Posts: 126
Threads: 42
Joined: Apr 2022
Reputation:
12
First bug of V4 ???
example:
Do
Print ".";
c = c + 1
If c Mod 10 = 0 Then Play "a64g64"
Loop
Play volume is not default 50% and not changeable with the Q## "play" variable. Tested with 3.14 as a baseline. V4.0 is much louder 100% and harsh distorted because of it. Sorry you got to kill the window, didn't leave an escape route.
Posts: 21
Threads: 4
Joined: Apr 2022
Reputation:
1
Happy Birthday, QB64PE v4.0!
Thanks to all people that made it possible.
Looking forward to try it out.
Posts: 69
Threads: 24
Joined: Apr 2022
Reputation:
2
Thanks for the hard work! All this effort is very much appreciated!
Posts: 71
Threads: 9
Joined: Apr 2022
Reputation:
11
12-15-2024, 11:10 PM
(This post was last modified: 12-15-2024, 11:14 PM by bert22306.)
Blistering fast:
I ran my number-crunching speed test, as I typically do with new versions of QB64. The test consists of timing how long it takes to compute all the prime numbers from 3 to x, where I run five different x values, and time each one in five separate tests.
All five tests showed obvious improvements. The punch line is that for the longest test, where x = 1,200,007, the time required went from ~ 1.8+ seconds for version 3.14.1 to 1.10 to 1.16something seconds in version 4.0.0. (Same computer, of course.)
Must be the new c++ compiler? Anyway, that's a really substantial improvement. So, thanks, developers!!
Posts: 188
Threads: 14
Joined: May 2024
Reputation:
20
thank you for the new release. channel 4 ho!
i must confess, i have been hesitant to upgrade after 3.11. it forced me to update my debian bookworm system for libpng. otherwise qb64pe compiled successfully and works without problems. i also tried it on debian bullseye system with lxqt. also compiled successfully with one warning (sorry didn't record where) but no problems and works as expected.
i believe the help page for _FLOAT should say that it's possible to have an "F" as sentinel of exponent for scientific notation instead of "E" for SINGLE or "D" for DOUBLE. this is because the new _TOSTR$ function could return _FLOAT value as a string including "F" as is demonstrated in one of the examples on its page.