Ran across this yesterday while looking for more Open Source PIM stuff, and it seems like it would make a logical front end to the versioning system I described on Wed. The more I think about it, so long as we can prescribe that all data be stored in XML format, I don't see why good-ol CVS can't do the job on the backend, it just needs some set of "wrapper" applications to make the interface a lot more transparent to the end user. I hate it when big projects start to seem approachable, 'cause then I find myself working on them. Well, sanity was nice while it lasted.
Mozilla themes fixed.
Geez. I thought that would never happen either. Fooled me. I think I'm having one of those hyper-productive spells where you just know that you can do anything. It's not so much that I'm a better coder than I was yesterday, or that I've suddenly gained deeper understanding into the DOM than ever before, it's just that I feel like I can do it, so I do.
Just wait, come my Calc exam on friday, I won't even be able to look myself in the mirror, let alone code. Better get 0.2.3 battened down while I still can. =)
Sometimes good things happen.
I've spent months, months trying to figure a way around the netWindows table-dragging bug in Gecko. I have assumed for some time now that the problem is in the lookup functions that NW uses in order to keep track of components. Last night, a friend of mine (a DOM newbie) wanted help with some DOM stuff, and so I oblidged him and walked him through some simple dragging stuff. He too came across the dragging bug, but wasn't using any of the complex object lookup routines that netWindows uses. I was looing in the wrong place. It wasn't the lookup, it was the table. Placing the table in a div, and moving the div solves the problem.
Why is it that the widgets I need to work on aren't the widgets that I want to work on? I think it has something to do with my complete and utter lack of respect for anything structured in my life. Anyway, stalled on slider-bar (for no good reason), working on a button widget instead. 1.5 hours into it, almost done...I love it when I code stuff that doesn't suck.
I want something better than Outlook, and I've been thinking about it since about mid July.
Here's what I think a true PIM will look like (so far):
- A format-aware, personal CVS type repository. Accessable from anywhere via encrypted link (ssh, ssl, etc...), this data warehoue will act as your personal information repository. No, it doesn't have a display format, no it doesn't act as a "database" (queryable), it's a information diff engine. The format aware part is pretty important here. We only have so many kinds of data in this world, and XML is making them all accessible to one another, so XML is a prime canidate for a "baseline" format from which others can be constructed or deconstructed from.This thing might be a huge app, requiring a dedicated server/person.
- A manaement layer. I'm thinking something along the lines of an information "landscape" that you "place" data on (or the manager places it on the landscape for you). Major life priorities could be given fixed points on the landscape, and other things "arranged" in proximity to them and whatever else they happen to be "associated" with. People in your life could also have "points", and data about, from, and to these people could be located at these "heaps" on the landscape. People could then be "placed" on the landscape in proximity to the things in you life they affect (family, work, side interests, etc...).
- Display layer. Current PIMs currently are far too tightly tied to their databases to be useful in any meaningful way. Displays for the new form of PIM can look like current displays, but can also live on any device with a connection (remeber, you should be able to "check out" your data from anywhere).
- more stuff I've forgotten.
I think it can be built. I really do. It won't be easy, and it won't be quick, but I think it can be done...I just don't know if I'm the person for the job.