Addinging future-presence based on airline and hotel confirmation emails

I can vouch that the iCal import works great on the Tripit iCal feed. I'm pretty sure it's going to result in my using both sites quite a bit more. You guys really nailed that feature.

Hi Matt,

So sure, you could rely on the output of TripIt and import it into Dopplr (maybe you already can).

And as much as that's "small pieces loosely joined" and all that, I kinda think it's a "core feature" you really should offer yourself rather than rely on someone else to do for you.

I say that because TripIt has to make money too, and if TripIt simply becomes an email -> iCal parser for a significant number of users who just use it for Dopplr import then the business will suffer because no one will be page-viewing the site. And if they are offering a valuable service then they deserve to monetize it (I don't work for TripIt, but in my line of work I'm mindful for the biz models of APIs).

Perhaps Dopplr and TripIt could contribute to an opensource repository of email templates for parsing?

Ben

I can see two sides to this issue:

1) "Two people being in New York or Shanghai at the same time is academic if they are staying in opposite parts of town": I don't fully agree with that rationale and I wouldn't trust any website to do such fine-grained matching for me. It may take 30min trip across a city to see somebody, but what if I'll have lots of free time anyway, or if I haven't seen that person for quite a while. DOPPLR seems to move in the opposite direction, with the fuzzier matching (in terms of both space and time), and I been given the options and making the final decision.

2) I haven't used Tripit yet (will check it out today) and I don't know what it does to all the details that it gets from confirmation e-mail. I wouldn't want everyone to know which hotel I'm staying in or where I'll be having my dinner. I can imagine sharing such information with people who are joining me in some part of my trip, but should DOPPLR store such info it would also need some smart privacy model to handle this.

Follow-up after checking-out Tripit: I forwarded two reservations. First is from GetThere (corp travel booking website) which is claimed to be supported by Tripit. Some things got recognised, but there are some silly mistakes: my hotel name is "Inn" instead of "Holiday Inn", and my airline is "Airways" instead of "British Airways". Second booking was from Eurostar, it was not recognised at all although Tripit claims to support parsing Eurostar e-mails. My verdict? I really can't bother with Tripit if it has such a success rate and I need to correct stuff all the time.

And now for some constructive suggestions:
- All you need to identify a hotel location is an approximation of the hotel name & a big database of hotels. Tripadvisor's database seems to be rather complete, but I can't see any public API that would allow querying it :(
- Same for flights, you can get the flight schedule if you have just a flight number and a lookup API such us http://www.flightstats.com/go/About/x... (seems to be commercial, unfortunately)

I would prefer to log in to DOPPLR and past couple of hotel names & flight numbers, rather than forward a booking e-mail and pray that XYZ Airline hasn't changed their e-mail template... :)

"I would prefer to log in to DOPPLR and past couple of hotel names & flight numbers, rather than forward a booking e-mail and pray that XYZ Airline hasn't changed their e-mail template... :)"

I think is an example of the spectrum of travelers and where Dopplr will have to decide who it ultimately caters to. I don't know how many times a month you fly, but I'm on an airplane and in a hotel at least once a week. I don't have time to book the damn flights sometime, let alone add the manually.

Personally, I've never had any problems with TripIt's parsing - but then I don't use corporate systems - just the sites directly (united, southwest, marriot, hyatt, etc).

I don't think it's a matter of 'praying' that the airline hasn't changed their template - there are only a finite number of providers and thus it's not that hard to keep such a database up to date. And if an OpenSource repository and reporting mechanism was created then everything would be even better.

Adding the dates and times I'm in a city to Dopplr is an order of magnitude less frustrating than booking the flights and hotels in the first place. The whole point for me is to let people know where I'll be in case they'll be there too. Ben, it sounds like you're not really bothered about that a lot of the time... I would totally cross town in New York or Shanghai to hang out with my Dopplr contacts! If the trip was too short to bother doing that, I probably wouldn't add it to Dopplr. If you don't want serendipity, likely serendipity doesn't want you :)

I guess it comes down to this: will TripIt add trip sharing and coincidence features quicker than Dopplr adds itinerary parsing? And which one would you stick with if they did?

"Adding the dates and times I'm in a city to Dopplr is an order of magnitude less frustrating than booking the flights and hotels in the first place."

Sure, but it's a necessary dependency (unless your driving there and staying with friends I guess). So if that dependency chasm is already crossed then I reckon it's important for Dopplr to let me extract as much value from that process as possible by utilizing the valuable information contained in those emails generated.

"Ben, it sounds like you're not really bothered about that a lot of the time... I would totally cross town in New York or Shanghai to hang out with my Dopplr contacts! If the trip was too short to bother doing that, I probably wouldn't add it to Dopplr."

In general I would too, especially if I'm traveling back to my native London (I live in San Francisco). But for short hops to places like LA which I do weekly, I can't realistically jump over to Pasadena when I'm holed up in West Beverly Hills overnight.

"If you don't want serendipity, likely serendipity doesn't want you :) "

I do want serendipity, but I don't see why that serendipity can't come for free/zero-time spent. In fact, that then increases the chances of serendipity for my contacts too. It's a two-way thing.

"will TripIt add trip sharing". They already do. It's not as good as Dopplr and also friend discovery is non-existent so I have to add friends by adding in their email address.

But with my developer hat on, it's not hard to improve the friend discovery - and easier than adding email parsing... which is why I'm concerned for Dopplr. TripIt actually has all the nuts and bolts in place, it's just poorly designed right now.

Maybe Dopplr is trying to go after a different kind of person. Could be it. If it is then the subtlety is lost on me (and no doubt others). I look at myself - a very frequent traveler, who lives abroad, has contacts around the world but very time poor. Up until know I thought and assumed I was bang on Dopplr's target market. But if I am, they can't expect me to manually enter my plans when the data is already there. It's not very Web2.0, dare I say it.

i think that adding itinerary information would expand Dopplr's utility because then one could use Dopplr to actually plan a trip with people or store reference information for one's own or someone else's trip. That would make things easier if you were picking up a friend at the airport because you wouldn't have to fish out the itinerary they sent you a few weeks/days ago. Or if you were meeting up with people at a house you rented you could just stash the rental info on Dopplr instead of in some email or group thread. Or heck, if you and your friends travel a lot, you might actually have a layover in common and could meet up in the elite lounge. It's also good for people at work or home that might need to find you.