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。

Specifying Spring '83

80 点作者 nexthash将近 3 年前

10 条评论

rgovostes将近 3 年前
The idea is appealing but I feel that there is a conflict between<p>&gt; I want this to work, furthermore, whether those people are sharing a random thought every day, a blog post every week, or an art project every two years.<p>&gt; More importantly, every board holds its place, regardless of when it was last updated.<p>I would not like to stare at the same board for two years between updates, so probably I would end up manually re-organizing my client according to the feeds that get updated, or just unsubscribe from the ones where the author seems to have dropped off the planet.<p>Newspaper classifieds and comic book ads are different each issue and therefore there&#x27;s some pleasure in scanning through them. Today&#x27;s algorithmic feeds on social networks may optimize too much for engagement at the expense of quality. (Twitter started putting complete strangers&#x27; tweets at the top of my timeline, on topics like &quot;Marvel&quot; that I have no interest in.) But the solution to this is not to avoid algorithmic curation completely.
评论 #32249489 未加载
doctor_eval将近 3 年前
I love the ideas here, but...<p>* if posts are limited to 2217 bytes with no external references then that excludes the richness of images, something I really enjoy about the modern web. Am I misunderstanding something?<p>* I think there&#x27;s a pile of room for the client to display things differently from that imagined here. ultimately we would end up with clients that implement their own &quot;algorithm&quot;, but that&#x27;s a good thing because we&#x27;re in charge of it.<p>* I always get irrationally stressed about the inefficiency of pinging a server every hour for a resource that only gets modified once a year, one of the use cases noted in the article. That&#x27;s a lot of wasted 304 not-modified&#x27;s.
lmm将近 3 年前
Full CSS and zero javascript is the opposite of what I want.<p>I love the idea of just seeing everyone&#x27;s current status, but realistically I want to be able to read people&#x27;s history as well, otherwise the FOMO will have me checking this thing manually every minute (maybe an effective growth hack, but sounds like it&#x27;d end up even worse than twitter if it worked at all).<p>And realistically an RSS reader with a &quot;tile view&quot; achieves everything that this does.
turnsout将近 3 年前
This is super fascinating, as it combines several trends (&quot;brutalist&quot; web design or Web 1.0 nostalgia, federation&#x2F;decentralization, ephemeral Snapchat-like &quot;Stories&quot; with no history) but feels authentic and compelling.<p>As ever, the issue will be achieving the critical mass needed for the network effect to kick in. But the system is simple enough to have a chance at succeeding (see Gall&#x27;s Law)
Cyberdog将近 3 年前
I don&#x27;t understand the cryptographic aspect of it. Unless I&#x27;m misreading things, which may be the case, it will take several minutes of basically cryptocurrency-style &quot;mining&quot; to find a key which matches a message on current relatively fast consumer hardware before a &quot;board&quot; can be posted. If the goal is validating a message hasn&#x27;t been tampered with, is that serving some goal that something like sha512 or even crc32 wouldn&#x27;t match? (It also means that it would basically be impossible to create a Spring &#x27;83 message using hardware actually contemporary to spring &#x27;83.)<p>Also, I&#x27;m not sure how important having the month and year in the identifier is to the whole thing, but using only two digits for the year has obvious ambiguities.<p>That the client must &quot;situate each board inside its own Shadow DOM&quot; seems to imply that all clients must effectively be web browsers too. A far cry from the original RFC 865 which could basically be done with netcat.<p>All in all this seems like a complex way to solve a seemingly easy problem.
评论 #32246471 未加载
smm11将近 3 年前
We could have gone so, so many different directions.<p>Still can, I&#x27;m thinking.
pc2g4d将近 3 年前
Top-level cryptographic identifiers that create an identity separate and distinct from a network location are basically de rigueur for network protocols these days. Almost as if it&#x27;s a missing layer between IP and TCP? Interesting to see the trend continue in this proposal.<p>Without an easy way to pay for content online, this will be subject to the same distorting, ad-infiltrating forces as the web in general. Charging even tiny fees for content online resolves a host of problems... while of course creating others.
dash2将近 3 年前
Something this has in common with similar efforts: the web page is treated as a document - here it&#x27;s a set of documents, i.e. &quot;boards&quot; - and not an application. As a result, commerce is excluded. That&#x27;s fine if it&#x27;s what you want, but I think it limits the scale and appeal to nostalgia. I deprecate lots of things about the modern web, but I like being able to buy things and perform transactions online.
syntheweave将近 3 年前
I mused on a very similar concept a few years back, without writing more than a brief treatment. I complicated it a bit for myself by also considering the idea of a periodic delay line loop that has some of the properties of a reblog - a way to circulate ideas repeatedly and add those of others to your own, making the page fresh even when unattended.<p>But I think the basic idea is fine.
sogen将近 3 年前
Sounds like Fraidycat