< Block Plugin (1.4)

Archive.png This article or section documents something not included in the current version of Scratch (3.0). It is only useful from a historical perspective.

This page is a list of bugs in the Block Plugin that allows text to be converted to blocks.

You can only add a bug if you have valid evidence that it is truly a bug. If you are unsure, check on the talk page to see if other people are able to reproduce the bug. Include a basic example of the bug if possible.

Bug reports should be moved to the Resolved Bugs section after they are fixed, or when a discussion on the talk page points out it is not a bug.

List of Bugs

  • Spacing of Scratch Blocks can often appear unusually and not flow with the articles.
when gf clicked
say [Hi.] for (1) secs
broadcast [next v]
stop script
when I receive [next v]
play sound [meow v] until done
say <mouse down?>


  • The When () Clicked block requires brackets around the sprite name, thus creating a text box or drop-down menu (depending on whether a v is put in or not) which is not what the block has in Scratch 1.4. (Shown below is "when [Sprite1] clicked".)
when [Sprite1] clicked


  • An Obsolete Block.png block shows in a darker color than the real thing.
obsolete!


  • C blocks don't have any empty space in their mouths when they are not holding any blocks in them.
repeat (10)


  • The puzzle-piece bump on the if block covers up the corner of the Boolean area if there is no condition.
if < >
end


  • When a Cap block is used within a C block, the puzzle-piece notch on the bottom is shown. (The notch should only be there when a Stack block is placed at the bottom.)
if <(Variable) = (1)>
  stop script
end


  • Content in a Reporter Block that wraps is not padded enough.
set [number v] to (join ((round (([10^ v] of (decimal places)) * ([sqrt v] of ((item (1) of [list v]) * (3))))) / ([10^ v] of (decimal places))) [])


  • Long text in hat blocks cause rendering problems. This is due to the hat block image being a certain length.
when [Lorem ipsum dolor sit amet, consectetur adipiscing elit.] clicked


stop all sounds
stop all pinks
  • Free-floating comments are rendered as an obsolete block linked to the comment.
//See?
motor on
motor off
motor on for (1) seconds
motor power (100)
motor direction [this way v]
  • There is no way to differentiate variable and list reporters.
(variable)
(list)
  • When forever and forever if blocks are used with "end" they have a puzzle piece stub on the bottom.
forever
end
  • Some blocks appear obsolete when there is a space between the argument and the last character.
set size to (100) %
<color [#FF0000] is touching [#FF0000] ?>
  • Reporters and boolean jump to the top of a script when not in an argument
next costume
next costume
(var)
<[] = []>
  • Sensor Value block renders as obsolete if used with < > and not ( ). Else it renders as a variable when used with ( )
([slider v] sensor value)
<[slider v] sensor value>
if <sensor [button pressed v]?>
 say [Hi]
end
set [sprite1name v] to [sprite1]
say ([x position v] of (sprite1name))
say ([sqrt v] of (5))
  • The lengths of the bottoms of C blocks do not match the lengths of their tops and are too short.
when gf clicked
forever if <loud?>
turn ccw (1) degrees
ask [Hello, I'm Scratch Cat. What's your name?] and wait
if <(answer) = [XYZ]>
say [Wow. Wow! WOW!!!] for (2) secs
end
  • Repeat, repeat until, if, and if, else blocks have no puzzle piece stub on the bottom when used without end. (Compare to last Scratchblocks entry.)
repeat (10)
move (10) steps
repeat until <touching color [#F1D58A]?>
point towards [Sprite2 v]
move (3) steps
if <key [space v] pressed?>
think [Bleep bloop blop beep boop bop.] for (10) secs
if <<(timer) > (100000)> or <(timer) = (100000)>>
say [100,000 seconds and counting!] for (5) secs
else
say [Wait...] for (2) secs
if <loud?>
say [Can you be quiet, please?] for (2) secs
end
  • C blocks lose their Boolean slot and render improperly when more than one of any condition block is used.
if << (x) < [1] > and < (y) < [1] >>
 say [Hello!]
end
  • Free-floating punctuation mark render with &nbsp in front.
ask [Move ? steps?] and wait
  • No empty dropdowns.
play sound [ v]
  • Long conditions in Boolean inserts can cause rendering problems.
when I receive [tween v]
repeat until <<([abs v] of ((target x) - (x position))) < [.25]> and <([abs v] of ((target y) - (y position))) < [.25]>>
 change x by (((target x) - (x position)) / (2))
 change y by (((target y) - (y position)) / (2))
end
go to x: (target x) y: (target y)
  • Else is found separately from if.
else
move (10) steps
end
if <>
else
else
else
end
  • In the forums, puzzle-piece stubs at the top of a script are white.
  • Bottom puzzle-piece stubs have a bar at the top in iOS devices.
  • Bad graphics in Internet Explorer (e.g. blocks are square, hat blocks are extremely long and have a huge gold chunk above the arch, puzzle-piece stubs are rectangular). This seems to be more of a bug with the browser than the plugin.
  • The rotation blocks' arrow images aren't vertically centered.
when gf clicked
set [d v] to (90)
set [s v] to (3)
reset timer
repeat until <(timer) > (s)>
turn ccw ((timer)*((d)/(s))) degrees
end
  • "...", when all alone, causes rendering problems.
...
  • You can have freely floating number inserts, string inserts, and dropdowns!
(...)
(2)
[hello world]
[dropdown v]
  • Empty Boolean inserts, when not in a block, render as an empty space.
<>
  • The parser ignores names of stack blocks shorter than three characters.
a
12
123
  • Comments glitch out of the page if they are too long.
move (1) steps // This block makes the sprite the script is in move one pixel, if you want it slower; you can also try and make the number larger, but I personally recommend one step.

