I cannot stress enough the importance of filing bug reports. The OpenAFS Gatekeepers and the rest of the OpenAFS community that contributes their time and energy to debugging the clients and servers and writing patches cannot fix problems that we do not know exist.

We all take great pride in the work we do. There is nothing that frustrates us more than seeing a question show up on the mailing list from a user that asks "does X work with Y?" for which the answer should absolutely be "yes it does" only to see a response show up later on that says "no it doesn't. we have been seeing Z for many months now and it still has not been fixed."

Folks. In all seriousness, we can't read your minds and we can't monitor your systems. If you have a problem and do not say anything it will never be fixed.

Of course, whether or not we can fix the problem based upon what we are told all depends on the issue and whether the description of the problem is something that we can reproduce or otherwise isolate so that the source code can be modified.

Even if you cannot isolate the problem yourself, it is worth filing a bug report because then at least we know that when working on the code we should be keeping our eyes open for issues that might produce the described symptom.

Of course, if you file a bug report we will pursue it and be pushy about obtaining the information necessary to fix it if it is not obvious. Remember, you are the one that can reproduce the problem not us. Therefore if you want the problem fixed you are going to have to find the time to isolate the conditions that result in the bug and collect the necessary debugging information. This can include fs trace logs, core images (UNIX) or minidumps (Windows), stack traces, network traces, or other details specific to the problem that is being encountered. The faster the problem can be isolated and the appropriate data collected, the faster a fix can be produced.

----

A quick comment on paid support. Those of us who make a living via paid support would appreciate it if you would pay for our services when we spending our time addressing a critical operational issue. Your user's time is worth money, your time is worth money and so is ours. Being paid for our time is what permits us to continue working on the development of OpenAFS. Those of us who are gatekeepers will work on the problem regardless, but we are happier doing so when we know that the beer we are drinking has been paid for.

Jeffrey Altman
Secure Endpoints Inc.

Attachment: smime.p7s
Description: S/MIME Cryptographic Signature

Reply via email to