Without Flash the web is more readable?
The web page above is essentially made of text and a few images. And the ads on the right too, are text based: these communicate their message using words, not animations. Still, the right part is entirely made of little Flash pieces.
We all know that Flash has been used for providing web pages with videos, interactive presentations, creating globally animated web presences. But what I didn’t realize until I started bookmarking pages “fragments” with our Licorize browser plugin is that Flash is widely used for writing text. Text that once clicked has some side-effect (that is why Flash is used), but still text.
The problem is that in writing text on pages through small Flash embeds, we have contents inside web pages that is not interpretable by anything else consuming the page, apart from a being capable of vision and semantics i.e. some living beings. But applications can’t do anything with that.
What I’m trying to say is that more and more we use the web to “read” the web, extracting contents from several sources. The words aggregation, curation are recurring frequently because of this. We even created a browser game where the web itself is used as a creativity source.
We can read the web through the web because… web contents are mostly accessible – this also partly explains the popularity of jQuery, a tool which greatly eases content navigation.
Having the Flash islands progressively replaced by readable HTML contents makes the web more accessible to applications. I know that by tagging the Flash embeds one could document their contents – but that is not what happens.
Making the web more accessible to applications can be described as the web is becoming more semantical – giving a clear, modest reading to the semantic web mouthful expression.
Technical part: how we try to understand whether a “piece” of the web is a video?
Here we try to explain some technical points that evidence the usefulness of the move to explicit tags instead of embeds by showing how complex it currently is to capture non HTML web parts. This is part of a recent set of updates to Licorize bookmarking due to @robicch and @pupunzi.
By “video” we mean is bookmarking any kind of “dynamic” content – could be a Flash game.
When we have textual Flash on the page, it confuses our plugin, as we assumed in construction that Flash is not used for that (not that it actually hinders the plugin functionality in any serious way).
“Mono video” pages
Pages whose main aim is to present a single video are somewhat easier to identify, and our latest plugin version does so on YouTube, Vimeo and many others. This also because there is some convergence towards the usage of tag “standards” set by Facebook (Graph API), so jQuery code selectors can work with expressions like
data.video = $(“meta[property=’og:video’]”).attr(“content”);
Other sites use older methods like “video_src” but these can be captured too.
Notice that the URL can be edited – say with your favorite embedding features. This is the large majority of cases.
P.S. In the same way we recognize “maps” web applications as Licorize location types.
Non-standard video embeds
Some pages still embed videos without declaring in tags the video content. In these cases we use a “DOM hunter” function, which is a heuristical mess – will be available (free: MIT license) on a coming update on Pupunzi’s jQuery components web site.
You have to consider that Flash sometimes even “kills” the DOM on click events – these are managed entirely it its small universe – making it harder to analyze behavior: another reason for preferring not to have such embeds on sites.