It’s time to summarize what we have done for the Open Source community. Scribd is pretty open company, we release a lot of code into the public after a time (sometimes it is short, sometimes it is not). Here I want to mention all the code we have opensourced. Please take into account that time is moving on, so we are publishing more and more code. I will update this post periodically, so stay tuned. Follow me on Twitter to get instant updates.
You are developing a large Web application. Controllers are full of complex data retrieving logic, views contain tons of blocks, partials, loops. One day you will receive an email with user complaints about some of your pages slowness. There are many profiling tools, some of them are easy (ruby-prof), others are large and complex (newrelic), but regardless of this it’s really hard to find the particular place where you have a real bottleneck. So we created really simple, but über-useful tool for ruby code profiling.
There was a lot of articles about HTML5 last days, so when Alexey Kovyrin asked me to help him with his new blog design I saw no other choice but using HTML5. There are a lot of new features added since HTML4, and some of them could be used today, like new elements <header>, <footer>, <nav>, <article>, <section>, etc. I think this is a nice addition to the HTML, because these elements add more sense to an unstructured markup. There is a buzzword “semantic” exists to describe this, but I don’t like buzzwords, so I would call it “sense”. So what features we could get from HTML5, that are supported by all modern browsers?
Sometimes we need to write small console scripts which do some background processing on our models. In Scribd we are using Loops to run these tasks (check Alexey Kovyrin’s introduction post). One of our scripts supposed to generate some sort of html reports, and we need to generate links to our site there.
