Hacker News Reader

Comment by Traster | Hacker News Reader

The thing that seems super strange to me is that they don't seem to have warned people they would do this, the first comment is

> This isn't very helpful, Amanda! I know that the wiki wasn't optimal, but there were many wiki pages developers like me had bookmarked for years beacuse they contained comprehensive and easy information, which is now missing. Why not just keep the wiki read-only online? Just to retain the old pages? I'm pretty lost right now without some of these articles and I don't understand why the only option you had was to completely disable it. Please think about opening it up again just for read. I don't care about the maintenance mode, but the wiki was an important learning point, which is now gone.

If you don't want to support the wiki that's fine, you don't owe anyone hosting, but if you're going to dump it, atleast give someone the opportunity to scrape the site and host it themselves.

View on HN

Lesson I've long learned: never bookmark anything. Bookmarks are to temporarily remember a URL, not to archive content. If the content is important or meaningful to you, save the page.

Yes this. Grab the 'singlefile' add-on[0] and train yourself to hit that instead of bookmarking. You'll be much happier if you ever need the information again!

[0] https://addons.mozilla.org/en-US/firefox/addon/single-file/

I'll bookmark that page, thank you!

I see what you did there. :-D

Naturally, you mean you'll SAVE that page, right? LOL.

Ah, right, of course, so that I can later access that page locally and install the extension whenever I need it. Then with that extension installed, I could even save that extension page!

Firefox: reader view then bookmark saves the content.

im surprised at myself for shilling a service i dont even use and dont have any stake in, but https://pinboard.in/ lets you bookmark things and then also archive those things so they dont go away

That site could do the same thing as the wiki.

You can download a complete archive of all the bookmarks and pages: https://alexwlchan.net/2017/07/backing-up-pinboard-archives/

I tried this in the past, but could not even get it to compile. Can you share your working setup, or have you just found this?

Not impossible, but it's a big part of how the site makes money.

Absolutely. When it comes to sites with mostly static content I have a habit of archiving them with HTTRACK, though I wish there was a more modern solution with better support for active content.

Too bad modern single-page apps make it very hard to just scrape the site.

For some inexplicable reason some pages add script tags with javascript. So what happens is that you archive a page. The page is gone and when you visit it you notice all the broken links pointing to javascript files that are gone as well.

Archiving even just a single webpage is non-trivial. You're dependent on the internet archive if you want a reliable backup.

Such a solution presents a bit of a challenge though, given that you'd be 1) broadcasting a security issue, and 2) possibly compounding it by presenting your audience with some really disagreeable news.

I can at least see why they'd hesitate to leave things up, depending on the anticipated risk and likelihood of addressing it in reasonable time.

Edit: Downvote if it makes you feel better, but this is really how groups execute on problems like this without taking time away from other important projects. "Security issue? Extensive fixes needed? Take it down!"

"We're going to shut down the wiki starting xx/xx."

Don't need to explain, don't need to make a fuss - just announce and move on.

Doesn't that also just open up a lot more community feedback/pushback? I'm thinking somebody saw it coming in any case and made the call.

Hackernews is shutting down in 45 seconds. Please save the pages of any bookmarks now

About as much as we have right now. The difference is people can at least make backups of the content they need.

Can’t you just lock the system down and isolate it enough so the security vulnerability is a non-issue? Certainly there’s an ops solution to things like this.

wget or other site rippers can just make static content out of it. You can write a script to put a notice/header at the top and host it on nginx .. or an S3 bucket.

I do not think your comment deserved to be down voted. That may have very well been their line of thought.

But I also think loktarogar provided a better solution.

And that is what a discussion forum is for.

They could provide a the public or a trusted third-party with a database dump, or put it on a separate hosting service. With a bit of effort either course would relieve them of security concerns

Reply on news.ycombinator.com