leo25 / arduino Goto Github PK

Add syntax highlighting for:

continue
break

The Servo library on the Mega should use pins 11 and 12, not 9 and 10.  

Some additional bugs in the original Ethernet Client library made it through
the revisions that were included in 0016. Basically the bugs are sanity
checks in status() and connected() to ensure that a socket is associated
with the client before dispatching a status request to the Wiznet chip. When
a client connection is closed and the socket released, the _sock variable
gets set to 255. Apparently when the invalid socket number 255 is passed to
the Wiznet chip in the getSn_SR(_sock) call, the result (for some unknown
reason) is 1 - which is the status for an open connection! (Sn_IR_CON).

So this results in a nasty situation where a correctly stopped and closed
connection will return true when the connected() function is called and
return an established connection result when status() is called.

Forum reference:
http://www.arduino.cc/cgi-bin/yabb2/YaBB.pl?num=1238640832/13#13

Below is a patch for Client.cpp (also attached as a file).  Basically it
just adds checks to return a default not connected result if the _sock value
is 255 (not bound to a socket).

--------------------------PATCH BEGIN--------------------------
--- Client.cpp        2009-05-30 14:50:02.000000000 -0400
+++ Client_patched.cpp    2009-06-09 21:05:32.000000000 -0400
@@ -113,13 +113,21 @@
 }

 uint8_t Client::connected() {
-  uint8_t s = status();
-  return !(s == SOCK_LISTEN || s == SOCK_CLOSED || s == SOCK_FIN_WAIT ||
-    (s == SOCK_CLOSE_WAIT && !available()));
+  if (_sock == 255) {
+    return 0;
+  } else {
+    uint8_t s = status();
+    return !(s == SOCK_LISTEN || s == SOCK_CLOSED || s == SOCK_FIN_WAIT ||
+      (s == SOCK_CLOSE_WAIT && !available()));
+  }
 }

 uint8_t Client::status() {
-  return getSn_SR(_sock);
+  if (_sock == 255) {
+    return SOCK_CLOSED;
+  } else {
+    return getSn_SR(_sock);
+  }
 }

 // the next three functions are a hack so we can compare the client
returned
---------------------------PATCH END---------------------------

Here are the revised functions after patching:

---------------------------------------------------------------
uint8_t Client::connected() {
 if (_sock == 255) {
   return 0;
 } else {
   uint8_t s = status();
   return !(s == SOCK_LISTEN || s == SOCK_CLOSED || s == SOCK_FIN_WAIT ||
     (s == SOCK_CLOSE_WAIT && !available()));
 }
}

uint8_t Client::status() {
 if (_sock == 255) {
   return SOCK_CLOSED;
 } else {
   return getSn_SR(_sock);
 }
}
---------------------------------------------------------------

Technically only the patch to the status() function is needed as it will
correctly make the existing connected() function work. Adding the sanity
check to connected() as well makes the code easier to follow.

From: http://www.arduino.cc/playground/Main/SuggestionsBugs?action=edit

Keyboard shortcut in the IDE for turning Serial Monitor on/off: to allow 
compile/upload/monitor without taking hands off keyboard

It had 9 votes.

What steps will reproduce the problem?
1. As of Arduino-15, Wire library doesn't work for Pro / Pro Mini
2.
3.

What is the expected output? What do you see instead?
Compiles, no errors. Wire library not functioning properly. Check your
#ifdefs for the Arduino MEGA in Wire code. The souce of the problem
probably lies there.


What version of the Arduino software are you using? On what operating
system?  Which Arduino board are you using?
Arduino-15
Mac OSX
Funnel I/O


Please provide any additional information below.

On the previous Bugs and Suggestions page, this was the most vote for item
(13 votes). People really would like a sleep function to save power on
arduino when used in certain environments, on battery, etc.

Many users use Processing and Arduino together and find that they cannot
control which program boots when clicking on a file.

This is a long-standing request from the playground suggestions list.

Paul Badger

What change would like to see?
the addition of a serial.end() function to turn of the serial port and
return the pins to digital function.
An implementation is shown here
http://www.arduino.cc/cgi-bin/yabb2/YaBB.pl?num=1232485147/1 which worked
with 0013

Why?
Sometimes you need the extra pins, sometimes you need to control the port
pins for custom handshaking.