Resolved Bugs

  • Blocks and arguments can be caused to appear in incorrect shapes.
pick random (mouse down?) to <[0]+[0]>
x position
y position
direction
costume #
size
answer
mouse x
mouse y
timer
loudness

pick random (mouse down?) to <[0]+[0]> x position y position direction costume # size answer mouse x mouse y timer loudness

are some examples.

Not a bug. The parser will identify the blocks by their names, not arguments.
3sal2 (talk | contribs) 21:04, 20 May 2012 (UTC)
  • The Set () to () can render incorrectly when long definitions are placed inside.
when gf clicked
forever
set [variable v] to ( (timer) + ( [10 ^ v] of ( (direction) / ( pick random ( length of [the] ) to ( length of [world] ) ) ) ) )
Not a bug. The blocks wrap.
Scimonster (talk | contribs) 18:42, 31 March 2012 (UTC)
  • Reporter blocks are too rounded on the edges.
(timer)
Not a bug. They'll look approximately like that from screenshots and the pre-alpha. And that's how it's done with CSS.
Veggieman001 (talk | contribs) 01:18, 31 March 2012 (UTC)
think [Hmm..] for (2) seconds
say [Hello] for (2) seconds
It's 'secs', not 'seconds' for both.
Veggieman001 (talk | contribs) 23:18, 29 January 2012 (UTC)
  • Answer (block) appears as the color of a variable block, not a sensing block. For example:
    (answer)
Fixed as of now.
JSO (talk | contribs) 10:04, 2 January 2012 (UTC)
  • Block plugin does not work when JavaScript is disabled. Could possibly be fixed by porting code to direct PHP.
The code was actually ported from PHP to javascript (and then greatly improved), so it could be used on all Scratch related sites without effort, and to put no stress on the servers. The amount of people visiting with javascript disabled is extremely low. Without javascript, the blocks are still easily readable in a monospaced font.
JSO (talk | contribs) 09:58, 2 January 2012 (UTC)
I put a hackish fix in place similar to the one used for the double '... of ...' blocks (sensing and operators). Basically, it now renders in dark orange if the second parameter is a dropdown, otherwise it will render in lightgreen. As Scratch doesn't allow you to drop a variable in that dropdown gap, it should be correct.
say (length of [list v])
say (length of [some random string])
say (length of (variable))

JSO (talk | contribs) 10:23, 2 January 2012 (UTC)
  • Blocks do not have puzzle-piece holes, to connect them.
Fixed as of recent update.
Lucario621 (talk | contribs) 01:13, 4 January 2012 (UTC)
Fixed.
Lucario621 (talk | contribs) 21:05, 7 January 2012 (UTC)
  • Colors can be entered into a text area of any block by using HEX format.
Not a bug.
Lucario621 (talk | contribs) 21:05, 7 January 2012 (UTC)
  • Cap Blocks have a puzzle-piece bump on the bottom while they are not supposed to. (Blocks can be stacked underneath them too; hopefully this should be fixed with the shape-fixing.)
when green flag clicked
stop script


Fixed as of now. Blocks can be stacked underneath them - it's up to the author to write a script correctly. The plugin will try to render it no matter what.
  • The () Mod () and Pick Random () to () blocks are not supported in the plug-in.
  • An empty number insert shows as a glitched variable.
  • An empty (or filled with spaces) textbox renders incorrectly.
  • Blocks with a boolean insert don't render properly without them filled.
if <<(condition) = [true]> and <>>
move (5) steps
  • Dropdown boxes in reporters are not padded enough.
  • The puzzle-piece bump on the "set [v] to ()" covers up the corner of the variable box

Scratch Blocks Bug.png

All fixed
JSO (talk | contribs) 00:21, 8 January 2012 (UTC)
  • Sometimes a dropdown at the beginning and a dropdown at the end of a block can cause rendering errors.
[Sprite1 v] go to [Sprite2 v]
I don't think this is a bug. The above block is not a real block, nor is there a stack block with a dropdown at the beginning and a dropdown at the end.
([x position v] of [Sprite1 v])
is the only other block with a dropdown at the beginning and one at the end, and it works just fine.
Mathfreak231 (talk | contribs)


  • Puzzle-piece stubs cover up dropdown menus at the beginning of the block.
[asin v] of (2)
Another bug of this case. There is no stack block with a dropdown at the beginning of it.
([x position v] of [Sprite1 v])
and
([asin v] of (3.5))
work fine.
Mathfreak231 (talk | contribs)
Cookies help us deliver our services. By using our services, you agree to our use of cookies.