Quesar Report post Posted April 30, 2011 (edited) Hi devs, There is a bug in console mode where old chat can show up in the bottom. With the resolution I run (1280x680) I normally have 3 blank lines in the bottom of the console. After the client has been running for a while, and the entire chat buffer gets filled and starts to loop around, I get old chat from the beginning of the buffer showing up on the bottom 3 lines. I've had it happen on another system in a different resolution (sorry can't recall details) where it had 2 blank lines instead of 3. This is what it looks like on my machine atm: As you can see from the mini-map, I'm clearly in neither PL sto nor C1 portals, lol. If I #cls, the problem goes away, but will come back if I leave the client running long enough again. Edited April 30, 2011 by Quesar Share this post Link to post Share on other sites
Entropy Report post Posted May 9, 2011 It used to crash when the buffer got full Share this post Link to post Share on other sites
bluap Report post Posted December 13, 2011 Fixed now. I rewrote the code to be a lot simpler. Instead of using a circular buffer, the oldest few lines are now deleted and the remainder moved up the buffer. I set it to delete 100 of the 5000 lines. Potentially more work (each time 100 lines are deleted) but still insignificant compared to everything else. Share this post Link to post Share on other sites
Cycloonx Report post Posted December 15, 2011 Fixed now. I rewrote the code to be a lot simpler. Instead of using a circular buffer, the oldest few lines are now deleted and the remainder moved up the buffer. I set it to delete 100 of the 5000 lines. Potentially more work (each time 100 lines are deleted) but still insignificant compared to everything else. A circular buffer might be "more complex", but the way you implemented it is rather slow and not recommended at all. Although, I appreciate your work ofcourse Share this post Link to post Share on other sites
bluap Report post Posted December 16, 2011 (edited) "rather slow" is not a useful measure nor perhaps is my "insignificant" assessment. However, I timed the new code and, even on my ancient laptop (1.6MHz Intel mobile) which is too slow to play EL, it took less than a millisecond to execute. As I said, I'm fully aware that the new code will be slower than the old, its just insignificantly slower. Also, it will only run once in 100 new lines for an already full buffer. I spent a significant amount of time attempting to find the real bug and no one else came forward with a fix in all the time since this was reported. Perhaps those that don't "recommend" this method could have a ago. I appreciate your comments of course Edited December 16, 2011 by bluap Share this post Link to post Share on other sites
Grum Report post Posted December 16, 2011 I haven't looked at the new code, but as the author of the original circular buffer, I'm glad to see it's gone The thing has always been more trouble than it was worth, and, as bluap found, too complicated to easily understand. I still think the idea was nice, but the code as it stood was unmaintainable. Share this post Link to post Share on other sites
Cycloonx Report post Posted December 18, 2011 I still think the idea was nice, but the code as it stood was unmaintainable. Ofcourse, bad written code is a pain. Share this post Link to post Share on other sites