October 31, 2005
Hmm. My last post from Sunday morning is no longer available. And here I was trying to get into the habit of writing every other day.
Such are the perils of relying on a (free) third-party service.
Update: Looks like it’s back. Not only was wp.com able to restore the old post, but after a short time of absence, this later post was also restored. Good for Matt. Usually when sites crash, you lose data from one of the three critical periods: pre-crash, during-crash, or post-crash.
October 30, 2005
For quite a number of months–since January 2004, according to my archive–I have preferred, and advocated for, Furl over Delicious. There were several reasons for this. I thought the UI was better; Furl includes full-text search; finding other people or articles related to my interests was easier; and it was much easier to create personalized, ad hoc hierarchical taxonomies.
This latter point is important because, in most contexts, I am more of a “hierarchical” minded person than I am a “bundle” minded person. (One big exception I’m realizing now is with bug databases, where hierarchy interferes with usability, and bundling is much more useful and intuitive. I may be convinced that the same applies to bookmarks. We’ll see.)
From an implementation standpoint, then, I’ve been hoping to see–or create–tag-based systems with many of the advantages that Furl has.
But lately, I’ve become increasingly frustrated with some of Furl’s limitations, and have considered switching to Delicious or Simpy. (Somebody at work also tells me that The Shadows is good, though I haven’t used it.) In fact, the thing that is keeping me tied to Furl during my busy times at work is really the lack of tools to import my Furl archive into any other social bookmarking service. Which, as far as I can tell, is certainly not Furl’s doing, since they provide their own XML export.
Here are what bother me most:
- Filing into multiple categories is hard. At first, this did not bother me, but as time has gone on I realize just how hard it is to apply multiple topics to an item using a select box. That I cannot change the size of the select area, which means I have to manually scroll through all my topics 4-at-a-time, which is more problematic since any ad hoc hierarchy is really a flat space, is a hindrence. More often than not, I don’t even try anymore, which means my bookmarks become more difficult to navigate later on.
- When I use the bookmarklet to “Furl” an item, and it has already been Furled, I get a nice little popup that tells me it has already been Furled. But then the edit box doesn’t reflect any of the information–like rating, or topics, or comments… and if I Furl it again, it’s a distinct item in my archive. All bad.
- Related to the above, editing (e.g., moving from “to read” to whichever proper category) is too difficult. Leaving aside my objection above, even going into my archive and trying to edit an entry requires so many clicks and contextualizing that, again, I usually don’t do it. So I still have things in my “to read” topics that I’ve already read, because it’s just too difficult to manage them.
- Recommendations–one of the things I thought was a huge advantage of Furl over other sites in the beginning–are not dynamic enough, and stabilize over time. It looks like their algorithm is not tuned to weight recent interests more, or to randomly boost certain recommendations every week, or any of a host of other things that could be done. What’s worse, their people recommendations–what they call “Furlmates”–has no management function. With article recommendations, I can Furl it to make it go away; or just click delete and it will go away. But with people recommendations, there’s no way to make a recommendation go away, whether I like that recommendation or not.
What’s the positive? The positive to this is it’s a great learning experience for me to see what it’s like, as a user, for a service you initially love gradually degrade relative to their competitors, not as a result of doing anything wrong, but just for being content. My employer has historically had the same problem, and sometimes, when you’re on “the other side,” it’s hard to really understand what’s going on in the minds of your most loyal members.
Plus, I think those of us who have used Furl, and not just followed the pack into Delicious and their imitators, have a different–and maybe better, in some respects–perspective on social bookmarking and related services.
October 28, 2005
I’ve been thinking a lot lately about communities / social networks as a means rather than an end, and what kinds of tools different communities will find useful. Of course, part of my job is to think about these things every day, but one tends to get bogged down in the mundane details of immense scalability and office politics, and needs a jerk back to see the bigger picture occasionally.
Like a lot of people (I’m betting), I was intrigued by the Mark Zuckerberg interview blogged by Jeff Clavier. Mark is the founder and CEO of the Facebook. During the interview, Mark asserted–correctly, I think–that one of the (primary?) reasons the “first generation” of social networking sites failed is that they didn’t provide a set of quality utilities. The network was the end, not the means. As Greg of Geeking with Greg writes:
Social networking sites like Orkut or Friendster have no purpose. Sure, it’s fun. You go there and, in a flurry of activity, set up your profile and list all your friends. It’s always good for a little ego pump.
But, then what is there to do with your social network? There’s no purpose, no reason to come back, nothing to do.
Think about this from an implementor’s standpoint. You spend your months (or years) creating a community with an awesome search and people finder. You have top-notch content, and maybe a great editorial team. Your members contribute their own outstanding material. So people find each other and are entertained by each other. And then…….? Hmm.
Then this morning on the Webshots blog (yes, Webshots has a blog now! I’d say “it’s not much yet,” but they already have more comments than I will in the next year), a member suggested a follow-up to our Hurricane Katrina/Rita/Wilma collection that features the devastated areas in better times. What an awesome suggestion.
Although the collection started with Katrina as a way of funneling donations to the Red Cross, Humane Society, and so on, it has become much more than that, and it only includes a handful of the public photos that can be easily found.
One can imagine follow-ups that track these areas during their rebuilding efforts, and promoting before-during-after photos a couple of years hence. The possibilities wouldn’t end even there.
But as good as Webshots’ editorial team is, and as much weight their collections carry, I can’t help but think that what members really want are tools to produce their own collections, and a means of promoting them (and maybe connecting them to external organizations).
It’s certainly more scalable, but, like television, it could create information overload, and with it, the risk that people will think there’s not much of anything good out there. Hence, the need for even better tools: tools not just to share and find–what we might call “first generation” social networking tools–but to promote, augment, incentivize, and (again with my silly ideas) connect with other communities outside of the walled garden.
What this spells to me is not just community, but also media: a platform for and a marketplace of mini-media outlets.
So how do you make that leap, make it meaningfully, and transition your existing communities to a better world without scaring them off?
October 22, 2005
I’m a little late on this, but Tyler Cowen of Marginal Revolution recently asked Should we confiscate Tamiflu property rights?. Tyler says no, and instead proposes to offer large prizes for speeding up Tamiflu production. This is not really any better than the proposal he counters.
Tyler’s motive is to change the incentives for Roche to manufacture more Tamiflu, and to emphasize the incentives for drug manufacturers in general to pour money into R&D of new drugs. Being an economist, Tyler understands the role of incentives.
Personally, I am afraid of any policy that reverses Roche’s patent. The argument runs that this is an emergency, so reversing a single patent in an emergency, which can help save lives, will have no effect on the incentives for drug companies to continue to throw money into R&D for new drugs.
But what’s getting missed is that mass-producing Tamiflu DOES NOT GUARANTEE AN END TO THE CRISIS. In fact, many experts are forecasting widespread resistance to Tamiflu should H5N1 or another strain mutate. In 12 months, any Tamilfu stockpile we have could be completely useless.
What does this mean? It means drug companies need to throw money into R&D for Tamiflu replacements, and/or into mechanisms by which Tamiflu or other anti-virals can be produced more efficiently/quickly. They need to do this now. But what is their incentive for doing so, if they know any patent claims to their results will be thrown out? That’s right: they have vritually no incentive.
Similarly, if governments offer large prizes for mass producing Tamiflu, guess what? Drug companies–even Roche’s competitors–will have little or no incentive to research and develop alternative, and possibly more effective, medicines. It’s clear to me, then, that while Tyler’s proposal might be ideologically purer, it’s still solving the wrong problem.
So what happens if, tomorrow, every member of the G8 decides to give every drug company free (or cheap) reign to mass produce Tamiflu (or large incentives for Roche to do so)? The result is we prepare for one scenario, at the risk of being massively underprepared for the other, say, 50% of likely scenarios.
