-
Notifications
You must be signed in to change notification settings - Fork 36.9k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
bitcoin: links reported still broken #1944
Comments
Upgraded -> Started bitcoin -> Tried donation links which failed -> Quit bitcoin -> Found multiple donation link processes and force closed them -> Tried starting bitcoin again, but would get no gui or unresponsive gui -> Tried your suggestion and got same result -> Reboot machine -> bitcoin works fine I'm not sure if it's a code bug this time, perhaps it takes too long to pass an URI to the message queue, which makes users impatient. And they click again and again, which can screw things up (which IS not good). |
Maybe do a check at startup so the client don't start more instances if it's already running? |
There seem to be way too many problems with boost::interprocess... how do other programs deal with this? Sending a message to a running program should be something trivial, in my opinion. Maybe it would require platform-specific code for almost any platform, but that isn't worse than what we have now. |
Qt supports named pipes on any platform that we care about, imo that's the Platform specific solutions are more work and probably require hacking |
@burger2 There is a check at startup, which just tries to send new URIs to the message queue used for handling URIs. If that succeeds skip further client startup and exit, if no message queue is available start a new client. To me it seems that check doesn't work reliable, when impatient users are clicking to fast on a bitcoin: URI (which is indeed no fault and should not crash the client though). |
I wasn't impatiently clicking on the link multiple times in rapid succession. I clicked it twice, the second time only after a few minutes. Rebooting the computer did seem to resolve the issue though, so it might have just been a specific issue to my machine. As I suggested, maybe rebooting a windows machine after upgrading should be recommended. |
I was not trying to say you are that kind of impatient user, I just wanted to clarify this can trigger a problem :). |
@gavinandresen As your paymentserver pull was merged, this can be closed IMO. |
… to their own targets c0c96dc [Tests] Split out zerocoin and sapling unit tests to their own targets (Fuzzbawls) Pull request description: with this, there are now 3 distinct targets for unit tests: - `check-standard`: Standard unit tests and util script tests - `check-zerocoin`: Legacy unit tests for zerocoin related functions - `check-sapling` : New sapling unit tests Running `make check` from the repository root will run the 3 targets in reverse order from the above to prioritize the longer tests being run first. ACKs for top commit: random-zebra: great. ACK c0c96dc furszy: cool, ACK c0c96dc and merging Tree-SHA512: d54676e7632bd6780598cba70fba499547c57f03cbe13760190a01abddb86d70064bc1537d87041f5c573d87d281a354679d4795818435977d8515461a81f87b
c80cba7 Revert "hack to prevent travis no outputs timeouts" (Fuzzbawls) 875a2d7 [Travis] Set stricter timeout for Dash 32bit job (Fuzzbawls) Pull request description: The unit tests for this job take upwards of 35 minutes due to 32-bit performance restrictions, so set a stricter build timeout. A build timeout of 600 seconds (10 minutes) was chosen as it is roughly double what a (currently) fully ccache'd build requires, so some leeway is left over to avoid unnecessary job restarts when a PR's changeset has a minimal impact on the build time. Additionally, revert the workaround "hack" to continuously output `.`'s to the console when running the unit tests as bitcoin#1944's reordering to run longer tests towards the beginning of the process has made it unnecessary. This also restores the easier to read output formatting. ACKs for top commit: random-zebra: ACK c80cba7 and merging... Tree-SHA512: 8c286660072064166573d017e4cde7e55830328d2badc377ecaafc33207e6dc7b11fcb63e0d1ec52e06c71510c998bdc4762a28bd50bda31712d20f2e87ce78b
LightRider on the forums reports:
"...clicking on the link multiple times creates multiple instances of the bitcoin process, but the user cannot see any affects of them starting. I didn't notice until I checked process explorer and noticed them running. Also, when the client is not running, clicking the link creates a process but no gui appears at all, neither the window nor the tray icon."
https://bitcointalk.org/index.php?topic=119277.msg1288375
The text was updated successfully, but these errors were encountered: