Hacker News new | past | comments | ask | show | jobs | submit login
Show HN: Mummify – Preserve web content, fight link rot (mummify.it)
90 points by zek on Oct 7, 2013 | hide | past | web | favorite | 66 comments

Its worth mentioning the http://archiveteam.org/ which does a lot of important web archival work.

They use the Web ARChive format (WARC http://fileformats.archiveteam.org/wiki/WARC ) which I hope mummify and other such services will standardize on.

The approach is nice, the problem being solved is real. Hopefully paying customers will flock in.

The question: is mummify.it itself going to go under some day?

So I'll wait for an open-source analog of this service, to run on my own tiny server. (Or carve some time to write it myself, of course.)

Wget is definitely a good start. What's needed is a shiny and usable UI, ways to catalog / tag / search inside your saved data, and a painless installation process. This could use come work.

I think Evernote was also intended to solve this. Unfortunately, it is also proprietary. So, there is definitely demand for this.

How is Mummify going to get around the copyright implications of ripping off the now removed piece of content which they don't own?

Also wondering: what happens if the publisher redirects the old URL to a new place -- maybe an "update"... or maybe a useless hub page?

Weird pricing plan. At even the most expensive plan (of 15$/mo) only 50 'mummies' a month. Seems like an arbitrarily low number

There are free alternatives, like Peeep.us [1] and Archive.is.

[1]: http://www.peeep.us/

[2]: http://archive.is/

From Peeep FAQ:

How long will Peeep keep my data?

Virtually forever. Nevertheless, we retain a right to remove content which has not been accessed for a month.

Trust us to save your content, except for the stuff that is least likely to be saved anywhere else.

I don't get the number of free/paid mummifications? Seems very low, space or bandwidth are abundant.

I think, and with respect to original developer, this is more a feature then an app and probably would help if it would be developed further to target more specific problem/group.

Having said that, I wish you best.

The OP depends upon Javascript for any and all content delivery/rendition. Given the topic at hand, I find this more than a bit ironic.

I gather from the comments that this is some sort of online storage of a copy. That may serve some use cases over the short term.

If you really want to avoid loss or "link rot", maintain your own copy on your own equipment.

I've been around to observe everything from personal interest changes, death, corporate policy changes, ownership transfers, deliberate manipulation... etc. -- you get the idea -- effect the ability to pull even what were formerly considered very stable and long-standing, aka "permanent", resources.

If you want to ensure you have access, save your own copy onto hardware that you own. End of story.

50 mummies per month for the highest plan is way way way too low. It should 10,000 or even 50,000. Think about it this way: the people who are willing to pay you for this type of service are probably huge publishers who would use this service to share their URLs to their followers in Twitter, Facebook, etc.

Doesn't help them much if a huge publisher is paying them $15 a month. That type of sale would require a SLA too, they really just need a contact email address.

Oh the price should increase too. Right now OP is targeting a imaginary customer. Who would pay $15 a month just for 50 mummies?

I would be happy with 1000 per month. Typically 10 from HN everyday and 10 from elsewhere. But I would not pay $50 per month for this.

and if Mummify goes down...

Would it be a better option if the "permanents" were shared across p2p/bittorrent and every unique item had at least 10 shares distributed across the globe, maybe a max of 20. When one share host goes down, it just picks up a replacement.

Just to second what donpdonp said (https://news.ycombinator.com/item?id=6509604), I think a service like this needs to offer a standard format WARC (http://archive-access.sourceforge.net/warc/) download.

The whole point of a service like this is long-term access and that really requires a data checkout option which can be used with other tools (e.g. https://github.com/alard/warc-proxy).

I use Safari web archives for a similar task.

But I wonder...isn't it safe to assume that, eventually, browser rendering engines* will change to the degree that something I saved 4+ years ago is essentially unreadable?

And doesn't that same potential problem apply to a hosted service as well?

*I'm using that vague term to describe everything I don't understand about how browsers render pages, markup, and javascript, which is a lot.

I agree. I think the safest bet would be to keep a static image as well. Without a static image would will always question if the rendering engine has even slightly changed the look.

Maybe save as a PDF if it's the content you're most interested in.

I save screenshots for that, but when I save an archive it's usually b/c I'm interested in referencing non-static elements. Like animation, transition, responsive behavior, etc.

Interesting in light of this discussion: https://news.ycombinator.com/item?id=6504331

But to trust that something like this to make a permanent copy of stuff I'm linking to, I'd need to know a bit more about them. Else this is effectively like using a link shortener -- a single point of failure.

Hey creator here, we realize there is a bit of a trust issue, thats why we have the paid plans. Our costs are low enough that, even if you were our only paying customer left we would be able to keep the service running just for you.

What Web services are you using in the back-end (that may potentially shut down)? I need my data accessible for low number of decades.

I have been using the Scrapbook add-on (see screenshots and manual here : [1]) in Firefox [2] for many years for this; it saves the web page to your local hard disk, and there are several types of annotations that you can perform on the saved page. One trick I use is to first run Readability on the page to get a clean version, and then save that to Scrapbook. With full-text and comments-only-search, this add-on, all by itself, kept me with Firefox even during the dark period when Chrome came in and thrashed Firefox on performance :-)

I used to use diigo.com, which does the job quite well, too, before I discovered Scrapbook.

[1] http://amb.vis.ne.jp/mozilla/scrapbook/ [2] https://addons.mozilla.org/En-us/firefox/addon/scrapbook/

I use evernote premium ($45/year) which is much cheaper and includes a tonne of extra features.

Compare this to the pricing plan of mummify at $15/month pricing or $180/year. for that price I could buy a pair of HD with several terabytes of capacity and copy/paste the whole webpage, code, files and all using httrack.

