Posted: December 1st, 2008
Right now I’m still sleep deprived having only slept 3h30 – something that happened by accident. I’m also a bit angry because my entry got arbitrarily bumped out of 3rd place by the competition judges. That sucks. Not just for the cash (100 eur) but mostly because it feels like a slap to face. I worked for 1 week non-stop (close to 20h/day for 5 days) and came all the way to Barcelona. I watched with excitement as my agent did better than most in the competition even if it was partially luck (as was for the 2 teams with better scores). And after all the sacrifice, tears and joy I don’t get to go to the final because of an arbitrary decision of a group of people. FUCK THAT. That’s why I dislike the existence of judges in a competition such as this. “Fixing” scores is cheating – regardless of weather it’s done behind the scenes or in plain view. Sigh. I guess it was too much for the final round to be 2 Portuguese teams, 1 Italian and none from the USA.
And worse, both of the entries from FEUP were from undergrads who only had 1 week to work on this while many of the other teams were from people who had masters and phds and the italian guy that got 2nd place was an associate professor and had 1 month to work on it (and said he thought it wasn’t enough in his presentation).
Oh well. At least the team that won was from FEUP (Team SpeedyGonzales) and I’m happy for them (congratulations again Alvaro, Fabio and Sara). The other team from FEUP (NetSqueak) sorta gave up so they got merged into my team and presented their results during the first part of my team’s presentation even though their agent wasn’t running in the competition.
My presentation sucked! I fell a sleep and didn’t wake up till 5 min after the competition officially started. I managed to arrive on time and do a few slides but I was way too tired to make a decent presentation.
Since I only had one week to prepare for this (from scratch) that meant 1 day to write the paper, 5 days to write code and 1 day to travel. Unfortunately my initial approach which consisted of using neural networks for motor control was a huge fail and I wasted a day with it. Still, in 4 days I managed to get probabilistic mapping, basic communication, and a bunch of behaviors (though many buggy) written. Apart from what I learned about cibermouse and robotics there’s also the experience of having an even more insane deadline than usual. It’s very much like a real time system: if you cant do it in this amount of time, it’s a fail. You have to give up on certain solutions if they don’t work because you can’t predict how much time it’s going to take you to get it to work while it’s a lot easier to estimate how long it’s going to take to implement a different solution (if you’ve done something similar before).
Honestly, a few minutes before the competition I thought I was going to be the worst entry there. There were so many bugs that remained unfixed that I had to disable a lot of my mouse’s intelligence – hell, during the first run, 2 of my agents crashed and I had to further cripple it by disabling the comunication system for the second run or risk losing points again (crashed agents collide with walls). Most of the entries had no intelligence anyway – at least not in practice. I was expecting to see some really intelligent agents but instead, the ones that did better where the ones closer to plain simple reactive. In fact, the robot that won the thing for SpeedyGonzales was a purely reactive agent (they had two wall followers, two explores and 1 simple reactive agent). Building intelligence into robots is a lot more complicated than I believed it to be. Also why the hell did so few of the entries bother to actually use the damn sensors properly to avoid collisions? Makes no sense.
I spent the past few hours at the hardrock cafe which is pretty much next door (like 30s walk) to the hotel I’m staying at. I liked it, the only other I had been to was the one in Oslo and that one was a HUGE FAIL – when I went there, it didn’t have Rock (they were wacthing football) or “Cafe” (coffee), the machine was broken or something. I was told the one in Lisbon was Bigger & Better (TM) than this one so I’ll have to check it out next time I’m in Lisbon. In the meantime I’m tempted to purchase a tshirt or something.
For the next 2 days I’ll get to see Barcelona (so far I’ve been stuck at the hotel working). So I
Posted: November 14th, 2008
The first project I’ve made (well technically second since it’s a support lib for another project) is now available (via SVN) at google code – irnotify.
A python library for making notifications. Currently implements notifications via XMPP (Jabber), SMTP (Mail) and Twitter direct messages.
Here’s the example code:
from twitternotify import TwitterNotify
n = TwitterNotify("twitter.conf")
n.notify("lrei", "How do you feel, Rei?")
from xmppnotify import XMPPNotify
n = XMPPNotify("xmpp.conf")
n.notify("luis.rei@gmail.com", "How do you feel, Rei?")