Would this cause any incompatibilities with previous versions?  If so, how
can these be mitigated?
Not sure, i doubt it.

Autosave of current project.  This can happen could be implemented in a
number of ways:

1: Auto save on verify or upload.  Configurable, select none, verify or upload.

2: Auto save at N minutes.  Configurable, user selects how often (0 means
never) project is autosaved.

3: Auto save on N chamges. Configurable, user selects how often (0 means
never) project is autosaved.

For option 1: auto may overwrite the current project (same as file/Save)

for options 2 and 3 auto save should create a new project with auto
increment numerical extension.  The number of versions saved should be user
configurable.


Why

Presently when you compile a project in the IDE the source is not saved to
disk.  If the host computer crashes (which can happen for a number of
reasons) the source code is lost.



Would this cause any incompatibilities with previous versions?  If so, how
can these be mitigated?

Shouldn't cause any incompatibilities.

What change would like to see?
A nice white background, and clear direct navigation to the oarts am
interested in.

Why?
A proper "clean" website, without the battle grey, otther formats or
restrictions.

As I've examining your html + code, and do appreciate the legacy, whatever
app it is.

Where I wpuld like to help is in a new design, more ammenable and simpler.
Full on CSSS and out with the <inline tags>

As the application appears to be Wiki then I am sure we can go places with it.

What steps will reproduce the problem?
1. cd ~
2. ~/arduino-0015/arduino

What is the expected output? What do you see instead?
The IDE should start.  Instead, you get:
Exception in thread "main" java.lang.NoClassDefFoundError: processing/app/Base

This is because the script assumes that it is run from it's own directory.
 The attached patch removes this assumption.

In File/Sketchbook/Examples, 

the entries starting with Library show the .svn directory.

What steps will reproduce the problem?
1. write a program that includes some html markup, or text that could be
confused to be markup. #include lines that use angle brackets fall into
this category.

2. select the program and choose "Copy as HTML".

3. paste the html into a file and observe the undesired behavior with a web
browser.

What is the expected output? What do you see instead?

I expected a colorized version of this program:

void loop() {
  Serial.print("<h1>hello world</h1>");
}

Instead I got a display with a giant heading in the middle of it. I've
attached a snapshot of the output.


What version of the Arduino software are you using? On what operating
system?  Which Arduino board are you using?

I am using 0016 for the mac, downloaded today.


Please provide any additional information below.

I believe that the "Copy as HTML" commands needs to do more escaping, or
that the escaping it is doing is getting lost somehow.

In Processing the edit menu has a block indent.

It's not there in Arduino 0016


What version of the Arduino software are you using? On what operating
system?  Which Arduino board are you using?

0016
Windows
2009/Mega


Please provide any additional information below.

Edit/Increase Indent is available in the Procesing environment, but not in
Arduino.

What steps will reproduce the problem?
1. start arduino0015
2. watch output

What is the expected output? What do you see instead?

proppy@kakarichou:~/Desktop/arduino-0015$ ./arduino 
Experimental:  JNI_OnLoad called.
Stable Library
=========================================
Native lib Version = RXTX-2.1-7
Java lib Version   = RXTX-2.1-7

