope that won’t happen.)

Luckily, user scripts have been invented to deal with situations like this. I found one existing script that kind of did what I wanted, but since it only supported public lists (mine are private), it didn’t meet my needs — so I ended up quickly putting together my own version. Welcome back, sidebar lists:

If you’d like to use the script, grab it from Userscripts.org or from GitHub. Tested to work in Firefox and Chrome.

Yesterday I had another crazy idea that ended up implemented as a Greasemonkey script. I was browsing some source code in Firefox, and I thought that the browser is missing one feature that I commonly use in an IDE, and that is the ability to open the same source file in two views displayed side-by-side. This is useful when you need to look at two parts of the code that are related, but are many lines away — like, a function definition and the place where the function is called.

For those of you thinking “meh, couldn’t you just open a second window?” — sure, and I sometimes do that, but it’s not very convenient, as it requires me to do some dragging/resizing to get the desired layout. I wanted an easier solution.

I implemented this in a very simple manner, by replacing the current page with a new document containing two iframes, and loading the current URL in both of them. To make it more useful, I also added the ability to resize the iframes by dragging the splitter bar between them. Here’s a screenshot of the script being used to view a source file on GitHub:

The script is available on Userscripts.org.

Here’s a slightly late announcement — a few days back, I added a new project to the projects page. It’s called MonkeyConfig and it’s a configuration dialog builder for Greasemonkey scripts.

As with most of my projects, it was born out of necessity. I wanted to add a user-friendly configuration screen to another script of mine, View as Source. I knew about the de-facto standard configuration library GM_config, but I had trouble getting it to do what I wanted — I couldn’t find an easy way to create a form field combined of two drop-down lists, and the generated configuration screen looked ugly for whatever reason (like some styles were missing). I finally gave up on it and decided to develop my own solution.

The just-released version is still pretty immature, but it’s already functional enough to be useful, especially for simple scripts with just a few configuration parameters. Here’s an example of a configuration dialog generated by MonkeyConfig for the View as Source script:

If you’re a Greasemonkey script author and would like to give it a try, you’ll find some basic usage instructions on the project page and in the documentation (it’s pretty minimal at this moment, I’ll be expanding it).

The script is also on Userscripts.org and GitHub.

Today I released the Greasemonkey script that I wrote while being sick on the snowboarding trip last week. It’s called “View as Source”, and it allows you to open any link in the page using the view-source URI scheme.

I developed it to resolve an annoying problem that I had every now and then with source code files linked on the Web. Although pretty much every source code file is plain text, and Firefox could display it as such, for certain MIME types it insists on opening the file in an external editor (or saving it). You know, like this:

Even if I just wanted to take a quick peek at a source file, and despite having a perfectly capable text file viewer in front of me, I had to launch another application.

I experimented a bit and found out that the view-source URI scheme can be an effective workaround for this. It’s utilized by Firefoxes “View → Page Source” option to produce a syntax-highlighted view of the HTML code of the currently open page. However, if used with a non-HTML file, it displays its contents as plain text — just what we need.

So, what the script does is it prepends the "view-source:" scheme prefix to links right before they are activated. It only does this on demand, as you certainly wouldn’t want to do this to every link on every page that you visit. You enable it with a keyboard shortcut (Shift + Alt + U), and then the script waits five seconds for you to click a link — anything you click within that time period is opened with view-source.

If you want to try it, the script is available at Userscripts.org. As usual, I’m looking forward to your feedback.

In my last post, I got somewhat excited about the idea of making a Greasemonkey script that would allow you to adjust the brightness of images on websites. One afternoon a few days ago, having nothing better to do, I wrote the script — and it sort of works. More precisely, it works only on some websites.

The script, like the imgFx plugin, makes use of the canvas HTML element. Among other things, canvas lets us retrieve the image’s pixel data as RGB values. To make an image brighter, we need to read all the pixels, increase their RGB values, and use the new pixel data to redraw the image. Sounds rather simple, but there is one fundamental problem. Due to security restrictions, Firefox will let us read the pixel data only if the image originates at the exact same domain as the website itself. Even if it’s just the sub-domain that is different, it won’t be allowed. So, if you have a website at somedomain.com, but your images come from images.somedomain.com, the script won’t work with it.

Anyway, in case you’d like to try out the script, here it is: brightenup.user.js

I’ve checked the most popular image hosting websites, and it seems only ImageShack meets the requirement of images originating at the exact same domain as the page on which they appear. Thus, the script is set to execute on *.imageshack.us. Here’s a sample image that you can test it on (no goatse, and safe for work, I promise):

http://img479.imageshack.us/my.php?image=letescaupeje0.jpg

Move the mouse cursor over the image, and two buttons will pop up in the top left corner. Press “+” and “-” to make the image more or less bright.

Update (Nov 8):
I’ve uploaded the script to Userscripts.org for the world to see — there it is.