I've gave PCALE a very good try. As implemented it suffers from several problems:
1. It is equipment specific and intensive. You either need an SGC tuner set up for bypass-on-receive (the only brand I am aware of that has this capability) or a special antenna that is resonant and efficient on each band you plan to scan. You can also set up RF switching to bypass the tuner on receive but that becomes even more complex. There was a computer controlled tuner on the market that could be controlled by MARS-ALE but MARS-ALE is not available to mere mortals and the tuner itself was buggy and is now out of production. 2. The link margins necessary for the calling waveform are pretty substantial. Those used to the relatively robust nature of RSID or any of the other common digital modes will be sorely disappointed. Even Winmor, while better than ALE, requires substantially better conditions for success. 3. The software itself is relatively complex to setup and operate. I'm sure Andy will argue to the contrary :-) However IMHO it's significantly more involved than just firing up Fldigi and banging away at some Olivia or PSK. 4. The widely shared nature of the ham bands makes collisions inevitable given the automation inherent in ALE (automation that is the whole point, in fact) and the limitations of even the best busy channel detection algorithm. This issue tends to generate a lot of hate and discontent. However this ought to be the least worrisome issue. With an appropriate band plan (which already exists for PCALE) the carnage can be limited to just the ALE calling channels and anyone who wants to use ALE should be expected to sign up for a certain amount of interference and not be whining about it as long as it stays on the calling freq's. In lieu of full-blown ALE consider the following idea: I'm no software engineer and beggars can't be choosers, so forgive me for making the following related suggestion (Patrick already laid into me on this once!) Consider that RSID is great for identifying the mode and that Call ID is great for identifying who is calling. Both use signaling standards and waveforms that are very simple and robust. But what is missing is an equivalent SELCAL (selective calling) signaling standard using waveforms and formats similar to RSID and Call ID. Imagine you wanted to find somebody monitoring the 3KHz of USB spectrum at 14070KHz dial freq. You could find a clear spot in the waterfall and transmit the SELCAL which contains the call sign of the station you wish to reach. At the receiving station the SELCAL enabled software would function in the same manner as that currently done for RSID, i.e. detect the call, display/sound a notification and provide automation for tuning and answering under operator control. Once an effective, simple and robust SELCAL standard is developed (again IMHO it should be a logical extension of the existing RSID and Call ID standards) it could eventually be parlayed into a more modern and effective variant of ALE. By using time synchronized band scanning and transmission (similar to WSPR et al) probability of intercept can be substantially improved. Neither the SELCAL or time synchronization represent new technology and both derive from proven, similar implementations. So if one were to make a SELCAL on 80M, for example, once the spot on the waterfall was chosen by the operator (because we can't rely on unreliable busy-channel detection technology) the SELCAL transmission would occur at say for instance 10 seconds past the minute. Synchronized scanning would put all stations on 80M at 10-15 seconds past the minute, 40M at 15-20 seconds, and so on. The last piece would be to perfect busy channel detection and automate the selection of empty places on the waterfall, but this part of the puzzle is useless with SELCAL (very useful by itself) and synchronized scanning/transmission. And once this last part was perfected we are back to requiring special tuner/antenna solutions.
