sfrrr - The superstores are governed by different rules, since in most cases the publisher sets a wholesale price and offers only a suggested resale price - retailers have an incentive to keep prices low in order to stay competitive with other retailers, and hence they're motivated to keep their markups fairly reasonable. Of course the margins do vary, they're quite a bit higher for software than for hardware since software sales are almost pure profit, and there are plenty of shady dealings involving large sums of cash being paid for prominent placement in catalogs or even in brick-and-mortar stores, so even with non-PDA software, if you buy it directly from the developer they'll make a lot more money than if you by it in a store.
It's been suggested that Palm/PocketGear + Handango would work a lot more fairly if they adopted the same model as other stores (developer sets a wholesale price and PG+HD compete with each other to offer the lowest markup), but since those two companies have cornered the market and their contracts guarantee that they won't have to worry about other stores undercutting them on price, there's not really any incentive for them to give up their guaranteed 40%. And for some reason there don't seem to be any pay-per-listing or user subscription fee catalogs like Download.com/Tucows/VersionTracker for PDAs (those sites actually do offer PDA listings, but few developers and fewer customers use them).
haraldalbrecht - agree with you completely about the declining quality of software on PG, though this trend has been going on for quite a while. A big part of the problem is that there just aren't that many different types of applications that people are interested in running on their PDA. So whenever a new product idea comes along there's a sort-of land rush and you end up with a zillion nearly identical products doing the same thing; there must be 50 different Sudoku games now, for example, and a similar number of programs to organize your movie collection. The difficulty of programming PDAs further exacerbates the problem, since a lot developers stick to easy-to-code apps like Bejeweled knockoffs or the aforementioned movie databases. The reason the Pocket PC game scene is so much richer than the Palm one is because it's actually possible for a novice programmer to write an original and graphically-interesting game on Pocket PC; on Palm, between memory restrictions, processor speed limits and the lack of built-in OS support for a lot of game-design necessities, anything beyond the level of a basic puzzle game can be extremely difficult to get working effectively. And similarly, the Palm OS "hack" scene was extremely lively back in the pre-OS5 Hackmaster days because you could completely replace any system function with very little programming effort.
Stacking the listings with multiple flavors of the same product is a long-established tradition; Beiks/Ectaco et al can get themselves weeks of free front page advertising simply by updating one or two products after they release a bug-fix for their software engine. During the few months we were lising Berlitz on PG (a desperate attempt to drum up sales - didn't help much, and this was back when they were only charging 30%) we even tried this ourselves, but stopped after a few days when we realized it made us look like complete idiots.
PG is indeed on the decline, they're fighting with Handango and the increasingly-independent large developers for a bigger piece of a declining market and they're not winning. Handango is trying to build up a business selling software for mobile phones, but there's no way the cellular carriers will let them keep making money off of that - there's a reason why they're so eager to lock out people from Bluetooth file transfers and the like. (so you can see why we're in no hurry to support phones that aren't Palm or WinCE-based) And once handtops like the
OQO get in the sub-$1000 range (they should be pretty close by next Christmas) the power users who buy most of the PDA software will start jumping ship to those and the picture will get even bleaker. Which is one of the reasons we're eagerly forging ahead with an x86 version of PlecoDict - it's not just for PCs anymore.