Posted: October 1st, 2008
How Digg Works is a nice blog post on the digg technology blog that gives an overview of the digg network architecture: load balancers -> application server (apache, memcached, gearman) -> database server (MySQL) / file server (MogileFS). Quoting:
“Digg uses Debian GNU/Linux across the board with a mixture of MySQL, Memcached, MogileFS, Python, PHP, Apache, Gearman and various appliances to serve up billions of requests a month (and more every day!)”
What’s New In Python 2.6. My highlights: new documentation format using Sphinx (which btw looks incredibly powerful – read “complex”), awesome (but still optional) with statment (from future) to replace try…finally blocks, multiprocessing package,and more.
I had more stuff to put in this post but it has been in my “draft” folder for more than a week now so I’m just gonna post it like it is.
Posted: May 11th, 2008
friendsnippets.com is for code what del.icio.us is for bookmarks. Just found it and fell in love with it.
link to my account.
Posted: May 11th, 2008
My “Computer Networks” (Redes de Computadores) assignment is writing an HTTP server that implements GET and conditional (if-modified-since) GET with persistent connections.
Here’s the code – Obviously, no guarantees.
Posted: May 10th, 2008
This one instead reads the values from a CSV file containing experimental data.
Link to my previous implementation of Simspson’s rule.
import csv
class simpsonCsv:
def __init__(self, filename):
reader = csv.reader(open(filename, "rb"))
xlist = []
ylist = []
for row in reader:
try:
x = float(row[0])
y = float(row[1])
except TypeError:
continue
except ValueError:
continue
xlist.append(float(x))
ylist.append(float(y))
self.len = len(xlist) - 1
self.xstart = xlist[0]
self.xend = xlist[self.len]
self.interval = self.xend / self.len
self.data = map(None, xlist, ylist)
def f(self, x):
for (u,v) in self.data:
if u == x:
return v
print "Bad x value (probably bad n): %s" % (x)
return None
def simpson(self, n):
"Approximate the definite integral of f from a to b by Simpson's rule."
if self.len % n != 0:
print "Error: %d mod %d is not zero but should be." % (self.len, n)
return -1
h = float(self.xend - self.xstart)/n
si = 0.0
sp = 0.0
xk = 0.0
for i in range(1, n, 2):
xk = self.xstart + i*h
si += self.f(xk)
for i in range(2, n, 2):
xk = self.xstart + i*h
sp += self.f(xk)
s = 2*sp + 4*si + self.f(self.xstart) + self.f(self.xend)
return (h/3)*s
filename = "integral_tabela_CO2_csv.csv"
s = simpsonCsv(filename)
print s.simpson(20)
print s.simpson(40)
print s.simpson(60)
print s.simpson(100)
print s.simpson(300)
print s.simpson(600)
print s.simpson(900)
print s.simpson(1800)
Posted: April 11th, 2008
def simpson(f, a, b, n):
"Approximate the definite integral of f from a to b by Simpson's rule."
if n % 2 != 0:
print "Ups: n must be even!"
return -1
h = (float(b) - a)/n
si = 0.0
sp = 0.0
for i in range(1, n, 2):
xk = a + i*h
si += f(xk)
for i in range(2, n, 2):
xk = a + i*h
sp += f(xk)
s = 2*sp + 4*si + f(a) + f(b)
return (h/3)*s
def f(x):
return x**4
ni = 50
nf = 1000000
n = ni
a = -20
b = 0
s = []
qc = []
ec = []
t = 0.0
div = 0.0
i = 0.0
while n < nf:
t = simpson(f, a, b, n)
s.append(t)
n = n * 2
for i in range(0, len(s)-2):
div = (s[i+2] - s[i+1])
if div == 0:
break
t = (s[i+1] - s[i]) / div
qc.append(t)
t = div / 15
ec.append(t)
for i in range(0, len(qc)):
print "%.12f, %.12f, %.12f => qc=%.12f, e=%.12f" % (s[i], s[i+1], s[i+2], qc[i], ec[i])
Posted: April 4th, 2008
# this requires numpy get it from http://numpy.sf.net
from copy import deepcopy
from numpy import *
# this function, swapRows, was adapted from
# Numerical Methods Engineering with Python, Jean Kiusalaas
def swapRows(v,i,j):
"""Swaps rows i and j of vector or matrix [v]."""
