As my work and life is spread across a few computers, keeping track of my bookmarks becomes a bit of a nightmare if I just use the browser. Even if I’m on my laptop, I have to remember to look in Safari and Firefox in case I’m in the other browser when I’m looking for a site I bookmarked.
This isn’t really a problem, as sites like Delicious have been around for many years now and are very good at storing bookmarks online, and helping me keep them organised.
Delicious has competition from Blinklist, Stumbleupon, A1 Webmarks, and a plethora of other sites. Even Google has a bookmark service. There’s also open source software for it so you can host your own version if you really want.
Last year I started hitting problems with bookmarks, and wanted features these sites don’t offer. Being a developer, I immediately started wondering if it was worth writing my own bookmarking site, or at least hacking around with an existing project and extending it the way I wanted.
Wants:
Keep a copy of the page I bookmarked – I am fed up with having a long term store of bookmarks where the page doesn’t exist any more when I follow the link. If I have a cached copy, it doesn’t matter if the original has disappeared. Potentially I don’t even mind if it’s just cached text, but having the images as well would be good.
Within this, I don’t want the page refreshed after I’ve bookmarked it, in case it’s changed to something I don’t want. Being able to bookmark the home page of a blog means I don’t have to think about finding the internal page where the content won’t change. I just have a copy of it in the state it was in when I bookmarked it.
Let me search on the content of those pages – now I have a copy of the pages I’ve bookmarked, let me search on their content. That’ll help me find the thing I was looking at the other day and stored, but can’t remember enough details to find properly. Having a filter so I can only search on bookmarks from a particular time period would also be good.
I thought Google bookmarks had this, but they run their search against their main search index, so if the page you bookmarked changes or disappears, they don’t have a copy of it any more. The search is great, but I don’t want the contents of it to change over the long term.
I thought up some other little features too, like having my account follow and bookmark links automatically from various Twitter accounts – both mine and other people’s, but they were really little gimmicks on the side of the main features I want.
Given that I was working very hard on client projects last year, then didn’t have much time this year as we’ve got a baby now, these ideas stayed in my notebook and kicking around in my head, gradually refining in to what I wanted. I did get as far as getting a domain for a test site and putting up a copy of ‘Semantic Scuttle’ but not as far as getting it to work properly (I don’t understand why they made the code so complicated to install for such a simple site.)
I started thinking about how I could get something like Solr installed to do the search side for me, and what sort of hosting it would need and whether it’d be simpler long term to ditch the Scuttle code and write my own, and go for a ‘NoSQL’ database store like Redis rather than MySQL.
Once my test site was up I saw one downside of running a bookmarking site: the darker side of the SEO industry started to post hundreds of bookmarks in to it. I don’t even know how they found it, I hadn’t linked to it from anywhere. All I’d done is loaded in my bookmarks from Delicious and let it sit there for a couple of months while I worked on other things. The site isn’t even working properly yet. They don’t get any benefit from being linked to from the site, but that doesn’t stop the links flowing in.
So that’s an issue. Will a new bookmarking site get over-run by people trying to use it to get the perfect link text to their own site? You can nofollow links, but that doesn’t stop people. A lot of the dafter, cheaper link builders actually doing the linking work won’t care or potentially know what nofollowing is about.
Anyway, I had my test site, albet a bit over-run with spam and not working. Soon after, I found out about two sites that pretty much solved my problems, with me having to do anything.
Two sites, two solutions
The first is Klektd, a bookmarking service being put together by Andy Kent.
It didn’t do everything I wanted when I signed up a few months ago, but it had a very nice interface – thumbnails of the sites I bookmarked generated very quickly and the promise of searching on the content of the pages, which wasn’t working then but is now and is exactly what I wanted. Although it wasn’t visible then, the search made me hope for caching of the original pages, and that’s now in with some good controls for refreshing them if you want – very handy if you want to be able to search comments on a page that have appeared since you first bookmarked it and you’re sure a refresh isn’t going to destroy the reason you had a cached copy of the page.
After finding Klektd, I then saw Historious.
Their interface is built around search, and they also show you the cached pages that they are storing to allow the search. It feels more like using a personalised search engine rather than a bookmarking service, which could be good or bad depending on what you like. This is a paid service with a free service for up to 300 bookmarks, which has only recently been introduced so I’ve only had a cursory poke about in the site.
They both have another feature I liked and has existed in bookmarking services for years: sharing links in public if you want it.
With only a few bookmarks in either, I like the interface of Klektd a bit more than Historious at the moment, but that may change as more sites get bookmarked in them and I’ve used Historious more. I was wondering how Klektd would handle having lots of bookmarks in it, as that’s lots of images to show, but of course they have a nice ‘load more’ function that lets you bring more in to a page rather than paging back and forth through your bookmarks.
Spam and adult sites
I’m not sure how either will handle spam. Historious has all it’s links to outside sites nofollowed, which might help reduce the amount of people opening accounts just for SEO linking purposes. Klektd doesn’t, but maybe that’s something Andy will bring in if mass linkers become an issue.
As they run thumbnails, Klektd may have some slight issues in the future: having certain accounts over-run by people linking to porn sites, which will leads to lots of adult thumbnails being on those account’s pages. This might or might not be an issue they have to think about if it becomes a business in it’s own right. It’s the sort of thing that bothers certain people and might prevent it becoming a really trusted site as you’re never sure what you might see if someone gives you a link to a Klektd page.
That said, Andy might have some sort of filtering going on to prevent thumbnails showing porn images, I haven’t tried adding a porn site to my bookmarks to find out.
Keeping going
When I was thinking about making a bookmarking site I thought about how to make money – advertising, a combination of paid & free(mium) accounts, not sure if I got much further. My scribbles decided paid & freemium was probably the best way to go, again just to reduce the problem of spamming. If people doing a lot of linking wanted to do it, they could just pay for an account like everyone else.
Historious seems to have come to the same conclusion, they started fully paid and have brought in free accounts to help introduce people to their service (from what I can see from some threads on Hacker News.)
Klektd seems to be a side project currently and still under quite rapid development given that status, so I guess it’s something they’ll tackle later. You can grow a bookmarking service pretty big on the cheap hosting that’s around these days, but it’d be nice to see some money being made to help with scaling if it does take off. It’s difficult to trust a bookmarking service that isn’t making money as you don’t want it closing and taking all your bookmarks with them because the owners are fed up with paying out for the hosting.
One less idea in the notebook
So, I have one less idea to work out. I could still do it, it’s not like you can’t build a different service to existing ones with pretty much the same features and still carve out a nice niche for the way you do it. But, this was always more of a mental exercise and not really a passionate need for me. I like the way both sites have brought in the features I thought were missing, and I’m looking forward to seeing what they both develop in to.