utility/w5100.c:30:19: error: stdio.h: No such file or directory
utility/w5100.c:31:20: error: string.h: No such file or directory
utility/w5100.c:33:27: error: avr/interrupt.h: No such file or directory
In file included from utility/w5100.c:36:
utility/types.h:19:26: error: avr/pgmspace.h: No such file or directory
utility/w5100.c: In function ‘IINCHIP_WRITE’:
utility/w5100.c:112: error: ‘DDRB’ undeclared (first use in this function)
utility/w5100.c:112: error: (Each undeclared identifier is reported only once
utility/w5100.c:112: error: for each function it appears in.)
utility/w5100.c:112: error: ‘PORTB’ undeclared (first use in this function)
utility/w5100.c:112: error: ‘SPCR’ undeclared (first use in this function)
utility/w5100.c:117: error: ‘SPDR’ undeclared (first use in this function)
utility/w5100.c:117: error: ‘SPSR’ undeclared (first use in this function)
utility/w5100.c: In function ‘IINCHIP_READ’:
utility/w5100.c:154: error: ‘DDRB’ undeclared (first use in this function)
utility/w5100.c:154: error: ‘PORTB’ undeclared (first use in this function)
utility/w5100.c:154: error: ‘SPCR’ undeclared (first use in this function)
utility/w5100.c:157: error: ‘SPDR’ undeclared (first use in this function)
utility/w5100.c:157: error: ‘SPSR’ undeclared (first use in this function)
utility/w5100.c: In function ‘wiz_write_buf’:
utility/w5100.c:194: error: ‘DDRB’ undeclared (first use in this function)
utility/w5100.c:194: error: ‘PORTB’ undeclared (first use in this function)
utility/w5100.c:194: error: ‘SPCR’ undeclared (first use in this function)
utility/w5100.c:201: error: ‘SPDR’ undeclared (first use in this function)
utility/w5100.c:201: error: ‘SPSR’ undeclared (first use in this function)
utility/w5100.c: In function ‘wiz_read_buf’:
utility/w5100.c:237: error: ‘DDRB’ undeclared (first use in this function)
utility/w5100.c:237: error: ‘PORTB’ undeclared (first use in this function)
utility/w5100.c:237: error: ‘SPCR’ undeclared (first use in this function)
utility/w5100.c:243: error: ‘SPDR’ undeclared (first use in this function)
utility/w5100.c:243: error: ‘SPSR’ undeclared (first use in this function)

What version of the Arduino software are you using? On what operating
system?

0015 on Ubuntu GNU/Linux Jaunty 9.04

What change would like to see?
Every nut and bolt and hinge+bracket.. documentated.. and shy it does it
Why?
Cos the docs atmo suck. We need to neutralise that in whaever format.

Would this cause any incompatibilities with previous versions?  If so, how
can these be mitigated?

Documentation is a core to adoption.

What steps will reproduce the problem?
1. Open terminal / command line.
2. Execute "svn checkout http://arduino.googlecode.com/svn/trunk/
arduino-read-only"

What is the expected output? What do you see instead?
Expected: get entire source tree
Instead: checkout gets stuck at file build/macosx/dist/librxtxSerial.jnilib
- I tried multiple times and aborted after waiting for half an hour for the
file (on a fast connection where checkouts of entire tree normally take 2
minutes).

What version of the product are you using? On what operating system?
Mac OSX 10.5.5 
svn, version 1.4.4 (r25188)


Please provide any additional information below.
I was able to check out the entire tree from Google code successfully on
Windows/cygwin a day or two ago.

This isn't really an arduino issue - but it is preventing me checking out
the code. I have the same problem that various other people have reported
here and there on the net: when I do a checkout of the latest trunk it gets
as far as trunk/build/macosx/dist/librxtxSerial.jnilib and then hangs. I
have tried the checkout now over ten times on two separate days. Perhaps at
some point soon the problem will go away - but at the moment it's
definitely reproducible...

I am trying to checkout using svn 1.4.6 on an Ubuntu linux box.

What steps will reproduce the problem?
1. After applying the most recent Java security patch from Apple last night
I started getting this error when starting:

Uncaught exception in main method: java.lang.UnsatisfiedLinkError:
/Applications/arduino-0016/Arduino
16.app/Contents/Resources/Java/librxtxSerial.jnilib:  no suitable image
found.  Did find:  /Applications/arduino-0016/Arduino
16.app/Contents/Resources/Java/librxtxSerial.jnilib: no matching
architecture in universal wrapper


What version of the Arduino software are you using? On what operating
system?  Which Arduino board are you using?

Arduino 016, MacOSX 10.5.7, JVM 1.5 32 bit, Arduino Duemilanove

Please provide any additional information below.

Processing is working fine, including serial communications on the same system

The dependencies of the software (rxtx, antlr, others?) shouldn't be
packaged with the software but called out in the installation instructions.
 This makes it easier for package managers to install it system-wide and
deploy on different architectures.

An enhanced preferences page.

