I whole heartedly concur with Chris. You haven't a prayer. Any project is won or lost according to the scope of work document that is agreed upon before you start. Unfortunately for yo this will entail considerable research (of which, I hope we are all provide use bits for you), to determine what's feasible. Clients always want the world and by lunch on Friday would be nice.
You must learn a whole new way of doing things, and whole new IDE's. Your only way of looking competent (note that I did not say looking good), is to show that you have enough sense to nail the scope in excruciating detail. What pda's - just PalmOS V5+, list one or two models, list adders for each additional model or for a port to another OS. Buy your way into it, getting the client to buy items that you may not need in future projects, but will require solely for this client to meet their demands. Buy a package, such as what has been described in this thread, that has the wireless libraries and the database support built in. I use Codewarrior solely, but I couldn't do this app in four weeks as I have little wireless experience. So, buy a full featured IDE, buy a installation package (Installwise or Catapult, depending upon your need for high end installation features. Contract someone to help, or spit off a piece to farm that out. And even if you can't charge this to the client, make sure that they know these expenses were required due to their unreasonable demands. This can, at least allow you to charge a higher "expedited" rate. I own a business with 3 offices across the state and this is a common practice. If you want us to start earlier than 2 weeks ARO, you pay 133% for that time. This helps the buy vs. build decision considerably. After writing my first app, it took me about the time in your described schedule to learn and write just the conduit! Now I am not the caliber of most of the folks you'll find answering questions on these forums (and yes this is the best place to post), but I'm a bit smarter than the average bear. Even so, conduits are an odd and fussy critter - this would be a good piece to contract out. Good Luck and tell them what you'll deliver and when. Then the decision is theirs. Regards, Randyp When I > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:bounce-414647- > [EMAIL PROTECTED] On Behalf Of Chris Tutty > Sent: Friday, March 11, 2005 11:17 PM > To: Palm Developer Forum > Subject: Re: Newbie question > > From: "Fernando" <[EMAIL PROTECTED]> > > I need your opinion. I have just been assigned a task that requires: > > Executive summary of the comments below: > - the training time is too short by a month or two. > - the 'all platforms' requirement is impractical and will destroy > the project. > - the time estimate is in the right ballpark for delivery of the app > itself by an experienced Palm OS developer. Testing time is > additional and varies proportional to the number of devices > supported. > > My advice: > - hire a contractor with experience in Palm wireless connectivity > to deliver that portion of the app and to help you get up to > speed. Pay good money for a week or two with someone that > really knows what they're doing. Alternatively find someone > that can spend a day a week with you for the duration of the > project. > > In more detail: > > 1.- Front end to a Client's orders database on a Tungsten C > (snip) > > 6.- For Palm 4.0 OS or higher > > 7.- That works across all platforms > > > Eh? They either want it for the T/C or they want it for everything. > Don't let them throw in 6&7 as 'nice to have's. They will > increase the testing and debugging time enormously. You should > assume that all testing will need to be repeated for each combination > of manufacturer, OS version and connectivity mechanism. It's > easy to see how quickly that can push the project duration from > weeks to months. > > If it has to be done in four weeks specifiy delivery on the T/C > ONLY. > > > They have given me(40 hour work week) > > 1.- 1.5 weeks of ramping up(throughout the life of the project) > > 2.- 4 weeks for doing the app from start to end > > > On my past experience training developers in > Palm OS, people seem to take eight to twelve weeks to get up > to speed with Palm OS development. It's a new platform, a > new development IDE, a new debugging infrastructure and a > new SDK - a fair amount of detail to get familiar with. That's > for VB programmers and your C experience will make a > significant difference, but that 8-12 weeks was also to the point > where they could work on projects where I provided the complex > database and back-end stuff as code modules. You're going > to be getting into connectivity and syncing issues that aren't > trivial. > > Another issue to take into account is that testing becomes very > time-consuming when you're doing connectivity stuff across > devices from multiple manufacturers under multiple OS versions. > While four weeks is probably ok to write the app, testing could > easily be that much time again, and that's assuming that your client > will provide a full range of devices to test with. > > > My questions are: > > 1.- Is the amount of time given to me sound right? > > > No. The learning time is simply absurd. Ask what it's based on. > > The biggest problem is that there are a whole stack of simple things > that can stop you dead for hours. For a tight deadline you need > access to an experienced developer to get you quickly past the > introductory gotcha's and then provide on-call asistance for > those times when you hit a brick wall. > > > 2.- Are there any other Forums besides the one in Palm that I > > should look at. > This forum is your best tool. I'm continually amazed by the > speed and accuracy of the information provided by this > development community. On the other hand they can't look > over your shoulder while you code. > > > 3.- Are there any books that you might recommend > > > No idea. The SDK tutorial and sample code is a good start. > > > 4.- What tools should I use for development > > > CodeWarrior is a mature product that most Palm OS developers > are familiar with. If you're under timeframe pressure I'd stick > with it. > > > The things I am concerned on implmenting > > 1.- Synchronizing of the Data wirelessly and via the PC > See below. > > > 2.- The wizard installation > I've worked for companies that have invested months in this, > particularly if the wizard will include registration and setup > of the wireless connectivity. > > > 3.- The "works on all platform" requirement > This will crucify the project. It's not even practical. Ask for > a list of supported devices. More practically, refuse to deliver > on any device that isn't provided for testing and estimate > a week for testing, installation tweaking and debugging for > each separate device. Combining this with the 'wireless' > requirement is particularly nasty as that's the sort of stuff > that is more likely to vary from one manufacturer to another, > produces difficult to debug faults and often requires debugging > on the hardware rather than using a simulator. > > Chris Tutty > > -- > For information on using the Palm Developer Forums, or to unsubscribe, > please see http://www.palmos.com/dev/support/forums/ -- For information on using the Palm Developer Forums, or to unsubscribe, please see http://www.palmos.com/dev/support/forums/
