Tk Source Code

View Ticket
Login
Ticket UUID: 220203
Title: Width=0 canvas lines are drawn width=1
Type: RFE Version: None
Submitter: welch Created on: 2000-10-31 22:42:38
Subsystem: 05. Canvas Items Assigned To: dkf
Priority: 1 Zero Severity:
Status: Open Last Modified: 2003-11-13 21:32:39
Resolution: Invalid Closed By: welch
    Closed on: 2000-10-31 22:42:38
Description:
OriginalBugID: 2298 Bug
Version: 8.1.1
SubmitDate: '1999-06-28'
LastModified: '1999-10-21'
Severity: LOW
Status: Closed
Submitter: techsupp
ChangedBy: hobbs
OS: Windows 95
OSVersion: 4.00.951
Machine: X86
ClosedDate: '1999-10-21'


Name:
John Nakulski

ReproducibleScript:
canvas .c
    pack .c
    .c create line 1 1 50 50 -width 0 ;# Draws a line of width 1

ObservedBehavior:
Canvas lines of width<=0 are drawn as if width=1

DesiredBehavior:
If I specify line width=0, I expect the line to be invisible (but to
    exist, so that I can later configure the width to be positive).
    
    If I specify width<0, I expect to get an error.



This is actually a feature of X that you can say 0, but you always
get at least a 1 pixel line.  The ability to make items invisible
in general is being considered for a future release of Tk. 
-- 10/21/1999 hobbs
User Comments: dkf added on 2003-11-13 21:32:39:
Logged In: YES 
user_id=79902

Looks like a docfix to me

dkf added on 2003-09-10 17:47:58:
Logged In: YES 
user_id=79902

Note that you see the difference between lines with -width 0 
and -width 1 when you generate and scale PostScript.

dkf added on 2003-02-23 01:19:01:
Logged In: YES 
user_id=79902

Invisible lines can be obtained by setting -fill (or
-outline, depending on item type) to the empty string. 
Given my experience with X, I'd expect zero to draw a
"narrowest visible" line (in some device-dependent sense). 
Negative widths should always be errors, as they're nonsensical!