It would nice to see some of the more useful parts of the prefs file moved
onto a preferences page. Such as the definition of tab stops, for only one
example.

Or is it there stated / unstated policy that Arduino IDE should be lashed
to the Processing IDE as much as possible? If so then asking for these
improvements really only makes sense in Processing forums?

Paul Badger 

What steps will reproduce the problem?
1. Connect to Arduino Mega
2. download program that writes to Serial
3. In IDE click "monitor serial port"

What is the expected output? What do you see instead?
IDE crashes with not error dialogs.

What version of the Arduino software are you using? On what operating
system?  Which Arduino board are you using?
Arduino Mega & Arduino-0015
Ubuntu 9.04 64-bit

Please provide any additional information below.
Compiling and downloading to Arduini Mega work fine. only crashes when I
try to monitor the serial port.

using another application to monitor serial port works fine.

What change would like to see?
The Arduino IDE should have its own file type (.ade)

Why?
So that files open in the correct IDE (not Processing)

Would this cause any incompatibilities with previous versions?  If so, how
can these be mitigated?
These would be mitigated by users opening and resaving in the Arduino IDE as 
.ade files or by 
renaming them to .ade, and by poping up a message on first use of the new IDE

Some duplication of bugs and enhancement requests is being encouraged on
the Playground wiki: http://www.arduino.cc/playground/Main/SuggestionsBugs

I'd like to suggest that page be closed and people be encouraged to use the
issue queue here instead. Then people could be encouraged to move issues
they are concerned about in here as a vote for continuing to give that
issue attention.

When will there be a 64-bit version available?  I assume that there will be
one as the previous version had one as well...

Compiling the standard MegaServo code from arduino.cc will cause servos to
jitter in Arduino-0016. The Servo library on the other hand works perfectly.


What is the expected output? What do you see instead?
Expected: flawless servo turning
Instead: Jittering, servo is "trying", but is not getting there.
See http://www.youtube.com/watch?v=ooXQ2H03oXw for a demo on jittering.

Topic is hosted on the forum as well :
http://www.arduino.cc/cgi-bin/yabb2/YaBB.pl?num=1244582253


What version of the Arduino software are you using? On what operating
system?  Which Arduino board are you using?

Arduino 0016
Windows Vista
Arduino Duemilanov

Is anyone working on a debian package for Arduino ?

If not then I wouldnt mind having a go at the calleng.

This was submitted to the mailing list by Christopher K. Johnson. I'm just
adding it to the issue list since the decision was to have patches moved
here to keep things organized.

-----------
In testing use of the i2c (twi) interface to control a PCA9505 I had to dig
into the Wire library to implement the correct twi bus sequence for reads
from the device.  The 9505 like some other devices such as eeproms and 9535
expects that after the slave address+w a single byte is written to go in
the control register specifying what register offset the read should start
at, and whether the addressed register should be automatically incremented
with each byte read.  Then the master should without sending a stop, send a
repeat start on the twi bus and the slave address+r and begin the read.

Write operations similarly require register offset and AI bit, but that was
possible to accommodate with the existing Wire functions by adding it as
the first byte on the write, and increasing the count of bytes to be
written by one.

I added support for the correct read sequence in twi.c and a readFromReg
function in Wire.c, and have fully tested read and write operation with the
9505.

I also found a pre-existing bug in twi.c wherein it would actually read via
twi one more byte than was requested, and save it in the input buffer due
to the fact that on reads an ack/nack occurs before the ISR is invoked, so
when the next to last byte is received the bit needs to be set to cause a
nack of the last byte.  Previously it was waiting until the last byte
requested was received (and ACK'd) before setting the bit to nack when the
next byte (one too many) was received.

I improved some comments and indentation.

I made a couple of other code tweaks in twi.c that should not alter behavior.

I set the default twi bus clock (TWI_FREQ) to 400KHz instead of 100KHz
because I believe support for that speed to be the norm, not the exception.
 That is the only change which might adversely affect existing uses of the
Wire library.

A recursive diff of changes is attached.  I am new to this list and not
sure what the best way is to submit such changes for review and
incorporation in future releases.  If there is a more acceptable way for me
to do so, please advise.  Furthermore I am not sure where to begin to make
changes to the online documentation for the library if someone could point
me in the right direction.