if len(v) == 1:
v[i],v[j] = v[j],v[i]
else:
temp = v[i].copy()
v[i] = v[j]
v[j] = temp
def pivoting(a, b):
"""changes matrix A by pivoting"""
n = len(b)
for k in range(0, n-1):
p = int(argmax(abs(a[k:n, k]))) + k
if (p != k):
swapRows(b, k, p)
swapRows(a,k,p)
def gauss(a, b, t=1.0e-9, verbose=False):
""" Solves [a|b] by gauss elimination"""
n = len(b)
# make copies of a and b so as not to change the values in the arguments
tempa = deepcopy(a)
tempb = deepcopy(b)
# check if matrix is singular
if abs(linalg.det(tempa)) < t:
return -1
pivoting(tempa, tempb)
for k in range(0,n-1):
for i in range(k+1, n):
if tempa[i,k] != 0.0:
m = tempa[i,k]/tempa[k,k]
if verbose:
print "m =", m
tempa[i,k+1:n] = tempa[i,k+1:n] - m * tempa[k,k+1:n]
tempb[i] = tempb[i] - m * tempb[k]
# Back substitution
for k in range(n-1,-1,-1):
tempb[k] = (tempb[k] - dot(tempa[k,k+1:n], tempb[k+1:n]))/tempa[k,k]
return tempb
def residue(a, b, c):
"""Calculates the residue of a system solved by gauss elimination"""
n = len(b)
t = a * c # t is the A with the values of x replaced (an [n x n] matrix)
s = []
for i in range(0, n):
s.append(sum(t[i])) # s is the solution
res = b - s # res is the residue
return res
#a = array([[1.0, 2.0, 0.0],[-1.0, 2.0, 3.0],[1.0, 4.0, 1.0]])
#b = array([3.0, -1.0, 4.0])
#a = array([[-1.414214, 2, 0],[1, -1.414214, 1], [0, 2, -1.414214]])
#b = array([1.0,1.0,1.0])
#a = array([[2.0, 2.0, 2.0],[1.0, 1.0, 5.0], [2.0, 5.0, 1.0]])
#b = array([6.0, 7.0, 8.0])
a = array([[1.001, 2.001, 3.001],[0.999, 2.0, 2.999], [1.002, 1.999, 2.999]])
b = array([4.003, 4.001, 3.999])
x = gauss(a, b)
print "Solution = ", x
#sol = linalg.solve(a, b)
#print "linalg Solution = ", sol
y = residue(a, b, x)
print "Residue = ", y
u = gauss(a, y)
print "Residue destribution = ", u
z = gauss(a, b+y)
print "New Solution (with added residue) = ", z
y2 = residue(a, b+y, z)
print "Residule of new solution = ", y2
if linalg.norm(y2) < linalg.norm(y):
print "New solution has a smaller residue."
else:
print "Original solution has a smaller residue."
Posted: March 31st, 2008
JabberLogBot is a jabber bot that records messages sent to it in a database.
The idea came from Nuno Dantas who wanted a jabber bot to record quick notes. He talked about it at a Prt.Sc dinner last Wennesday.
There’s also a simple PHP file in there that displays the data. That file is currently just for show as it is rather “plain”.
UPDATED: added a screenshot of the web viewer.
Posted: March 28th, 2008
The second beta version of the iPhone SDK is now available and includes Interface Builder, a powerful tool that allows you to visually build your interface and makes creating a UI as simple as drag and drop.
Screenshots:
Download at iPhone Dev Center
Posted: March 27th, 2008
I finally decided to take the Numerical Methods (MNUM) course. It turns out it’s a lot more fun than I thought. There is programming involved but you can chose to use whatever language you want. This is yet another nice excuse for me to use Python instead of C++ or Java. Last semester I was able to use Python to implement the game logic for Software Application Laboratory (LAS), which is mostly an OpenGL course with IPC via sockets thrown into the mix, and to write an article on dynamic languages (focusing mostly on Python) for Software Engineering (ESOF).
But back to this semester, 3 classes into the semester and the teacher is already said something like “I’m going to learn python now. I didn’t believe when I heard someone saying it was the best language in the world, but now I see there might be some truth to that claim”. That and I suspect his next laptop might be a macbook but that’s another story.
There are a few things that make Python great for Numerical Methods. In my opinion, Python’s clear, easy to understand, syntax is the most important one.It makes algorithms easier to implement. The syntax ends up being very close to language neutral pseudocode available in numerical methods books. Also Python’s datatypes as well as those provided by other libraries can be very useful.
The following code implements the stuff in chapter 2 (determining zeros) of the course. The methods implemented are Bisection, Rope and Newton. The function returns both the solution and the number of iterations necessary to get to that solution.
