Sharing static HTML of the whole output

[IanHopkinson]

I'd like to be able to share a URL for my summary. At the moment I screen shot then crop in PowerPoint, then email...

[zarino]

On the live data, or on the data as it was at the instant you "shared" the summary?

[IanHopkinson]

On the data at the instant I shared the summary. The images/media have hyperlinks which I'd like my sharee to have access to.

[frabcus]

So this would make a kind of static copy of the HTML, and let you share that?

[IanHopkinson]

Yes, zarino and I discussed this a bit. Making the summary "live" is not really desirable.

[frabcus]

Here's a prototype of one page frozen that I made yesterday for fun. I made it by the browser saving a frozen version of its HTML (i.e. $('html').html())

https://premium.scraperwiki.com/bl6ssyc/psrihabwlxvmrgm/http/x.html

Does that seem fit to purpose? Note that it doesn't have any of the post-facto interactive javascript (hover on the charts) because of how it was generated. There might be a way round that, but there isn't a cheap one :(

My idea would be to have a "Share" button that opens that up (possibly with a header to help you share it on Twitter, embed it or wthaever).

[frabcus]

Note for me: scraperwiki.readSettings() being null is the test for whether in frozen version or not.

[IanHopkinson]

That's progress! I note that my use case typically involves taking one of those panels to embed in a blog.

[frabcus]

So single panel is more important than whole page?

And actually, having a PNG file probably better than HTML? (To avoid any issues of HTTP vs. HTTPS)

[zarino]

If the HTML fragments you're currently saving aren't interactive (because they don't contain javascript) then saving to a image is indeed simpler and more compatible.

Are you going to just call the Google Image Charts API and generate a similar chart? Or are you going to try and convert the DOM into an image (using something like this)?

[frabcus]

@zarino, why I've not seen that HTML-renderer-in-Javascript-canvas, that's craziness! Alas, if you paste the URL https://premium.scraperwiki.com/bl6ssyc/psrihabwlxvmrgm/http/x.html into http://html2canvas.hertzen.com/screenshots.html you get a bit of a mess. Mainly in the SVG!

To be fair, you get some beauty too!

[frabcus]

Someone (@pwaller or @paulfurley) suggested using https://github.com/alphagov/screenshot-as-a-service from GDS. If it is super well made, it does feel like a service ScraperWiki would have found very useful in the past. (Our attempts on Classic often broke, as the underlying QT screenshot thing was flakey, and it needed a fake X window)

[frabcus]

BTW, @IanHopkinson those URLs on cobalt also work without https, e.g. http://premium.scraperwiki.com/bl6ssyc/psrihabwlxvmrgm/http/x.html - does that help with embedding?

[IanHopkinson]

@frabcus the embedding problem is that the Wordpress blog is on HTTP.

I think it was me who suggested screenshot-as-a-service ;-) @pwaller and @paulfurley will have a good opinion as to whether it is worth using.