Re:My blog can now be browsed with cURL
by FroshKiller on Wednesday January 15, 2020 @06:35AM
(#22053375)
Yes, but linking to the text version instead of simply providing the text is hostile to the user. You know what resource I requested. You know where that resource is. Give me the resource.
[ Parent
| on HN ]
Re:My blog can now be browsed with cURL
by anderspitman on Wednesday January 15, 2020 @01:30PM
(#22056928)
I agree it would be ideal to return a static HTML file for direct requests. I'm working on adding that back in. But I think the /txt version of the site is still valuable. I tried to come up with a way to return it without requiring a separate URL, but short of doing magic on the server (such as detecting user agent) couldn't think of a way to do it without having nasty HTML in the result. Any suggestions?
EDIT: I updated it to HTML-redirect to /txt/feed for noscript users. Still not ideal if you're trying to find a specific post, but it includes instructions to help. I think it's a reasonable compromise.
Re:My blog can now be browsed with cURL
by FroshKiller on Thursday January 16, 2020 @08:52AM
(#22064469)
Yeah, that works. I'm glad you're willing to put in the work, but I have to tell you, if you'd just focused on delivering the content and followed the principle of progressive enhancement, you'd be delivering a better experience for probably the same amount of effort at this point.
Re:My blog can now be browsed with cURL
by anderspitman on Friday January 17, 2020 @01:09AM
(#22072321)
I think something that got lost in the conversation is that a big part of my personal site is experimenting with different ways of delivering content. My current experiment was to see if I could serve a git repo as-is without any build step, and what the trade-offs would look like. Unfortunately everyone just assumed I was ignorant or lazy.