4 ms·

This is the site that got me to move away from tables. My how time flies! Remember the days when teeny tiny microfonts were in style? Also Photoshop orbs... Eve

by eksith 13y ago

This is the site that got me to move away from tables. My how time flies! Remember the days when teeny tiny microfonts were in style? Also Photoshop orbs... Everywhere, there were bloody orbs; from gaming forums to "tutorial sites" (that linked to other sites). Part of me feels nostalgic, but another feels it's good to grow out of all that. In many ways, we were entering a Goldilocks zone for a semantic and accessible web, but I fear we're starting to slip back a bit into a client-side dynamic web phase that isn't accessible, not semantic and overall app focused. Which is fine if you want the web to be a series of apps, but there's a heap of content in there too. And it would be a shame if it only became accessible via apps. "Webkit-only designs will be discarded with prejudice... Graceful degradation is acceptable, and in fact highly encouraged." Can't tell you how happy I am to see that they haven't lost any of their sensibilities.

modarts 13y ago
> but I fear we're starting to slip back a bit into a client-side dynamic web phase that isn't accessible, not semantic and overall app focused I agree with the overall sentiment that content should be accessible; but one thing you aren't taking into account with the above statement are the channels by which we access the content. Today REST APIs are that channel, and serve as the backends for the dynamic heavy client apps, as well as a medium for other services/applications/users to interact with and consume that content. I'd argue that it allows for an even more semantic manifestation of data and content, since we aren't limited to defining semantics with just a series of tags wrapping content; served to clients in large homogenous blobs of HTML in readonly fashion. Instead we expand our conceptual space to include HTTP verbs, headers, and URLs to provide an interactive interface to data that far eclipses what the "Semantic Web" movement was ever able to deliver.
- eksith 13y ago
  Fair enough. But then a REST API isn't a panacea since browsers(or content parsing apps) need go through a new iteration from rendering content "as-styled" to "as-meant" for a semantic web. We don't have that yet and it doesn't look like vendors are going down that path any time soon. Then there's the question of text-only browsers like Lynx. A semantic web as you describe isn't the current web at all, but a new completely new layer on top of it.
  - dredmorbius 13y ago
    Readability, Instapaper, and Pocket all do a pretty good job of "as meant" for sites. The do this through a server-based solution, with some reliance on standards, but also, I'll suspect, a large number of site-specific parsing and formatting rules.
- gambler 13y ago
  even more semantic manifestation expand our conceptual space What exactly does all of that mean? REST is just a subprotocol of HTTP. It does add some guarantees about server behavior, but it does not make chunks of JSON any more semantic than if they were served over FTP. The simplest example of why HTML is semantic (and the underlying benefits) is the very textarea I'm writing this in. I can resize it to accommodate larger messages I'm typing. It does spell-checking. The functionality was developed once and now benefits thousands of people visiting thousands of websites. Moreover, I can rely on it and control it. All of this thanks to semantic difference between textarea and some random div. This wouldn't be possible with some random JSON format or custom JavaScript control.
dredmorbius 13y ago
I wholeheartedly agree with you that CSS / Ajax design are jumping the shark. I've found myself for the past couple of years increasingly fighting websites and their design in the interests of simply being able to read content. I've flagged numerous submissions to HN on that basis alone. And joined the Contrast Rebellion (http://contrastrebellion.com/ http://contrastrebellion.com/). After using element editors to interactively correct issues, userContent.css to override settings (especially font face, and size), and other hacks, I stumbled across Readability. First the bookmarklets and browser plug-ins that simply reformatted pages to be more readable, then finally finding that the full-blown service, including saved reading lists and tags, was worthwhile to me. Which addresses another failing of browsers: the insanely poor UI/UX of bookmarks management. It's not just the sharing aspect, but being able to enter, arrange, tag, annotate, deduplicate, etc., bookmarks. I've been seeing a place increasingly for two modes of browser: an app-mode browser for highly interactive sites, and a content browser, whose aim is to provide access to, and management of, informational content. One mode I've found Readability hugely useful for is in bookmarking, categorizing, prioritizing, and reading a large volume of literature related to research I'm doing. And presenting that information in a way that's tolerable to read. The only downside so far: a per-user limit of 500 tags. I'd also love to have something resembling Readability (preferably itself) for PDFs, which are absolutely awful to read online, and impossible on mobile (yes, I've got Adobe's PDF reader for Android, it's freaking useless). ePub and readers such as Calibre and Moon+Reader are another approach to online documents I very much appreciate. Some associated G+ posts / rants: Readability: two days in -- VERY nice https://plus.google.com/104092656004159577193/posts/PhQR421etdT https://plus.google.com/104092656004159577193/posts/PhQR421e... Dear Internet Desiners: We need to talk https://plus.google.com/104092656004159577193/posts/NFEFMZdkd5q https://plus.google.com/104092656004159577193/posts/NFEFMZdk... Annoying Web annoyancs are annoying: CSS Interstitials https://plus.google.com/104092656004159577193/posts/j357ACUsejY https://plus.google.com/104092656004159577193/posts/j357ACUs...
- fractallyte 13y ago
  I checked out contrastrebellion... and was amused to find HN highlighted as one of the examples of bad design in the 'No Contrast' section ;-)

