Hi Hugo,
Yes, indeed, quite often the recommendation is to create a support case and then neither Support nor the thread starter informs here if the problem was solved.
Maybe Support will comment this additionally, but once they already answered that quite often the problem appears to be specific for the given tested applicationand that they request confidential information for the problem investigation.
From my experience, I can confirm that several customers I worked for had a serious problems with their applications tested with TestComplete and that we provided SmartBear with the confidential information to help them to solve our problem.
For example, one Delphi application with devExpress components started to crash with the new version of TestComplete. We failed to reproduce the problem with simple samples, so the customer provided SmartBear (AutomatedQA at that time) with a copy of their real application. Then we received one or two more patches and finally the problem was solved and did not happen with future versions of TestComplete.
Considering this case:
-- I don't know what was corrected in the final patch, how well it was tested and thus how safe it was to made it public for those who did not have the problem like we had;
-- Obviously, the information that was provided to Support while working on the problem was confidential without any chance to be exchanged via forum;
-- Considering, that the problem never happened with the future TestComplete versions, I assume that the patch was somehow applied to the main code branch, so no patch, neither private nor public was needed anymore.
Finally, it is my opinion and my guess, that it is much more safe from the support point of view to release patches privately only for those who really need them than to make them public. Indeed, without a strict automated patch management and install subsystem (like the one from MS), there is a high chance that customers will apply all released patches without considering whether they really need this or that patch or not. Frankly speaking, when you apply a patch from MS that 'resolves security issues', do you have any idea of what is corrected and do you really need this? But this is MS and it is probably better for them to implement and support Windows Update than to deal with reported crash dumps every one of which may contain otherwise a unique combination of system components installed on the given client's box.
Considering that people when reporting their problem on the forum quite often do not mention even the major version of TestComplete, I think that it would been a real nightmare to investigate a reported problem if a user can have a unique set of TestComplete modules (consider efforts to just create a test environment to try to reproduce the reported problem).
On the contrary, when once MS released a .Net update that made all .Net applications closed for TestComplete, the patch was released very quickly and made public through the forum as well.
Summarizing, nobody is perfect :), but personally I see some reason of why patches are not made public but are distributed only to those who explicitly requested them.
P.S. Considering the exact problem stated as the thread subject, I did not meet it for a very long time (several years, I think I can say), so unfortunately, I have very little idea of what might cause it with the current TestComplete versions. Thus I just believe that Support will comment the current state of the problem considering their internal issue database (e.g. how often the problem was reported for the past month or two, for what configurations, etc.).
