link sharing
social bookmarking

Interview: A conversation with Larry Halff about the relaunch of Ma.gnolia

Next Story

Update: Netflix announces exactly who won that $1,000,000 prize

logo_300Many of you may remember Ma.gnolia—the nifty social bookmarking tool that unfortunately imploded at the beginning of this year. Founded by Larry Halff almost 4 years ago, the site had a different aesthetic and attitude toward sharing information. It was one of the more community-minded tools I remember from that era, offering features like the ability to “thank” the sharer of a useful link, for example. It also possessed clean design and careful site organization. In my opinion, its take on sharing data really differentiated it.

Like many great things, Ma.gnolia didn’t start out to be big, but rather started out to be good—and it was. And, as is often the case with things that are good, Ma.gnolia become big by virtue of that goodness. Ironically, even though the membership of the service reached hundreds of thousands of account holders and tens of thousands of regular users, the infrastructure supporting the site was still incredibly small. It was run almost solely by Larry and the hardware and bandwidth he could support by himself. Unfortunately, there were some technical limitations to the honorable yet fragile DIY set-up running behind the scenes that ultimately led to the site’s premature demise. I was really bummed to watch the VOD-cast explaining the catastrophic nature of the data loss back in February and have thought about the site often, since that time.

I was able to catch up with Larry a while back and talk with him, not about what went wrong with Ma.gnolia 1.0 but rather what is in store for Ma.gnolia 2.0, if anything, and also pick his brain about the future of social bookmarking. If you were a fan of Ma.gnolia in the past, you will be happy to know that it is scheduled to relaunch September 22, by invite only.

JD:
What is your take on the future of social bookmarking?

LH:
Well, I think there are two ways it can and is going. One way is the sharing of links. You know, sharing a link with someone may help you record a memory of something but it’s not really an archiving activity, and I think the biggest link sharing site out there is probably Facebook with people posting items to their wall and sending recommendations to each other. And then I think there are the sites that are more about saving and recording things and keeping them around for future reference. I think that’s definitely Delicious and the slew of other traditional social bookmarking or tagging sites out there. I think Ma.gnolia was sort of in between, in a way, in that it didn’t know which it wanted to be because it had a lot of the social functions built in, with sending bookmarks to people and sharing them with groups and thanking people for bookmarks. There was some of that stuff built in that sort of straddled the line between those two things [social functions and archiving]. I think, as “The Web” progresses, and as things become more component-ized people will start assembling the services and applications they want to use around “Identities”, whether they’re OpenIDs, Google IDs or Facebook IDs. We’ll start seeing “sharing and storing” services evolve along those directions. I don’t know that there’s really a huge “future of social bookmarking”—I think it’s that sharing and saving links is just one of those things that people do and I don’t think it’s going to be like Search where there’s a huge industry built up around it.

JD:
OK, so social bookmarking is more of an evolutionary meme or practical tool but not some revolutionary business model? Is that what you mean? Like in the case of link sharing on Facebook—it seems casual yet still purposeful.

LH:
Yeah. I think the idea of social bookmarking as this distinct thing apart from the services that we have gathered around us, is going to disappear. The direction I’m going with Ma.gnolia, with reviving the service…it’s going to be private and focused on the community of users who are drawn to the social aspects it offers as opposed to trying to be a big service all on it’s own in a market that would be seen as big as the Search market.

JD:
That leads to another question I had for you. Are there any ways you want to point out that the new version of Ma.gnolia will be different from the old version?

LH:
Initially when I launch Ma.gnolia soon—imminently, it will be the same Ma.gnolia with a few tweaks, like making registration by invitation, fixing a few items…but it will be the same code and rather than pursue the rewrite from the ground up, as initially planned, I’m just going to be refining the code base and aiming to get existing code to a point where it can be open-sourced  and look at iterating that more toward a distributed, component-type model.

JD:
So, one thing people can expect is not to see a complete re-write… the general concept of it worked well enough before, so you may keep a lot of that infrastructure? Is that what you are saying?

LH:
Yeah, exactly.

JD:
OK. And after you re-launch it, you want to open it up to the community for development and improvements?

LH:
Yeah, but I need to polish the code base to the point where it can be opened up to the community.

JD:
Great. Can I get an invite? (laughing) I’m shameless.

LH:
Yeah, definitely (laughing).You know, I’ll shoot you a note when it’s ready to go. I have 21 days until the end of summer. [our conversation took place on Aug 31]

JD:
And that’s the zero hour?

LH:
My goal (laughing) was to have it up by the end of summer.

JD:
You know, one thing I wanted to point out was that Ma.gnolia was different for me in a couple of ways, namely that it embraced OpenID early on (I seem to recall that it’s the first site I remember using OpenID—it actually got me to sign up for myOpenID) and also it really embraced aesthetics. I come from a design background and I always appreciated that aspect of it.  Did I read correctly that Jeffery Zeldman helped design the front end?

