Aleksandar Fabijanic

I'll leave this open for now if anyone care to carry the changes from develop over to devel (sorry that 2.0 did not happen, but we simply don't have resources...

It is a [fix](https://github.com/pocoproject/poco/commit/1b2bd22d349e3fb58912c35c3e6d1be16bed9d54#diff-a2d2d62278c17ea0cf59aacacea99676) for bug. doing this right is hard because different back ends have different ways of determining the length limit. fix has been tested with Oracle, SQLServer,...

@sil-mk something along those lines was done in #1403 . can you check and see whether/how it relates to your issue?

I currently have no time nor platform setup to do something here. @RangelReale can you help with this? @sil-mk if you'd like to see the fix in 1.8, please find...

You should fix it and send the fix as pull request to devel branch. Or wait until someone else does that.

I'm not following, `onIdle()` was called when there are no socket handlers - who's there to handle the idle notification? When there are registered handlers but no activity, there is...

In any case, `SocketReactor` is a `Runnable`, which means user can override the `run()` to do whatever necessary for their requirements - we can't be all things to all people.

From just skimming over the issue (I have no time to examine or try to reproduce right now), it sounds like you expect all the data you request to be...

Looking through the code, it seems there's no way to control it. Perhaps `Application::processOptions` could be made virtual and protected. I've done some work recently where I found the logic...

I guess OCSP does pretty much the same thing.