Going about beta-testing a program... advice?

[Madison Kelly]

Thnk you for replying!

Andrew Hammond wrote:
> My advice would be to answer the question: what niche of the market are
> you aiming for? I'm not seeing any features that aren't already provided
> by other backup packages. To break into the enterprise market, your
> application would need some incredibly compelling feature that can't be
> replicated in 6 months by your competitors. And if you're aiming for
> Linux as a platform then you've pretty much ruled out the low end
> market. They're too cheap to hire the skills necessary to run linux.

   The reason the program was written was because I had spent some time 
trying to get other backup programs like Amanda and Bacula working. When 
they did work recovery of data required special knowledge of Linux. 
Third there wasn't a backup program I could find that used 
(single/multiple) USB/Firewire connected drives as it's storage medium. 
Finally I wanted to have a couple of features that I haven't seen before 
on any backup program, mainly the ability to search for files that are 
on media that is offline and identify which drive I need to bring online 
in order to recover the data once I found it.

   To this end the program we have written is used through a webpage 
front-end (a la CUPS/Webmin) so a user who doesn't know (or is 
intimidated by) Linux will be able to use it once it's setup from their 
desktop. This includes selecting what to backup, when the jobs are to 
run, search for backed up data and recover it from the media back to the 
system all from the web front-end. The only special knowledge needed is 
the initial install.

   The program also can backup multiple sources simultaneously to 
multiple destinations. The program handles all of the spanning when 
needed and subsequent jobs will look for the data and update it when 
found regardless of how or where the media is connected rather than 
blindly writing the data out to the first destination with enough space.

   So the two main features that I think set it apart is that it is easy 
to use, very flexible and it uses inexpensive drives (when compared to 
tapes like Ultrium). Obviously it is not for everyone but I expect some 
will find it useful.

> So you want an admin / MIS / CIO to gamble their reputation by
> recommending an unknown, unproven back-up system with unclear licensing?
> There's already a good selection of solutions which are well known and
> proved. A few thousand dollars a year for software licenses is nothing
> if it buys a solution. Saving a few thousand dollars in software
> licenses at the cost of risking business critical data (and your job) is
> insane.

   At this stage, no, definitely not. The post was to ask for advice on 
interesting beta-testers and how to manage they're feedback. That is 
also why I mentioned that the program is, at this point, at an alpha stage.

   The licensing thing is a pain in the ass and I have argued the best I 
could to get it as close to GPL as I can. That is how I got it to be 
free for home users and to make the code available to view. The problem 
at this point is that my boss sees what he has spent to develop this 
program and he worries about getting a return on his investment. Once he 
sees that he can infact see a return simply on support and development 
contracts I fully expect that he will GPL the program. He just needs to 
see it happen first.

> Bugzilla or you could use RT (which I prefer).

   Ah, RT! I forgot about that, I had looked into it before. Thanks for 
the reminder!

> I contribute to GPL projects because by enriching the commons, I enrich
> myself. What do I get out of contributing to your project?

   Well, though it isn't GPL there are numerous other split-level 
licensed programs that are popular in the Linux community (MySQL comes 
to mind). The community (and in turn you) will benefit by having a 
freely available, easy to use backup program for home users. Companies 
will be required to pay for the most recent version but they will now 
have one more backup option to chose from which may well help a Linux 
advocate to convince their company to migrate (as Suse OpenExchange 
currently does).

   Thank you again for your feedback!

Madison
--
The Toronto Linux Users Group.      Meetings: http://tlug.ss.org
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://tlug.ss.org/subscribe.shtml