I have been actively involved for some years as a user and contributor to open software. Often I have picked up stuff which is useful to me and not known in my own discipline of chemical engineering. I have interacted freely with the message groups and newsgroups and learned to listen as well as to contribute.
In the middle of last year I needed to learn some new software (which will remain nameless here.) I purchased a copy and started to use it. It crashed, because two of its components were incompatible. It asked me to report the crash, so I did. I investigated on the web site and found that the crash was acknowledged in an earlier version of the product with the work around (!) to not do that action at all.
I took this up with them by Email and had an acknowledgement of the problem. I spoke to the U.K. distributors of the product who mainly dealt in books, so they didn't really know the technicalities but said they would look into it.
When a colleague went to a conference I wrote a report about it all for him to pass on. I have been able to characterise and reproduce the fault and have even written software to get around it.
The thing I want is a version of the product which does what is says it does. The main response I have got it to (a) amend the web site to say the problem has gone away when it hasn't and (b) replace the version of the product with a differently functioned one, for which they want more money.
I know what the code is doing because it has an external interface which I can interrogate and find the answers I need. They are one level of indirection down from the standard level which is why the interface fails.
I am being deliberately ignored and sidelined. They are happier if customers like me do not exist. For my part I would rather use someone else's software, but this happens to be something useful for me.
There, I feel better after all that. It also feels to me like a part of the discussion on the wikiWiki.
-- JohnFletcher
I don't think the Cathedral ignores the customer any more than the Bazaar. Instead, both methods run the risk of ignoring the user/customer if they are not valued and part of the development processs. Unfortunately, many companies use marketing more than they do customer support as a resource for mining requirements from. Most commercial software companies don't even bother to conduct interviews of existing and potential users during elaboration or transition. However, I do think the bazaar method runs the same risk. Software can end up being slanting to those who are taking part in programming the software instead of those using it. It might be better to say that many MarketingDriven? and TechnologyDriven? companies ignore the customer and would do better to be UserDriven?. Companies are generally more interestedin CrossingTheChasm than answering their user's needs. --RobertDiFalco
See also TheCathedralAndTheBazaar