Tirsvad
375cc0b5f3
The game updates the price and owned text at once but the affordable flag only at its next redraw, so a second purchase in the same visit used a stale flag. Buying at most one product per visit cut the failed purchases from about 40 percent to 1 of 49 and raised the CPS from 16.6 to 29 in the 300 second Docker run. README, screenshots and the wiki in en, da and ms show the new result. Task: MIL-006#1 Task: MIL-006#2 Task: MIL-006#3 Task: MIL-006#4 Task: MIL-006#5
28 lines
1.5 KiB
Markdown
28 lines
1.5 KiB
Markdown
# Step 5: the five-minute run
|
|
|
|
🌐 **English** · [Dansk](Step-5-Five-Minute-Run-da) · [Bahasa Melayu](Step-5-Five-Minute-Run-ms)
|
|
|
|
`run_bot` plays for 300 seconds in one loop: a burst of 20 clicks, then every second one visit to the store: upgrade crates first, then at most one product. At the end `get_cps` reads the cookies per second. The run in the screenshot reached **29 cookies per second** with 29 cursors, 17 grandmas and 2 upgrades (the first release reached 16.6).
|
|
|
|
**Screenshot**
|
|
|
|
|
|
|
|
**Function**
|
|
|
|
```python
|
|
while now - start < duration_seconds:
|
|
click_cookie(game, CLICKS_PER_BURST)
|
|
now = clock.monotonic()
|
|
if now >= next_store_check:
|
|
_buy_while_affordable(game)
|
|
next_store_check = now + STORE_CHECK_INTERVAL_SECONDS
|
|
```
|
|
|
|
## What we gained
|
|
|
|
- **One loop, no threads.** The earlier solution used two threads and an extra package. One loop is easier to read, needs no extra dependency and is easy to test.
|
|
- **Inject the clock.** `run_bot` receives a `Clock`, so a fake clock tests a 300 second run in milliseconds.
|
|
- **Run the GUI in Docker.** Headless Chromium in a container never takes your mouse or keyboard, and the screenshots and the log come out as files.
|
|
- **Where to go next.** Buy by cost per cookie per second and click the golden cookie: each is a new function behind the same ports.
|