At work we have a couple of projects which have leapt to the forefront of our priority list, which basically means that I won’t be able to work on Beyond at work for the next two weeks or so. But fear not, that’s what spare time is for. :) I wish I didn’t have classes this term, though, because homework’s eating up a lot of it.

Anyway, I’m looking into SOAP vs. REST for the web services part of Beyond. Looks like Google’s built on SOAP, Flickr provides both SOAP and REST interfaces, and Ruby on Rails seems to favor SOAP as well. The general feeling I’ve gotten from reading blog posts and articles about the debate is that SOAP is more complicated. What I’ve seen of REST certainly was simpler. I’ll look into it some more and see if it would be worth my time to support both, like Flickr. Anyway, once I get some time I’ll try to cook up a little test app to make sure I understand how web services work, sending XML back and forth like tennis players at Wimbledon.

Technorati Tags: SOAP, REST, Google, Flickr, Ruby on Rails, XML

Five minutes ago I got an e-mail from one of the genealogy lists I’m on (for my Crowder surname), and it looks like the person is doing research in the same area I am. So I started a reply, and then when I got to the point where I was going to describe my line there (e.g., John Franklin Crowder b. 1887, William Crowder b. 1809, etc.), I realized I couldn’t remember some of it. No big surprise there. :) And I wanted to open PAF to see what I had, but I’m on my Mac and I couldn’t. My next gut instinct was to open a new tab in Firefox and view my genealogy in Beyond. Easy, quick, and portable.

Too bad it’s not quite there yet. :) Yet another motivation to press onward…

Okay, I fixed the problem. Here it is (click on the image):

It doesn’t work quite right in Internet Explorer, so I recommend using Firefox. (Keep in mind that these are just mockups and that the real thing will of course support IE.) The colors will change. I’ll have other mockups soon (of the pedigree page, the research page, and a couple others). There’ll be more information on the dashboard, I think. We’ll see.

Thoughts on making mockups of interfaces: it’s really, really tempting to fine-tune everything at this point, especially since I do a lot of graphic design on the side. I want it to be beautiful. But that’s not the point of the mockups — it’s more about getting the flavor, the general idea, not the specifics. It makes sense, too. Why would you want to spend a lot of time tweaking something you’re going to end up throwing away? The point of making these mockups is to see what’s necessary and what isn’t, and some of what appears will almost certainly get thrown out. That’s why I’m not going to worry about getting these mockups to work in Internet Explorer — that can wait till the end.

It’s exhilarating to see it start to come to life. :)

I spent a couple of hours this afternoon putting together an HTML mockup for the dashboard, but then had a unexpected two-hour-long meeting. I’m relatively pleased with the results so far, especially for such a short amount of time. There are still a few little problems with it, though, so I’ll wait to post it till later. (And I haven’t done the mockups for the other parts of the app.)

Technorati Tags: genealogy

Ran across a genealogy/Google Maps mashup today. That got me thinking about maps. The key about maps as far as genealogy goes is that things change over time. Boundaries move back and forth — just look at Alsace-Lorraine! — and using a modern map isn’t always helpful because of it. Not to mention the varying levels of detail you’d want, and other sorts of non-political boundaries (diocese boundaries, for example). Aye, it’s one heck of a problem. I don’t think the data itself is necessarily a problem, nor storing it — the problem is that the data isn’t freely available. Google Maps is doing a good job but it’s all present-day (which is fine for its main use, of course).

I’ve thought about user-contributed map data, but no matter what angle I look at it from, it seems like a bad idea. It’s the sort of thing that should be left to cartographic experts.

In an ideal world, all the data would be available for free, and you’d be able to pull in just the part you want — say County Down in Ireland in 1810, for example. (Meaning, the individual counties and parishes and such would be discrete entities able to be separated from the rest. It seems like much of the free boundary data out there just describes the boundary lines themselves. But I could be wrong.)

Ah, if only this were an ideal world…

Technorati Tags: genealogy, maps, Google Maps, cartography

I got a PowerMac G4 at work today. It’s old, sure, but it’s running Tiger and has the latest Xcode, which means I’ll be able to develop the OS X client in parallel (in Python with PyCocoa, I think). I’ll still focus on the web client (which I’ve dubbed Orion, since saying “the web client” all the time is a) boring and b) longer than “Orion”), but once I get far enough along it’ll be nice to work on the Mac client (no name yet) so I can make sure the API is actually usable.