Other changes I would like to discuss:
1) Consider changes to enable twi hardware response to twi bus general call
sla+r/w requests.
2) In a multi-master twi bus environment there are some weaknesses in the
current design - race conditions on testing then altering twi_state in
functions not protected from interruption as well as changes to twi_state
within the twi isr, and the lack of a way to indicate failure of a read
operation due to multiple possible factors such as bus contention.

Sincerely,
Christopher K. Johnson

If $LD_LIBRARY_PATH is empty, it could allow arbitrary code to be executed
from the directory it's run.  Line 13 should read:

LD_LIBRARY_PATH=`pwd`/lib${LD_LIBRARY_PATH+:$LD_LIBRARY_PATH}

[deleted issue]

Many expect Shift backspace to behave as unshifted backspace, not as an
alternate for delete.  This is maddening if it behaves as delete.  A
preference to specify the desired behavior in the editor would be a godsend.

What change would like to see?
The IDE uses AWT-Dialogs for file-open file-save. Replace these dialogs
with their swing counterparts.

Why?
The look and user experience of the AWT-dialogs does not integrate with any
of the linux desktop themes in use.    

Would this cause any incompatibilities with previous versions? 
No.

Two files attached:
AWT.jpeg shows the current look on linux 
swing.jpeg shows a standard javax.swing.JFileChooser (from another app)


Eberhard

Can the Arduino Team please post a policy about who has access to the code
and can make changes in code?

If David Mellis is to be the sole arbiter of Arduino code can that please
be stated somewhere public so potential coders will at least know how the
bureaucracy works. In a similar vein, a sentence or two on philosphy, and
what is likely to be added to core or libraries would also be helpful for
potential contributors.

On another note opening up the code would be a way to get a lot of the
issues we see in this list implemented quickly.

Paul Badger

What steps will reproduce the problem?
1.check the source of arduino-0015 with the static code analysis tool
cppcheck (http://cppcheck.wiki.sourceforge.net/)

What is the expected output? no errors expexted.
What do you see instead?
the output is as followed:

arduino-0015$ cppcheck -a -q -j2 -f -v .
[./hardware/libraries/Firmata/Firmata.h:106]: (all) Memory leak:
FirmataClass::firmwareVersionVector
[./hardware/libraries/Matrix/Matrix.h:34]: (all) Memory leak: Matrix::_buffer
[./hardware/libraries/Sprite/Sprite.h:36]: (all) Memory leak: Sprite::_buffer
[./hardware/libraries/Wire/Wire.h:30]: (all) Memory leak: TwoWire::rxBuffer
[./hardware/libraries/Wire/Wire.h:35]: (all) Memory leak: TwoWire::txBuffer



What version of the Arduino software are you using? arduino-0015
On what operating system? Ubuntu Linux (Jaunty 9.04) 
Which Arduino board are you using? ---

Please provide any additional information below.

Is there an official todo/wish list for version 17 yet? If so can someone
please post a link. Maybe that should be a top level link somewhere on this
list.

Prioritizing what needed to be done might give potential contributors some
targets to aim at, and also indicate what projects had the blessing of the
administration.


Paul Badger

The MacOS Arduino version has similar issues to
http://code.google.com/p/arduino/issues/detail?id=4.

The RXTX library shipped with Arduino-16 is very old, and doesn't work on
64bit Macs ( http://www.arduino.cc/cgi-bin/yabb2/YaBB.pl?num=1246332304/1#1 ).

A work around is to check "Open in 32 Bit Mode" for Arduino, but updating
RXTX is a better option

The current librxtxSerial.jnilib only has i386 and ppc support and is from
2006.

You can grab a new one from http://iharder.sourceforge.net/current/java/ which
supports i386, amd64, ppc, and ppc64.

It looks like just including the new version will fix the 64bit intel
problem... I
only have a 32bit intel Mac, so I can't tell for sure.

What change would like to see?
Arduino icon fixed for mac client so there is no white border around the edge.

Why?
It is a minor visual issue which would make it appear in the dock neatly, 
perhapse even a redesign 
like Processing.

Would this cause any incompatibilities with previous versions?  If so, how
can these be mitigated?

None.

What steps will reproduce the problem?
Standard java dialogs (File-open,Print) are displayed in the local language 
while the rest of the IDE displays in english.

On what operatingsystem? 
Linux Windows

Fix locale to English on startup, or i18n the whole IDE

Eberhard

EmilyJane from the arduino forums noted that the 3.3V 328P 8MHz arduinos
were communicating at 2X the specified bit rate.

e.g.
Serial.begin(9600);
Serial.print("string");  // would be transmitted at 19200 bps on 3.3V 328P 8MHz

The bootloader for the 3.3V 328P 8MHz is created with the flag
"-DDOUBLE_SPEED" which sets the U2Xn bit in UCSRnA register.

Instead of overhauling the bootloader, HardwareSerial.h/.cpp has been
altered to explicitly set the U2Xn bit in UCSRnA register to 0.

New HardwareSerial.h and HardwareSerial.cpp files attached.

Can also be downloaded from:
http://mediafire.com/Arduino under "HardwareSerial u2x fix"

b

What steps will reproduce the problem?
1.Changed to the Windows build directory
2.Run ./make.sh
3.Extracting avr tools ...

What is the expected output? What do you see instead?
c:\cygwin/home/rhinerfeld/arduino/app/preproc/PdePreprocessor.java:32:8:32:21:
Semantic Error: You need to modify your
classpath,sourcepath,bootclasspath,and /or extdirs setup. Jikes could not
find package "processing.app" in:
..\build\windows\work\lib\RXTXxomm.jar
..\build\windows\work\lib\mrj.jar


What version of the Arduino software are you using? On what operating
system?  Which Arduino board are you using?


Please provide any additional information below.

In Arduino-0014 in Client.cpp from the Ethernet library there is an overflow 
issue with the _srcport 
in the connect() function. The _srcport is incremented on every connect without 
any limits. Since it's 
a uint16_t the assumption is that it would just overflow back to zero without 
any problems.

This would normally be okay, but the code then adds 1024 to the _srcport during 
the socket() call. 
This is correct to keep the port out of the privileged range. The problem is 
that once the _srcport 
exceeds 64111 the addition of 1024 will make the _srcport overflow back into 
the privileged port 
range.

There needs to be a check added after the _srcport++ to reset the value to 0 if 
it is > 54511.

(this "defect" is partially to try out the google defect tracking system
If a sketch folder includes a file "wiring_options.h", it will be added as
a "forced include" during the compilation of the wiring library files (ie
"-include <path>/wiring_include.h")  This would allow a sketch to override
default values of #defines in the wiring files (assuming that the files are
converted to allow such overrides, which is a separate issue.)  The most
obvious example would be the RX_BUFFER_SIZE used in the serial libraries.

(This patch is from the trunk circa Arduino14.  It may or may not still apply.)

Windows has alternate shortcuts for Cut Copy Paste as listed below.  Old
habits die hard.  It would be great if the editor could support these.

Ctrl-Insert = Copy
Shift-Delete = Cut
Shift-Insert = Paste

The firmware upload baud-rate should use the functionality in setbaud.h
http://www.nongnu.org/avr-libc/user-manual/group__util__setbaud.html

The firmware baud-rate calculation was changed for:
http://code.google.com/p/arduino/source/detail?r=581
but this is just a one-time fix. If you use the macros in util/setbaud.h,
you will not run into this same issue again for other upload speeds.

Also,
the firmware is using the old
  (F_CPU/(BAUD_RATE*16L)-1)
which doesn't approximate the baud rate as closely as
  (F_CPU/(BAUD_RATE*8L)-1)/2
or
  (((F_CPU)+8UL*(BAUD_RATE))/(16UL*(BAUD_RATE))-1)

which means that uploads will be corrupted more than they need to be.

From: http://www.arduino.cc/playground/Main/SuggestionsBugs?action=edit

Either implement a proper search on the forum, or just make the search box 
use google. Currently a lot of stuff just doesn't show up in search. OR, Move 
the forums to a platform that has better searching and anti-spam 
capabilities.