LH:
Happy Cog did our original design and interface design. And you know, there’s no reason that the tools you use should be ugly and unpleasant to look at so that was one of the way I felt that [Ma.gnolia] showed consideration to the user.

JD:
OpenID integration and design aesthetic were two ways Ma.gnolia was different for me, as a user. How was it different for you, being behind the scenes?

LH:
I think we really saw our users as people, in social context. When we built out Ma.gnolia, everything from OpenID to the way groups ran to the “Give Thanks” feature…it was [always on our minds] “how does Ma.gnolia exist in the greater context of our members’ lives”. OpenID was an important part of it because we wanted it to seamlessly integrate with an identity system that they were in control of, that they felt ownership of…that YOU all felt ownership of. We were not there to gather up identity, we were there to help you use your existing resources. Also, with “Give Thanks” we wanted to encourage positive social interactions among our members rather than forcing the social stuff in a marketing-heavy type way. You could instead discover other members by finding out who appreciated what you were doing.

JD:
Kind of like a reverse lookup by appreciation or something?

LH:
Yeah. So instead of being like “Here, follow my stuff. I’m important. You should listen to me”. Instead it’s more like “Oh, I think YOU’RE important and I really like what you’ve been doing. I really like the stuff you’ve found”.

JD:
I ditched Google Notebook for Ma.gnolia. After Ma.gnolia though, I never really found a replacement (I re-flirted with Delicious and Digg and even Evernote). I have since moved on to Pinboard (at Michael Arrington’s recommendation). What do you use to bookmark and share links these days? What would you recommend, in the meantime?

LH:
I actually just use my browser.

JD:
(laughing) OK. That’s cool.

LH:
I’m also using a “not-ready-for-prime-time” instance of Magnolia. I’m alpha testing it myself.

JD:
I want to ask you about the greater issue of “Sharing” in general. Sites like Sharemo in Japan, for example, are bringing the concept of sharing out of the “data” world and into the real world with the sharing of used physical items—books, clothes, etc. The concept is not new—people have been trading things forever—there are just new ways to organize and deliver the concept. The sharing of goods, data, whatever is efficient and necessary in all kinds of ecologies, online and offline. Here’s the question…what else is going on with this kind of sharing of goods and data? Is there beauty in tracking what is shared and following that huge pattern? I get the idea from what you have said that maybe it’s greater than even that… that there is no mystery in the sheer numbers but rather in the the emotional connections that take place around shared items. Does that make sense? How would you respond to that question?

LH:
[With regard to data sharing] I don’t think, at this point, it’s possible to gain any kind of bigger pattern out of that kind of sharing and it’s something we never really attempted with Ma.gnolia. I think it’s a misguided fantasy because whenever you start measuring [in that way]…you have no real quality of measurement…you have no idea really what you are measuring other than that these links are being sent around. You don’t even know if they are actually being shared. You don’t know if the receiver cares about them, you don’t know if the sender cares about them, you don’t know if the sender and the receiver are the same person just behind different accounts. On that scale, I think its impossible to draw any conclusions about what’s happening beyond the fact that your system resources are being used  to send these URLs around. I think that sharing only really makes sense to the people who are the senders and recipients of those links when they are being shared. I think the sites that try to harness that sort of flow in information, like Digg, have a real hard time doing it and they have to put in all sorts of guards and protections and try to figure out gaming patterns and I don’t think, even then, you can draw any sort of meaningful conclusions from what happens there. In the end it’s just another game to watch.

JD:
So you can’t get qualitative out of quantitative?

LH:
Yeah, exactly. I think what you get is information about how people play that particular link sharing game. You don’t get any particular info about the links that are being shared, themselves.

JD:
Do you have any plans for doing any kind of mobile integration? People are coming into contact with and sharing information despite their location.

LH:
We had a mobile version of the site that a lot of our members were pretty happy with. It was just an alternate view in the browser. I think that, considering you view links from a browser at this point, it satisfied a lot of people. Of course, it would be fun to have an iPhone app that could search and get through your stuff in a more native friendly interface, but that’s not a top priority. It’s on the list but I don’t know how important it actually is.

JD:
Is there anything else you’d like to tell me about Ma.gnolia or otherwise? Any new ventures that may cross-over with it?

LH:
No. You know, I’ve just been recovering, taking care of other things in life and figuring out a strategy for the best way to bring Ma.gnolia back. And I’ve weighed some of the pitfalls of our past setup. It was a very heartening experience in the sense that the community was, overall, very present and understanding of what happened and most people pretty much said “there isn’t a replacement for Ma.gnolia and we want to see the service come back”. And that’s sort of what I keep hearing and it definitely…I mean, I put three years of my life into it [Ma.gnolia] and seeing that it really did matter to people was a wonderful thing. I’m really looking forward to having Ma.gnolia up and running again and having a good community of people using it.

blog comments powered by Disqus