TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Ask HN: Did APIs Kill the Semantic Web Dream?

4 点作者 casper345超过 6 年前

1 comment

PaulHoule超过 6 年前
Yes and no.<p>If the &quot;semantic web&quot; is dead it is that people are sick of updating standards. For instance, we know now that we should have SHACL before we had OWL (forgot the old &quot;garbage in garbage out&quot; thing) but it is too late. Probably the only semantic web &quot;standard&quot; that realizes that RDF can do everything JSON can do is the XMP metadata standard that Adobe has pushed. (Somehow the librarians who made Dublin Core didn&#x27;t seem to think it mattered what order you listed the authors of a book in -- but the one guy Adobe had thinking about it did.)<p>One strange thing that turned up with the semweb for instance is that maybe 80% of the time people use an ordered list in JSON they don&#x27;t really care about the order -- representing facts with set semantics actually works most of the time but unfortunately when you do 80% of the work to get to an MVP it is still not viable.<p>People at this point might be vaugely aware that SPARQL would be a lot more useful if they added &quot;object-relational&quot; features similar to N1QL but the standards fatigue is too set in. (Eg. you should be able to write queries that know about ordered lists)<p>The Object Management Group has long been following a path which is parallel to the semantic web. For instance, if you have machine readable API schemas (eg. represent the semantics of the API) you could have API clients that &quot;code themselves&quot;. Thus there really is no contradiction between APIs and the &quot;semantic web&quot;.