by eksith 13y ago

modarts 13y ago

> but I fear we're starting to slip back a bit into a client-side dynamic web phase that isn't accessible, not semantic and overall app focused I agree with the overall sentiment that content should be accessible; but one thing you aren't taking into account with the above statement are the channels by which we access the content. Today REST APIs are that channel, and serve as the backends for the dynamic heavy client apps, as well as a medium for other services/applications/users to interact with and consume that content. I'd argue that it allows for an even more semantic manifestation of data and content, since we aren't limited to defining semantics with just a series of tags wrapping content; served to clients in large homogenous blobs of HTML in readonly fashion. Instead we expand our conceptual space to include HTTP verbs, headers, and URLs to provide an interactive interface to data that far eclipses what the "Semantic Web" movement was ever able to deliver.

eksith 13y ago
Fair enough. But then a REST API isn't a panacea since browsers(or content parsing apps) need go through a new iteration from rendering content "as-styled" to "as-meant" for a semantic web. We don't have that yet and it doesn't look like vendors are going down that path any time soon. Then there's the question of text-only browsers like Lynx. A semantic web as you describe isn't the current web at all, but a new completely new layer on top of it.
- dredmorbius 13y ago
  Readability, Instapaper, and Pocket all do a pretty good job of "as meant" for sites. The do this through a server-based solution, with some reliance on standards, but also, I'll suspect, a large number of site-specific parsing and formatting rules.
gambler 13y ago
even more semantic manifestation expand our conceptual space What exactly does all of that mean? REST is just a subprotocol of HTTP. It does add some guarantees about server behavior, but it does not make chunks of JSON any more semantic than if they were served over FTP. The simplest example of why HTML is semantic (and the underlying benefits) is the very textarea I'm writing this in. I can resize it to accommodate larger messages I'm typing. It does spell-checking. The functionality was developed once and now benefits thousands of people visiting thousands of websites. Moreover, I can rely on it and control it. All of this thanks to semantic difference between textarea and some random div. This wouldn't be possible with some random JSON format or custom JavaScript control.

dredmorbius 13y ago

I wholeheartedly agree with you that CSS / Ajax design are jumping the shark. I've found myself for the past couple of years increasingly fighting websites and their design in the interests of simply being able to read content. I've flagged numerous submissions to HN on that basis alone. And joined the Contrast Rebellion (http://contrastrebellion.com/ http://contrastrebellion.com/). After using element editors to interactively correct issues, userContent.css to override settings (especially font face, and size), and other hacks, I stumbled across Readability. First the bookmarklets and browser plug-ins that simply reformatted pages to be more readable, then finally finding that the full-blown service, including saved reading lists and tags, was worthwhile to me. Which addresses another failing of browsers: the insanely poor UI/UX of bookmarks management. It's not just the sharing aspect, but being able to enter, arrange, tag, annotate, deduplicate, etc., bookmarks. I've been seeing a place increasingly for two modes of browser: an app-mode browser for highly interactive sites, and a content browser, whose aim is to provide access to, and management of, informational content. One mode I've found Readability hugely useful for is in bookmarking, categorizing, prioritizing, and reading a large volume of literature related to research I'm doing. And presenting that information in a way that's tolerable to read. The only downside so far: a per-user limit of 500 tags. I'd also love to have something resembling Readability (preferably itself) for PDFs, which are absolutely awful to read online, and impossible on mobile (yes, I've got Adobe's PDF reader for Android, it's freaking useless). ePub and readers such as Calibre and Moon+Reader are another approach to online documents I very much appreciate. Some associated G+ posts / rants: Readability: two days in -- VERY nice https://plus.google.com/104092656004159577193/posts/PhQR421etdT https://plus.google.com/104092656004159577193/posts/PhQR421e... Dear Internet Desiners: We need to talk https://plus.google.com/104092656004159577193/posts/NFEFMZdkd5q https://plus.google.com/104092656004159577193/posts/NFEFMZdk... Annoying Web annoyancs are annoying: CSS Interstitials https://plus.google.com/104092656004159577193/posts/j357ACUsejY https://plus.google.com/104092656004159577193/posts/j357ACUs...

fractallyte 13y ago
I checked out contrastrebellion... and was amused to find HN highlighted as one of the examples of bad design in the 'No Contrast' section ;-)