After our recent news story on the WAN acceleration company that's come up with a software client, enabling mobile users and teleworkers to take advantage of technology normally only used between offices or data centres, reader Michelle Munson, who runs software house Aspera, wrote in to say that Orbital Data's version isn't the first, and anyway it's not all-software - it still needs a data centre appliance.

She adds: "Our company has an all-software transport (targeting bulk transfer applications) that has been on the market for the last two years, has a lower price point, runs today on all major operating systems (Linux, Solaris, Mac OS and Windows) and out-performs the Orbital appliances for high-speed file transfer over long distance, high bandwidth, and lossy networks, in every customer comparison brought to our attention."

Of course, she's right - but then so was our news story, and that's one of the challenges when it comes to understanding the fast-growing WAN acceleration business. Aspera's technology is indeed software-only and therefore inherently cheaper, but as Michelle says, it's aimed at long-distance bulk data transfers between two specific machines.

By contrast, although the schemes offered by the likes of Orbital Data, Expand Networks, Juniper (ex-Peribit) and Riverbed use many of the same techniques as Aspera's FASP, their aim is different. In particular they accelerate applications running over the WAN - and a big hole in that area has been that the high cost of hardware appliances has prevented the technology's deployment to individual remote users.

When you then add in yet another category of WAN optimisation, which is the web acceleration and load balancing represented by companies such as Zeus, Juniper (ex-Redline), Array, Crescendo Networks and F5, the potential for confusion is greater still. (One difference is that these are single-ended systems, which focus specifically on publishing web-based apps, whereas the double-ended technologies can accelerate other things too.)