processing4/build/shared/bugs.txt

KNOWN BUGS FOR THE MOST RECENT RELEASE

this was moved out of the release notes since it's getting too long,
and since only a handful of things change between releases


..................................................................


COMMON PROBLEMS (NOT BUGS)
things that are often perceived as bugs, which means maybe they'll be
changed in the future, but for now, they're not actually 'broken'.


- names of sketches cannot start with a number, or have spaces
  inside. this is mostly because of a restriction on the naming of
  java classes. i suppose if lots of people find this upsetting, we
  could add some extra code to unhinge the resulting class name from
  the sketch name, but it adds complexity, and complexity == bugs. :)

- if you get a "ClassCastException" when trying to run your code in a
  browser, you're probably trying to run non-java 1.1 code. this 
  means any classes that didn't exist in java 1.1. or, if if you're
  exporting as applet, variables that are 'final int' need to be
  'static final int'. otherwise they'll have trouble with a java
  1.1-only machine: (i.e. running microsoft's java on windows). you
  should also avoid using "final" variables *inside* of methods, 
  since those aren't supported by microsoft's vm (and that's the 
  one most widely installed).

- 'export to application' remains unimplemented, though it's closer,
  and should be ready before beta.

- 'preferences' will get better in a future release

- on macosx, the error message:
  ## Component Manager: attempting to find symbols in a component
  alias of type (regR/carP/x!bt)
  is not something from processing, it is a bug in toast 5.2's video
  cd support extension. to suppress the messages, move "Toast Video CD
  Support.qtx" out of /Library/QuickTime (at least until you need to
  do video cd work).

- on windows, sometimes your programs will run in a window that has a 
  java coffee cup icon, instead of the usual processing icon. this 
  simply means that the application is being run outside of processing
  (by java itself) rather than from within processing. this happens
  when extra libraries, multiple source code files (more than one tab), 
  or extra code files in the 'code' folder are employed.


..................................................................


QUICKTIME FOR JAVA / VIDEO PROBLEMS

- quicktime for java is still required on windows. it's required on
  macosx too, but it's installed there by default. the requirement
  will be removed in a later release for people who don't want to use
  video, but the fix is more difficult than anticipated, so it will
  probably happen after the initial beta release. 

- on windows, if you can't install quicktime for java yourself, all
  you really need is a copy of QTJava.zip from the quicktime install,
  which needs to be at %WINDIR%\system32\QTJava.zip. WINDIR is an
  environment variable on windows, open up a DOS prompt and type 
  echo %WINDIR% 
  to find out what it's set to (prolly something like C:\windows on
  windows xp/me/98 or c:\winnt on windows 2000).

- on macosx 10.2.x (jaguar), quicktime for java was missing from the
  initial release of quicktime 6.4 on the mac. an upgrade to quicktime
  6.4 that includes qtjava is now available from software update. or
  you can download it from versiontracker, since software update
  doesn't show the update on some machines:
  http://www.versiontracker.com/dyn/moreinfo/macosx/21302
  you can also upgrade to panther (macosx 10.3) which doesn't have
  this problem.

- video sometimes likes to crash the application completely on
  windows. this seems to be due to problems with quicktime for java
  (since we shouldn't be able to write code that crashes using
  java.. heh right). so this is probably outside of our control.


..................................................................


NETWORKING API BUGS

- net is fairly horked, you should probably just use the java networking 
  api until we get a chance to fix it.

- network examples haven't been updated for the latest api, so they
  will likely break in releases 60 and higher.

- net needs work, it's currently sending unicode, instead of ascii or
  UTF-8, data. this can cause problems with other programs.


..................................................................


KNOWN BUGS
the real deal, keeping us awake at night.. this is only a partial
list, see the bboard for more, and inside ben's head for others.


- beautify doesn't retain the *exact* position of the caret within the
  text. 

- beautify has problems with indenting when lines with a brace are
  commented out, and also ignores switch statements.
  http://processing.org/discourse/yabb/YaBB.cgi?board=Proce55ing_software_bugs;action=display;num=1067309918;start=0

- clipping planes are not yet implemented. meaning that objects may
  sometimes fly into the camera and turn upside down or do other tricks.
  http://processing.org/discourse/yabb/YaBB.cgi?board=Proce55ing_software_bugs;action=display;num=1058491568;start=0
  http://processing.org/discourse/yabb/YaBB.cgi?board=Proce55ing_software_bugs;action=display;num=1052313604;start=0
  http://processing.org/discourse/yabb/YaBB.cgi?board=Proce55ing_software_bugs;action=display;num=1042699742

- there is a bug that when smoothing is enabled, sometimes thin lines 
  will be visible within a concave polygon. this will just have to be 
  fixed in a future release. however, as long as stroke() is enabled 
  on the polygon, the lines should not be visible.

- the app doesn't provide an option to 'cancel' when it so politely asks
  if you want to "save changes before quitting?". this is because of
  java weirdness that we're trying to straighten out. the vm seems to
  be force quitting itself, even if cancel is selected. 

- lines with stroke weights that aren't 1 are sometimes strange.

- single pixel lines still have some problems with drawing over things.

- the size of exported applets has jumped in recent releases. this will
  get smaller again in a future release. we're in the midst of
  transitioning between two different sets of rendering/graphics code.

- curveTangent() is not yet implemented. contributions welcome.

- strokeMiter() and strokeJoin() exist but are not implemented.

- textures warp in strange ways due to perspective problems.
  http://processing.org/discourse/yabb/YaBB.cgi?board=Proce55ing_software_bugs;action=display;num=1050650262

- smooth() is not completely implemented. there will be quirks in how
  it draws (like a weird flat line at the end of a shape, or areas
  that are less than smooth).

- on windows, if you hit 'run' and no display window comes up, it probably
  means that you have processing inside a folder with non-ascii characters
  in its name. for instance, if the name of your user folder uses japanese
  text, or it has an umlaut or cedilla or anything else fancy, and processing 
  is on the desktop, or somewhere inside that folder, you're gonna
  have trouble. the fix is to place processing inside the root level of 
  your C:\ drive (or whatever drive you use) and it'll be ok again.
  http://processing.org/discourse/yabb/YaBB.cgi?board=Proce55ing_software_bugs;action=display;num=1062794781

- jikes is ~9 megabytes in size, which is a bit obnoxious. it doesn't
  hurt anything but we plan to put it on a diet for a future release.
  http://processing.org/discourse/yabb/YaBB.cgi?board=Proce55ing_software_bugs;action=display;num=1062581723

- sound doesn't always stop when the 'stop' button is hit (i.e. the
  mixer example seems broken). for now, you have to first quit
  processing to stop it. 

- the horizontal scroller for the text area has some weirdness, this
  may not be fixed until after beta, depending on how bad it is.

- using saveFrame() or saveBytes() or some of those things will cause
  trouble inside applets. we'll try to do better error trapping soon
  so that your applet doesn't crash when this happens. 
  http://processing.org/discourse/yabb/YaBB.cgi?board=Proce55ing_software_bugs;action=display;num=1059928189;start=0


..................................................................


BUGS THAT WE PROBABLY CAN'T FIX
these are things that are out of our control


- "create font" crashes sometimes on windows, bringing down the whole
  environment. this seems to be a jdk bug, because it's not a java
  exception, but a full crash.