Hello everyone! (sorry for length of this post).
This thread was pointed out to me privately, so I have registered to address this issue, which is a cause of concern to both ourselves and ROOL - we certainly don't want anyone to feel disappointed. We're all massive RISC OS fans, and our hope is that as many people as possible will come to enjoy RISC OS on the Raspberry Pi, and the NutPi project is most definately a hugely important part of this. I'm in discussions with ROOL as to how to improve the NutPi web pages, and hopefully that will bear fruit over the weekend.
Firstly, to address the specific concern about Messenger Pi. If anyone feels unduly inconvenienced by the IMAP cut, please contact us privately, and I'll try and resolve this - the IMAP component is actually a standalone part, which I can supply to those who need it desparately.
When developing for a project like NutPi, we were approached by ROOL to deliver software to add as much value to the NutPi but within budget constrained by the NutPi's RRP. Clearly the goal was to deliver as much value as possible, with as much breadth as possible, so we had to make decisions as how to best achieve this. In Messenger's case, the only "transport" normally included in the full price was IMAP (this has now changed in v7), with the more common POP/SMTP being part of NetFetch.
This is a historical peculiarity that stems from the modem/dialup era, where people used separate "Internet Suite" products which managed connection and fetching. In order to offer POP/SMTP as part of NutPi, we therefore had to include our NetFetch as well. Since that's a whole lot of expensive software, we felt that a fair compromise was to cut IMAP from Messenger Pi since most people can use POP+leave-on-server to achieve similar results.
In Messenger Pro 7 (which has been released since NutPi was created), we took further steps to address this for new users, as it had clearly become a problem. MP7 includes portions of NetFetch as part of the package, to ensure that POP/SMTP/IMAP are all supported "out of the box". However, please remember that Mpro7 alone costs more than NutPi retailed for!
Anyway, I'd like to re-iterate my offer to help anyone who feels that they can't get by without the cut feature(s). It certainly isn't our intention that users be inconvenienced by this - when creating software to a budget, it is very difficult to gauge which features will be needed. I'm quite happy to produce a future update for the NutPi version based on feedback on this issue.
It is my belief that Pi has a huge "evangelical" role to play in spreading the RISC OS word to new and returning users (especially the latter). NutPi is, in many ways, the "ace in the hole", as it gives quality software for the same price as a meal-out.
However, one does need to stop and think for a moment - if most new sales are likely to be Pi-based, what incentive is there for anyone to buy the more expensive editions of software? Are we essentially saying RISC OS software is only worth 5-10pounds an app (or less) - that's less than a Happy Meal!
This is a very difficult conundrum, and one we're all wrestling with. Our aim with NutPi was to give as much as possible/sensible in the NutPi suite, whilst still keeping some value in the full copies. We need that to stay in business, unles NutPi is hugely successful (I have no clue on sales figures yet). This is something that we're revisiting and rethinking all the time. As one of the few "full time, with paid developers" companies within the RISC OS world, it is something which is accutely relevant to myself!
I'm certainly not saying we've made the perfect decisions, and we'd certainly welcome feedback. So far, we haven't seen any upgrades (to my knowledge) from NutPi, which either means the versions on NutPi are meeting users' needs, or that people are so upset that they're not interested at all. I tried to price upgrades such that it was cheaper to buy NutPi and upgrade a couple of apps, than buy them full-price. That seemed a fair compromise (ie. it makes NutPi extremely desirable even if you plan to buy lots of RISC OS software separately). The way I saw it, people could buy NutPi and receive a lot of quality software at a low price, or buy NutPi and upgrade to get latest/best editions, and *still* be better off!
One final thought - for existing RISC OS users with existing collections of R-Comp software, it'll still probably be cheaper to get upgrades directly from us, rather than going NutPi (although there's lots of great non-R-Comp software on NutPi which more than makes up for this). That's the result of low-cost upgrade policies (which is very different to the PC world where companies like Adobe charge huge fees for yearly updates). We tend to do major version updates (usually priced 20ukp or less) every 3 years or so, with free updates in between.
I know this is bad form to say, but please note that I don't tend to follow forums/newsgroups as I'm usually too busy with email/work/development etc. However, I'd be more than happy to discuss things with anyone on our usual R-Comp email, and once again, my goal is that NutPi users be thrilled with all things RISC OS!