From a risk avoidance perspective, I’d rather take my chances of being able to (a) avoid contact with others, and (b) acquire Tamiflu if I need it, in the scenario where Tamiflu remains effective, but is underproduced. That is preferrable to having a stockpile of Tamiflu that could be administered to every human on Earth, but have no alternative if and when the dominant flu strain becomes totally resistant.
And that’s to say nothing for the ability of governments to administer Tamiflu efficiently to those who need it even if we had a stockpile large enough for everybody. Or of the increased likelihood that the dominant flu strain will develop complete resistance should every government and health worker on Earth start administering Tamiflu–and often ineffectively–to every infected or potentially infected person.
I’m hoping that politicians from every nation understand this, are developing their strategies accordingly, and are not coerced into going along with everybody else. The result could very well be a host of different strategies applied in different regions, with the most effective winning out in the long run. Why? Because there’s not a single solution that will save the lives of 100% of infected persons, or prevent wasting valuable resources on solutions that turn out to be useless.
October 3, 2005
A quick run-down of my blog so far, more for myself than anybody else.
Number of entries: 3
Frequency of posting: about once every 2 weeks
Number of links: 2
Number of links into ongoing conversations: 0!!
Average length of entry: long
Focus of entries: Self, mostly.
Style: Somewhat formal.
Voice: Not sure.
Number of comments: 0
I’m not sure what other metrics to use, but I’d summarize my first month as a timid, uncommitted exercise that is too insulated and not reaching out into ongoing conversations happening “out there.”
I’ll act to change that in the coming month.
October 2, 2005
Maybe this is a silly question, but I want to follow this line of thought.
Let’s start with this answer: search is a textbox, into which users enter text queries in order to retrieve relevant results. Sounds kind of boring, doesn’t it? Not that there aren’t a lot of interesting challenges here: content relevancy factors, social network analysis, targeted relevancy based on submitter behavior, and so on. In fact, a lot of resources have been poured into improving upon the “search is a textbox” mode of operation, and a lot of really cool innovations have come from it. The result is a firm mental model among users that search == textbox.
What about tagging? Tagging, like search, is an explicit activity: both in tag creation, and tag browsing. It is a bit more compelling than plain “search is a textbox,” because it offers additional navigation elements that are closely linked with the content. Some problems that are intermediate or difficult in the “search is a textbox” model of full-text indexing–such as clustering, related content, and so on–become technically simpler with tagging, and integrate well with the existing tag browsing UI. Users also gain more benefit from tag systems if they have a relationship with the organization. So if you’re a tag system developer, are you developing a search system? Does it matter whether the back-end is primarily MySQL, or Lucene, or Berkely DB?
The next level “above” tagging is faceted navigation, especially where tags (or metatags) become facets. Faceted navigation merges tagging and textbox searching into one UI. One way of thinking about faceted navigation is as an interactive boolean query builder. Faceted systems make a ton of queries into the back-end to fulfill their promise of providing context-sensitive navigation, and benefit from efficient search algorithms. A faceted navigation developer is obviously a search developer, even if all the context is pre-generated (good luck with that strategy, though).
The context-senstive links in faceted navigation are, of course, primitive (or not-so-primitive, depending on POV) recommendations, which brings us to recommender systems. Recommender systems usually rely heavily on statistical analysis, but then so do more traditional search features such as targeted/personalized relevancy and clustering. Recommender systems can be entirely driven by profiles consisting of a small set of keywords based on past (or recent) behavior. Recommender systems can be used to recommend just about anything: news stories, user-generated content, fellow members in a community, editorially “programmed” content, advertisers.
But now we’re talking about more than a textbox with a submit button. We’re talking dynamically generated content, which might even look like ordinary content, with ordinary navigation choices. That’s the heart of personalization, isn’t it? Isn’t that the dream we’re all (users, producers, pundits) sharing? That can’t all be driven by something as mundane as “search,” can it?