So, in no particular order… Things I would really like to spend some time hacking on, but I have pretty much zero free time at the moment between moving, marrying, releasing at work, etc.
Asbestos
Yeah, good old Asbestos. It’s been…a long time since I’ve had a chance to commit anything. There was a period where there was some Ubuntu user who was all excited about things and looking to send me some patches, but that fell through. Add to that the fact that Network Manager has pretty decent vpnc support, and I’ve really had little need to bang on Asbestos. That being said, I still think it’s actually a pretty handy-dandy little tool, and has its appropriate niche it can fill.
I’d like to:
- Get Network Manager integration ala vpnc; I really wanted to do this before I’d seen how nicely it could work, so I definitely want it now.
- Update to the latest dbus. Got, could that stuff hit api stability sometime, please? I really don’t have time to chase the changes, but I want to use it.
- Link in appropriate libraries to do ssh in process rather than spawning it; by doing this, I could more easily do some fancier things such as…
- Allow live modifications to Asbestos tunnels; it’s a real PITA to have to drop the whole tunnel in order to add a new thing to it.
I’m sure there’re more things (hell, if I dug out the source, I probably have a TODO file with a bunch of these things already), but those are the big ones off the top of my head.
Scanning Support
The state of things in the linux world with scanning is pretty hokey. I’ve seen people complaining, but as far as I’ve seen, nobody’s really stepping up and doing what I want. I want proper HAL integration for scanners. I want a scanning frontend that doesn’t terrify me (really, XSane, I don’t need all that stuff you throw at me. I really don’t care about being able to scan at 297 dpi instead of 300, etc.)
I’ve got notes about what I’d like to do, and probably if I go poking around, I have the start of some code to wrap the sane api into a glibby thing. The sane api is really a bit of a pain; there’s only about a dozen calls to it, because everything’s a matter of “oh, query the scanner to see what thousand options it has, now do some string matching to figure out if any of those are interesting, and make sure to store the indices to them, since you’re going to need those to poke at the values”. Yes, it’s very flexible. Yes it means that the same API can be used for a zillion different things. But really? At this point, I care about a simple API to talk to a flat bed scanner. I really don’t care about hand scanners, treating a web cam as a scanner, etc.
I really want to start banging on this stuff, because it’s an important problem space, and could do a lot of good for people. I also want to get my foot in the door before I languish on it long enough and somebody else beats me to the punch. I want some celebrity, damnit!
Synergy
I’ve mentioned in the past how I use x2x to glue my laptop and desktop together. The other day I was pointed to a more fancy-pants solution to the problem (which I’ve probably also talked about wanting to implement myself. So with Synergy, you run a daemon, have clients connect to it, blahblahblah.
It seems to work pretty well at first blush (at work I now have my mac glued to the left of my workstation glued to the left of my laptop), so I’m fairly excited. It needs some usability love, though; you need to write a config file describing all the screens and their relationship to each other (and it’s not reflexive; I have to say that my laptop is immediately to the right of my workstation AND that my workstation is immediately to the left of my laptop), etc.
I’d potentially like to clean that up some (or who knows, maybe that work’s already in progress; I haven’t honestly poked at the project too much at this point. See also: my total lack of time), but what I’m even more interested in doing is sorting out mDNS and making a clever little thing so that you can look for synergy servers, register with them, etc.
In my mind, I see something like this:
- I fire up my laptop at work, and NetworkManager brings up the network
- My synergy lurker catches the dbus signal that the network is up, and checks with mDNS to see if there are any synergy servers on the network
- My workstation goes “Oh look at me! I am!”
- My laptop and workstation have already been paired, so it does some quicky authentication and poof, I’m glued together
As for the pairing bit, I see it as something sort of akin to bluetooth devices; I can initally get a list of open synergy servers, tell them to pair, either end wants soms pin to prove that I’m kosher, etc. After that, they’ve got some sort of stored state between them, an auth key, or whatever, so my laptop can always go “Hey, you know me, I’m that laptop that sits to the right of you”, and it can all work, even if I happened to dhcp to something different.
I could see it being very slick, and it seems like a reasonable excuse for me to figure out how to work with the mDNS stuff; that always seemed cool, but I hadn’t come up with a practical demo to play with until now.
Set Up LDAP
For years now, I’ve been wanting to finally bite the bullet, set up an LDAP server, and start stashing things in there. I’ve got a handful of boxes I’d like to run with shared authentication, I’d like to do server-side address books, etc. Yes, I could cobble together something stupid for myself, but I really want to do it right, otherwise I won’t have time to maintain it. So LDAP it is.
Eventually.
Set Up Bugzilla
I actually did this once upon a time; I set up a bugzilla server for my own person stuff, so I could do things like file a bug saying “Fix the toilet” or “But toilet paper” or some such. It’s really just me on my personal quest to find a way to manage my todo lists, etc. I just figure that maybe it’s worth trying it again, especially since I’m building up a rather large mental list of things to do now that we bought the house. (Building up and….forgetting. Crap.)
Win The Lottery
Seriously, man. There’s so much FOSS stuff I’d love to just spend all my time on, but I can’t right now. Looking at the current jackpot, it’s $25 million. If you divided that by four (losing half of it to get as a lump sum, and then losing another half to taxes), that leaves ~$6 million. If you shoved that into something giving 4.5% annual interest without touching the main lump, you’d sill be making ~$280k a year on interest. I could find a way to scrape while making over a quarter million a year with 0 hours work.
And then I really would spend time doing FOSS hacking. I could really just make it my day job to work on some of this stuff. I know that sort of risks one of the big problems in FOSS; since there’s no monetary incentive, I would be at risk of abandoning things in search of the next shiny thing that caught my interest. But at the same time, I’d actually be able to invest the time and effort into things to get them into a much better state before I did that anyways; a couple weeks of serious time spent on things is more than the same time spent scraped out of free hours on the evenings and weekends.