Bertjan has already done a very nice job of giving an overview of where we are with this year’s GSOC project and where KPilot is in general. I wanted to add a couple of things to this train of thought too…
First off, this is the second year that I’ve worked with Bertjan on a Google Summer of Code project and I have thoroughly enjoyed both years. Last year, our goal was to create a base conduit framework that could be used to unify KPilot’s conduit code and push down into it all of the synchronization logic that was previously spread across all of the individual conduits. This was a success, and Bertjan also wrote a new Keyring conduit that implemented the new base conduit infrastructure. Unfortunately, these were dark days for KPilot and neither Bertjan nor I nor Adriaan had any more time to give to our little KPilot friend. KPilot was semi-ported to Qt4/KDE4, but none of our KDE PIM conduits worked. And so KPilot languished pretty much untouched and unloved between last year’s GSOC project and this.
This year, Bertjan has built upon last year’s work with our base conduit core code. He’s ported the contacts, addressbook, and todo conduits to use both the new base conduit infrastructure and the new Akonadi back ends that KDE PIM 4.2 will be using. As a satisfying proof of our design work from last year, I don’t believe that we’ve hit any synch algorithm problems, yet. In addition, Bertjan has done the necessary work to finally have 2-way category synchronization working with KPilot and our conduits, YAY!
As Bertjan said, there’s some rough edges still, 2 weeks away from the conclusion of this year’s SOC project, but we’re hopeful that we’ll be able to iron KPilot’s side of these out before 4.2 is released and also hopeful that our KDE PIM apps will be ported to use Akonadi’s back-end so we can again synchronize between our PIM apps and Palm devices.
Now, there’s a few things I’d like to really ask for help with, so if there is still a KDE Palm community as I have seen glimpses of before, now is the time for you to speak up and help us make KPilot a first-class citizen in the KDE4 world. To put it bluntly, we need the KPilot community to help or we won’t make it.
- We need help in testing with devices. Qt4 has changed the way that process threading works and I think I’ve done it right, but have only been able to do very limited testing with a very limited set of devices. I know our configuration/device-detecting wizard is broken and has been for quite some time, so it would be really great if people could help test and send fixes in for this.
- We need help testing our new conduit code. Bertjan has written unit tests (awesome!) and we will be doing ad-hoc testing, but obviously we’ll miss something unless we have the broader KPilot community’s help. If you have time to spare and an old device to test with, please, please help us test the new ToDo, Contacts, and Calendar conduits. The Memofile conduit hasn’t had to change that much, so I’m pretty comfortable with it, but the others are major rewrites.
- We have an awesome little Keyring conduit that we simply don’t have the time to fully test, polish, and support for KDE 4.2, which makes us very sad. It does seem to work, mostly, and Bertjan has even written an almost completely functional Keyring viewer/editor in our KPilot code! But we just don’t have the time and manpower to make it as bullet-proof as we’d like before KDE 4.2 is released. But this is good news! If you use Keyring and would like to help us get this finished, polished, tested, and ready for prime time, we’d LOVE your help! Please let me know!! Unfortunately, if we don’t get the help we need with this, we’ll be forced to disable the Keyring conduit and nobody will know it’s even there. =:(
In addition, let me just add that while Palm devices were once the coolest geek device on the planet (after the Newton went buh-bye, of course), and lots of people were interested in using them and subsequently a higher number of people were interested in coding for and helping out with them, this is definitely not a true statement today. Everybody and their dog is all excited about the iPhone, and I cannot say that I blame them. Palm seems to have totally lost its coolness, geek appeal, good business sense, and anything else that would keep it otherwise relevant in the current technology climate. I won’t digress too much here, but if you’ve ever tried to endure through Opera’s constant crashing of your Treo or Blazer’s general suckiness for a couple of hours while everyone else around you is experiencing pain-free Safari-surfing on their iPhones, you might be able to appreciate the frustration and rage that accompanies said experience.
But back to subject at hand…. The affect that this all has, though, on those of us who are trying to maintain software for the Palm platform (KPilot, here), is that it’s pretty de-motivating when nobody seems to care about the software you’re writing anymore. Synchronization and PIM software is about as un-sexy as you’d want to get, quite honestly, and KPilot seems to be the ugly poster child for it.
So, reiterating my cries for help, if you use and care about KPilot now and in the future, please please please let those of us know who still try to care for the little guy. And if you truly care and have even small amounts of time (and maybe a spare couple of devices to test with), please let us know that too. I’m not asking for people to send me or Bertjan or Adriaan devices. We have some already. What we don’t have is the time and breadth of use cases to adequately test KPilot and make sure that it’s not going to eat your data for lunch, and that’s something we’d really like to not see happen after KPilot is released with KDE 4.2
Thanks for making it through this lengthy post and as a reward, here’s one of Google’s image search results for KPilot (????) =:)