4 ms·
I've thought a bit about this, churning around various product ideas. In general I'd say that if it's a typical SaaS, I would probably only trust it with my dat
by e12e 13y ago
I've thought a bit about this, churning around various product ideas. In general I'd say that if it's a typical SaaS, I would probably only trust it with my data, my time (learning curve, adapting workflow etc) if it was a available under a licence that would allow me to self-host it if it was discontinued (or pay someone else to host it for me). This is for stuff like mail, images and other DAM (Digital Asset Management) systems, chat/im, bug trackers/knowledgestores and databases etc.
Note, that I don't see any reason to trust Google either. Remember wave? Remember XMPP compatible gtalk?
My conclusion has been that if/when I finally get started on my first product, I'll probably license it under AGPL, and make sure there's a realistic self-host story for users -- and then charge for the SaaS/PaaS-bit, but probably not for the client (or server side software under AGPL). Clearly it's a little more challenging -- you might end up with people competing with you for hosting -- but I can't see any other way to build customer trust. Nothing any company would say to me as a customer would make me trust them -- how could I ask anything else from my potential users?
Since people still pay for hosted email (even if many, many do run off to freemium services) -- I'd say the business model is viable -- those that can afford to pay, and know that value can be added by proper hosting -- and not having to do everything themselves -- will consider becoming a customer. And I strongly suspect those kind of customers are the best kind of customers. Especially for a start-up.
Even though I wasn't a user, the most dreadful death-by-buyout I remember is DabbleDB[1] -- compounded by the fact that the founders went to twitter -- presumably helping with the biggest non-product of our decade -- rather than allowing dabbledb to live on in any form.
Also note that having a published API/open protocol/standard isn't
enough. For example, the IMAP standard clearly states that message ids
aren't necessarily in chronological order. Some IMAP servers make sure
that they are and/or that messages are returned in a list ordered by
(message) time -- and some will return the message that was latest
modified first (eg: a messaged moved into a folder). Some IMAP clients
get this right (sort as the user wanted, regardless of server
implementation) -- some get it wrong. A small difference (arguably
client bug) that can make or break certain use cases and processes for
dealing with email.
For a b2b setting -- one could reasonably offer source under a different license -- if AGPL seems too open for enforcing whatever business model one has decided on. But I suspect that AGPL would work fine in b2b settings as well. The key aspect is to offer source code and documentation in a form that offers a viable self-help scenario in the case of a bankruptcy or "death-by-acquisition".
[1] http://www.youtube.com/watch?v=MCVj5RZOqwY http://www.youtube.com/watch?v=MCVj5RZOqwY