Number Master

This is a Mastermind-style code-breaking game using digits 1–6 instead of colours. The program generates a four-digit secret code stored in A$, then gives the player up to nine attempts (lines 80–300 loop from column 21 down to 5 in steps of −2). Exact-position matches are shown with a filled block character and misplaced-digit matches with a dotted block character, displayed to the right of the guess. Line 50 uses the idiomatic VAL “RND*5+1” trick to embed a numeric expression as a string, saving memory and deferring evaluation. The scoring logic modifies working copies B$ and C$ in-place to avoid double-counting matched digits across both the exact and misplaced passes.

---

Program Analysis

Program Structure

The program is organised into four logical phases:

1. Initialisation (lines 10–60): Prints the title banner in inverse video, dimensions two four-character string arrays, and fills A$ with a random secret code.
2. Game loop (lines 80–300): Iterates over FOR A=21 TO 5 STEP -2, giving nine rows on screen, each handling one guess.
3. End-of-game (lines 310–4020): Prints either “GAME OVER” or nothing (using the AND boolean idiom), reveals the secret, shows the number of tries, pauses, clears, and restarts.
4. Alternate path (lines 5000–5010): Dead code reachable only by direct GOTO — never actually reached in normal play because line 310 handles all endings. STOP prevents further execution.

Secret Code Generation

Line 50 builds each character of the secret code using:LET A$(N)="123456"(VAL "RND*5+1")

The VAL "RND*5+1" idiom evaluates the expression at run-time while storing it as a string literal in the token stream, saving a few bytes of RAM. Slicing a string literal with a computed index is a clean ZX81/TS1000 pattern to pick a random element from a fixed set — here digits 1 through 6, corresponding to Mastermind’s six colours.

Scoring Algorithm

The scoring is a two-pass algorithm that correctly handles duplicate digits:

1. Pass 1 (lines 130–190): Scans for exact matches (right digit, right position). Matching positions in both B$ (the guess copy) and C$ (the code copy) are overwritten with a filled block graphic \;; to prevent them being counted again in pass 2.
2. Pass 2 (lines 210–290): Scans remaining positions for misplaced matches (right digit, wrong position). Each match overwrites the relevant position in B$ with a dotted block \.. to prevent double-counting, and prints the misplaced indicator.

The variable B (reset to 0 each round at line 90) acts as a print-column accumulator for the scoring markers, advancing by 2 per marker. When B>=8 all four positions are matched and the game ends in victory via GOTO 310.

Screen Layout

Column range	Content
Col 9–12	Player’s four-character guess (B$)
Col 0, 2, 4, 6	Scoring markers (exact = \;; ▐, misplaced = \.. ▄)
Col 6, row 1	Title “NUMBER MASTER” in inverse video

Key BASIC Idioms

• PRINT "..." AND condition — prints the string only when the condition is true (non-zero), a standard ZX BASIC space-saving trick used in line 310.
• VAL "11-A/2" in line 320 computes the number of tries from the loop counter A without a separate variable. Since A steps from 21 down by 2, the formula recovers the 1-based attempt count.
• PAUSE 40000 at line 4000 provides a long but finite pause (approximately 11 minutes at 50 Hz) before auto-restarting — effectively “press any key to continue” without an INKEY$ loop.

Bugs and Anomalies

• Variable B used for two purposes: B serves as both the scoring-marker column counter and (implicitly) a “score” register. This works because it is reset to 0 at line 90, but the dual role makes the code harder to follow.
• Line 105 validation: The guard against spaces in the input only checks for a literal space character; it does not validate that each character is in the range 1–6, so invalid input (e.g. letters) is silently accepted.
• Lines 5000–5010 are unreachable: Line 310 always prints the game-over state and falls through to line 320; no code path issues a GOTO 5000. This block appears to be a leftover from an earlier version.
• DIM B$(4) at line 30 allocates only four characters, matching the four-digit guess. However, INPUT B$ at line 100 does not constrain input length, so entering more than four characters will cause a subscript error when lines reference B$(5) etc. implicitly through the four-character DIM. In practice the DIM limits the usable string to exactly four characters.
• NEXT F placement: NEXT F at line 280 is inside the body of the inner loop but positioned after a GOTO 290, meaning it is only reached when no match is found. This is an intentional early-exit loop pattern, not a bug.