One point to note is that a DMCA takedown targeted at Mummify.it will remove the content just as it would from the nytimes.com.

If I manually save that content to disk then any DMCA take down doesn't affect the content stored on my local hard disk.

If it's something personal, I use http://archive.is with their bookmarklet in Chrome (free; unlimited archiving). It immediately renders the page, saves a copy with a unique url, and gives me a .zip link to download the archive.

If it's something I want to submit to the Internet Archive, I use wget with WARC extensions (http://www.archiveteam.org/index.php?title=Wget_with_WARC_ou...), submit the archive to the IA and notify them it needs to be merged in, and keep the .tar.gz archive.

Eventually I'll webapp/one-click the whole thing, with an archive to S3 and/or Glacier.

Disclaimer: archiveteam participant


Good point. Perhaps a "download catacombs" feature?

This feature would make it much more useful.

What if the server was not in the US? A DMCA takedown notice would not apply.

That's actually worse, since if there's no DMCA-like safe harbour, they can be sued immediately.

Most of the time, one will not know which pages one links to will disappear in the future. This service therefore only seems really useful if you use it with every link you make. That would make the biggest plan short on "mummies" by a couple of orders of magnitude, at least.

Not every page one visits is worth storing. The whole point is to filter out (and save) the good from the rest.

An aside: Mummify needs browser plugins for all the major browser to make saving as seamless as possible.

> Not every page one visits is worth storing. The whole point is to filter out (and save) the good from the rest.

If this is a tool to "fight link rot", then we're not talking about every page one visits, we're talking about every page one links to. Presumably, every page worth linking to is also worth saving.

Really like this, nice work. Is there a way to automatically mummify every link on an entire website or directory? Your method/service could be an interesting, easy to use way to recall a read-only version of a site after it goes down.

Do you rip video on the page too? If so, I would definitely use this service. I've got videos with my company brand a bit everywhere and I can't find a good way to download it or link to it so that it stay there permanently.

Just a thought: the "Link Rot" link should really be a Mummified link. I know Wikipedia isn't disappearing any time soon, but it just seemed silly to me not to "realise" the use case!

Anyway, very sexy site design IMO.

This is pretty cool, I'd be happy with a service that mummified it to storage I control, since I've had pages that I've pointed to at archive.org vanish after the current robots.txt was changed.

The examples of the page don't work. Would be nice that they have some working examples right at hand to see how the redirecting or the mirroring work on each case.

Your logo looks like a dog with a boner. Cool concept though.

Haha I actually thought the same thing. It's like the FedEx logo... once you see the boner, you can't not see the boner.

Thanks. Fedex and Tom Cruise's center tooth.....

Ha! I didn't realize the Tom Cruise Tooth one until this: http://www.buzzfeed.com/lyapalater/you-may-never-be-able-to-...

Let's centralise the interwebs!

It seems like a bad idea, but they do have a point. Maybe when referencing a link also make a small note to one of these archive sites?

Maybe also offer a callback with the contents of the cached site, so developers can handle 404 redirects themselves.

Things like this should be required when posting links in Stack Overflow and the like.

Rant: I absolutely hate when someone posts an answer in Stack Overflow and it points to their blog, which then stops working at some point. What if Wikipedia was just a collection of links to blogs or other datasources (citations aside). Put your damn answers in Stack Overflow people!

do you get 10 additional mummies per month? Assuming this site stayed open could be a great help to sites like stackoverflow.

What about SEO?

A 301 redirect should pass link juice fine.

This is an example of how great branding can help explain a product. Love the name, love the look, the copy is clever... but I don't like the low number of mummifications you get per plan.

I suggest adding more value or lowering the monthly price.

Overall, very cool.

First I had to make an account.

Then it doesnt work. Stuck at "caching page".

I hate you.

From your profile:

> Intelligent but bad at communication. Trying to improve my communication skills.

Start by not posting "I hate you" to people on the internet.

Sorry you are having trouble with it. Sometimes the caches take a bit to generate. Please feel free to shoot us an email at help@mummify.it if your problem persists.

FYI, for those of us with Firefox's NoScript extension your site is completely broken. Nothing loads, it is just a blank page with a list of scripts you intend to run.

I wasn't curious enough about your service to whitelist your site, but I will leave a comment asking that you consider providing alternate content.

It is rare to have a site completely fail to load, even with all scripts blocked (though I accept I will have reduced functionality).

You're running an extension that purposely disables what makes the majority of the web work. Unless you disable it, you can't really complain about sites not working correctly for you.

HTML is what makes the majority of the web work, with additional functionality (optionally) provided by scripts. When I see a blank page I wouldn't call that 'not working correctly' I would call it 'not working at all'.

Maybe we agree to disagree now, and you go on blindly trusting all websites to execute random code on your machine. I am perfectly content to have new websites look a little funny the first time I visit.

"HTML is what makes the majority of the web work" - Yes, ten years ago.

The modern internet is javascript. The code is executing in a secure sandbox. If you can get it to do something random on your machine make sure to let Google know, they'll send you a pretty big check.

It's not just doing something funny with the machines, it's also doing funny stuff with other websites. Plenty of websites are still vulnerable to XSS and CSRF.

Having used noscript for a while, I've gotten into the habit of letting it block everything at first and only allowing what seems necessary for that site.

I'm perfectly okay with the 'broken until enabled' model... I used to use "Request Policy" in Firefox - the most granular control of script and XSS accesses I've ever seen. Miss it in Chrome.

Let's hope you never need to use a screen reader.

This is a red herring. The overwhelming majority of screen readers support javascript front ends. As in, over 98% of people who use screenreaders use one that supports javascript.

source: http://webaim.org/projects/screenreadersurvey4/#javascript

It's a Backbone site =p

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact