I did not intend to read this from beginning to the end. Talk about a sticky article.
Consider quoting the text directly though. Making images of text is probably the most inaccessible thing imaginable. Should take less than a hundred lines of CSS to create a believable HN container in a div.
> Making images of text is probably the most inaccessible thing imaginable.
Agreed. I love the concept of this site, but it would load significantly faster and provide better user experience if it utilized text instead. Loading this website took me 9.41 seconds.
Images don't scale, and people using screenreaders would be out of luck, for starters.
If you click on the images, it'll open to the actual comment in text form.
While load times for images are worse, this format provides some handy advantages, namely the ability to look through that thread and see more context, read the related article, etc.
No advantages that a css/html rendered version of the same thing doesn't have, with the loss of accessibility that you'd get using the alternative approach.
This is just "the easy way out". Justifying it otherwise is wrong, in my opinion...
This would be an easy thing to do and everyday occurrence if we had proper, first-class-citizen transclusion [1] mechanisms as part of the Web standards and their implementations.
That's fair, and true from an engineering perspective. I imagine for YC it comes down to, as with nearly every other company, a prioritization problem.
With the exception of companies like FB with millions of users, the priority of creating a css/html rendered version of text to prevent people from having to click through to copy is quite low.
I suppose I'm defending the image as being the more useful of the two "easy way outs" that are likely to be implemented. :)
Sharing the image retains attribution and source. Sharing the text alone opens channels of abuse and lack of context... I like this option more, personally.
On top of that, someone could probably script this process to extract the text from a link or create this page from a series of links automatically. Well, at least the quotes and links if they want to hand-write the intro or ending. The result would help both for people with weak/slow connections and any impairments that tools assist with.
Not that I'm for screen shot, I imagine they wrote a quick script that went through a list of links to the comment and produced the screenshot (which also resized the height of the window to fit the full comment).
The script probably utilized an integration with a screen capture library. I'd imagine someone can write a script to parse html and dig out the comment, but the screen capture script was probably faster to write.
A slower method would be to copy and paste the text by hand...
Gee, if only Hacker News had an API to extract the text and associated metadata from a comment, so that it could be presented in a textual way rather than as an image…
Consider quoting the text directly though. Making images of text is probably the most inaccessible thing imaginable. Should take less than a hundred lines of CSS to create a believable HN container in a div.