As long as session cookies are meant, the behavior looks correct to me. If not it is indeed a bug. (But I am a cookie hater: best (intention) and worst (mis-use) idea ever invented. Servers should maintain state, not clients)
So it is a bug indeed.
It looks like your bugfix did not solve the bug.
[...]
Maybe you should open the bug report again.
.(My fpc contains your bugfix).It was already applied, do not apply it yourself. 3.2.1 and 3.3.1 both have the patch applied. re-patching it may break your build since many more related things in TfpHttpClient are also fixed.
engkin has not logged in since April last year. I think you'd do better opening a new bug report and referencing the earlier one.I will, but I wanted be sure there is still bug. I thought , maybe I am making something wrong.
No Thaddy, I did not applied his patch, I downloaded version of 3.2.1 and compiled it. It already contains the patch. Maybe I said it wrong..(My fpc contains your bugfix).It was already applied, do not apply it yourself. 3.2.1 and 3.3.1 both have the patch applied. re-patching it may break your build since many more related things in TfpHttpClient are also fixed.
see https://bugs.freepascal.org/view.php?id=33451 which is back ported to 3.2.1 I believe.
With proper settings only the session cookie is dropped (as it should) I am using daily builds for 3.2.1 and 3.3.1 and so should you. I tested with trunk from yesterday, but not very intensive.
It looks like your bugfix did not solve the bug.
TFPHTTPClient still holds only last cookie in fpc 3.2.1-Svn 62267 and latest svn version.(My fpc contains your bugfix).
I uploaded test project with server and httpclient app. Chrome-Browser holds cookie until it is closed, but TFPHTTPClient loses it after second call.
Maybe you should open the bug report again.