As far as the server (which now needs a name :)) and Orion go, I installed Ruby on Rails today and can start learning the ropes. I’ll be prototyping the interface and will hopefully have most of that done very soon, so I can get some feedback. That’s all for now, folks.

Technorati Tags: Ruby on Rails, Python, PyCocoa, OS X

Installing Subversion on Bluehost (a shared host) looks like it’ll be difficult, at least until they offer it officially, so I’ve decided to register on SourceForge and use their servers instead. It takes up to two business days for them to process it, assuming that it gets approved. We’ll have to wait and see…

Anyway, I chose the GNU Public License (GPL) for the project. Why? Mainly because it’s the one I’m most familiar with (that’s what happens when you use Linux for eight years :)), and I didn’t want to read through all the legalese in all of the others (MIT, Mozilla, etc.). The stuff on the website itself will still be Creative Commons, though.

Technorati Tags: SourceForge, Subversion, Bluehost, GPL, Creative Commons

Here they are:

I’m still not sure yet what I think of this idea. There are clear advantages in the flexibility department (along with integration with research pages), but there are a few unresolved issues.

In the first (John Smith), the info is pretty straightforward. (Does the right alignment for the source stuff make it feel less cluttered?) The second (Jeremiah Shanks) makes the advantages of this webpage-ish method more evident: you can include to-do lists (or any kind of checkbox lists) right in there with everything else on that person, including things like tables (as you see with the census stuff).

So now the question is this: does all that stuff belong in there? Even if we don’t do this on the individual detail pages, it’ll still be there in the research pages, so it’s not a life-or-death question. I’m torn. On the one hand, it’d be nice to have a to-do list and research log specific to each person, right there with the other information. You can take it all in with a single glance. On the other hand, should research be kept separate from the conclusions? I don’t know.

Another related issue is that of stuff you’re not sure about. For example, in my PAF file (which I inherited from several people, collated together into a hodgepodge mishmash of data with very few sources), I don’t know what’s real and what’s not. It’d be nice to mark it all as unsure until I can go through and check everything (and add sources in). I suppose not having a source (especially if the data was greyed out) would be equivalent to being unsure. And yet it’d be nice to be able to rank bits of evidence on more than just a binary yes/no scale. Something like the iTunes five stars…

That’s easy enough for data that doesn’t overlap with anything else in your pedigree, but what if you’ve got two possible mothers for your great grandfather? There has to be a good way to allow for alternate pasts, “what-ifs” if you will, and make it clear that you’re not sure which it is. I’ve got a few ideas on this but they’re still kind of fuzzy.

Technorati Tags: genealogy

Yesterday I came up with an idea for encouraging people to enter sources:

Namely, grey out information that doesn’t yet have a source. For me personally, something like this would certainly make me want to bring all my information to “full strength,” so to speak, and it would remind me at a glance what pieces of evidence are still unsourced. Thoughts?

Incidentally, seeing it laid out this way has made me wonder if Backpackesque pages would be a good way to manage the individual detail pages. There’d still be a pedigree, of course, but each person would have their own page, and users would be able to edit them the same way you edit a Backpack page, ordering things however they want. Advantages: closer integration with the research pages, and you could put anything on the page however you want (flexibility), and it’s more in line with the nature of the web. And you could easily publish a user page (make it public), or send the link to someone else. Disadvantages: it seems like things could get rather cluttered. And I’m not sure if it would actually work. (The thing that I’m not sure about is the connection between the individual pages and the pedigree. I suppose there could be certain database fields that show up in the pedigree, or perhaps make it flexible and let the user choose which fields they want to appear there.) Hmm… This is mostly in my head still, so bear with me if it doesn’t quite make sense yet. :)

If you haven’t already tried Backpack, go register (for free) and play around with it a bit. What I like about it is that users are free to organize their stuff however they want. That’s important. People are different, and I want Beyond to allow people to work the way they want to work. That doesn’t mean shrugging off all design decisions onto the user, of course, but it does mean that the system should be flexible enough to provide for individuality in research method.

This’ll require some thought… (And a clearer explanation so you can see what I’m envisioning.)

Technorati Tags: genealogy

Here are some quick mockups I threw together for the lightbox ala-Expose feature:

I read 37signals’ Getting Real last week, and it’s influencing a lot of my decisions (for the better). I’m still working on other parts of the interface, trying to figure out what’s most important and what’s not as essential as I thought. More to come soon…

Technorati Tags: UI, Getting Real