It had 10 votes.

Currently Arduino expects to be installed and run from a single directory
tree (eg: /home/foo/arduino-0015).  If Foo wants to add a new library, he
puts it in /home/foo/arduino-0015/lib; if he wants to save a sketch, he
saves it in /home/foo/arduino-0015/sketchbook.  If Foo upgrades to a newer
version of Arduino, it gets installed in a different directory (eg:
/home/foo/arduino-0016), and he has to copy all of his modifications from
'0015 to '0016, maybe missing something.

While this is ok (if clunky) in a single-user environment, in a multi-user
environment (particularly one where the software may be installed by an
installer or package manager) this makes it very inconvenient to package. 
If Arduino recognized the concept of home or working directories, in
addition to the normal installation (system) directory, it would make the
tool significantly easier to package for installation and use with multiple
users and workspaces.  It would also make it easier for someone to upgrade
to later versions of Arduino.

To make this work, Arduino would need to support multiple paths for key
directories (definitely examples, lib, sketchbook, and possibly hardware),
and in the event of duplicate files, choose the user's file instead of the
system file.  In this scenario, Arduino would be installed somewhere
appropriate to the system (C:\Program Files\Arduino on Windows,
/usr/share/arduino on Ubuntu, etc.) and there would also be a set of
directories for the user's files (C:\Users\Foo\AppData\arduino on Windows,
/home/foo/arduino on Unix/Linux/MacOSX).  When Arduino needs a file, it
looks in the user's directory first, then the system directory.

What change would like to see?


Please have the IDE check the linker output, and report an error if the
SRAM usage exceeds the size determined by the CPU selected in the
preferences menu. 


Why?

Because this is an annoyingly difficult bug to track down if you are not
expecting it.  In my case I got weird output on the Serial line, and
thought I had messed up my interrupt code!

Would this cause any incompatibilities with previous versions?  If so, how
can these be mitigated?


It shouldn't, the only issues would be for people using the IDE to generate
files for odd ball AVR chips.  eg. using the settings for a 168 to program
an at90, or an atmega48 say...

What steps will reproduce the problem?
1. Open Arduino IDE
2. Open Arduino->Preferences (OS X) or File->Preferences (Win)
3. Try to select path to preferences.txt at bottom of window - not possible


What version of the product are you using? On what operating system?
Arduino 12 on Win, Mac, Arduino 13 from SVN sources on Win, Mac

Please provide any additional information below.

It would be great to be able to select the preference path to paste into
Explorer/Terminal/... Since you cannot modify preferences.txt while Arduino
is running, users have to manually write down the path or remember it (it's
long and uncommon) to browse to the file after closing the IDE. Allowing
users to place copy the path onto the clipboard would be much nicer.

What change would like to see?
An official "Arduino logo"

Why?
Cos I want it upon a cool.T.shirt on a wet hot sunny day

Would this cause any incompatibilities with previous versions?  If so, how
can these be mitigated?

ATMO thee is no previous version, maybe we need a design competition.

Best man wins and get a can of beer, 3 ards, 2 resistoers and a dark UV led ;-)

The design competion idea tickles me, and we can use that for some daft
reason ;-)

What steps will reproduce the problem?
1. Selecting ADC8 in a sketch results in the ADC0 output being selected due
a bit masking function in line 45 of
\\hardware\cores\arduino\wiring_analog.c

2. Change
ADMUX = (analog_reference << 6) | (pin & 0x07);
to
ADMUX = (analog_reference << 6) | (pin & 0x0f);

3. The original code was correct when AT chips only used ADC channels 0 to
7, but this masks out the new ADC8 internal temperature measurement channel.

What is the expected output? What do you see instead?
ADC8 output is expected; ADC0 output is provided

What version of the Arduino software are you using? V0015
On what operating system?  Windows; but the code is the same in Linux and
no doubt Mac.
Which Arduino board are you using? Diecimilla/BareBones Board.


Please provide any additional information below.
Temperature Sensing is available on ATMega328 (confirmed) and newer
production lots of its siblings (unconfirmed)
Refer to Arduino Forum › Software › Development › Internal Temperature
Measurement ATMega328.