UPDATE: forgot the book - Numerical Methods in Engineering with Python
Appendix A – mnum2.py
from math import log
def bisect(f, a, b, e):
""" Determines zero between a and b using Bisection. """
n = 0
fa = f(a)
if fa == 0.0: return (a, n)
fb = f(b)
if fb == 0.0: return (b, n)
while (abs(a-b) > e):
c = 0.5*(a+b)
fc = f(c)
if fc == 0.0: return (c, n)
n = n + 1
if fb*fc < 0.0:
a = c
fa = fc
else:
b = c
fb = fc
if fa < fb:
return (a, n)
else:
return (b, n)
def rope(f, a, b, e):
""" Determines zero between a and b using the Rope methode. """
n = 0
fa = f(a)
if fa == 0.0: return (a, n)
fb = f(b)
if fb == 0.0: return (b, n)
while (abs(a-b) > e):
c = (a*fb - b*fa) / (fb - fa)
fc = f(c)
if fc == 0.0: return (c, n)
n = n + 1
if fb*fc < 0:
a = c
fa = fc
else:
b = c
fb = fc
if fa < fb:
return (a, n)
else:
return (b, n)
# Note: must verify that for the function f and guess c
# the method will _converge_.
def newton(f, df, c, t):
""" Determines zero between a and b using Newton """
n = 0
fc = f(c)
if fc == 0.0: return (c, n)
while (True):
fc = f(c)
dfc = df(c)
if dfc == 0:
print "dfc is 0"
return (0, -1)
dc = -fc/dfc
c = c + dc
n = n + 1
if abs(dc) < t: return (c, n)
##Tests
#def f(x): return -log(x)+4.0
#def df(x): return -1.0/x
#x= bisect(f, 1, 70, 0.00000001)
#print x
#x = rope(f, 1, 70, 0.00000001)
#print x
#x = newton(f, df, 0.1, 0.0001)
#print x
Posted: March 26th, 2008
Ok, from the comments on the previous post it seems people are NOT “getting it”. This is NOT a “problem”, this is funny.
So I guess people don’t know what the functions min() and max() do or are just confusing them
max([...]) - this function usually takes a list of numbers and returns the highest number in the list.
e.g.
if x belongs to [0,20] then y = max(x, 18) => y belongs to [18-20]
max(20,18) = 20
max(19, 18) = 19
max(10, 18) = 18
min([...]) - this function usually takes a list of numbers and returns the lowest number in the list.
e.g.
if x belongs to [0,20] then y = min(x, 18) => y belongs to [0-18]
min(20,18) = 18 [<- fixed thanks Mind Booster Noori]
min(19, 18) = 18
min(10, 18) = 10
This means that according to the function the function displayed in my previous post, according to that function, ALL students grades will be between 18 and 20. That means if you have a freaking 0 in the exam and a 0 in the Assignment you’ll get an 18/20 (which is an excellent grade). i.e. THE MINIMUM grade is 18. It’s a simple mistake in the equation. What the page should say to be correct is
Nota Final = min(Nota do exame final+Nota do trabalho, 18)<- this would be the correct equation
And hell this was posted under “Entertainment”.
So the previous post wasn’t supposed to be about something “unfair” it was supposed to be about making fun of a simple mistake that under the circumstances really is funny.
Posted: March 25th, 2008
I just got an email (via the faculty-wide spam network) about the ACM International Collegiate Programming Contest (ICP) or more precisely about SWERC 08. For a few moments I thought “hey, this might be fun”. A second later I read something like
“Languages allowed: C, C++, Java or Pascal”
Yay it’s 1995 again!!!
Pascal? Didn’t Pascal die like more than a decade ago?
C++ and Java are so 90′s… where are the cool languages? Python? Ruby? Heck even Erlang and Haskell?
Considering the “retro” theme, I would’ve was expecting Lisp and Fortran to be in there. (Not (that I have anything against (lisp))).
So basically the point of the contest is seeing how well you do with a handicap? Kinda like the Olympic 100m with a broken leg (or 2 if you’re using C)?
Sounds very interesting but I rather go cut myself with the portuguese tokio hotel fans (via Tiago Farrajota)…
Posted: March 24th, 2008
I mentioned in a previous post that I was using a private WordPress blog to keep my notes. Not anymore. I migrated to Evernote.
Thanks to Maria Joao Valente for sending me the invite to evernote.
Evernote is a note organizers, similar to Journler which I used a while back.
Check out the About Evernote and their screencast. My highlights:
* Web client
* Desktop client
* Works with Mobile Devices
* Painless, automatic synchronization (think gmail + IMAP but better)
* Notes can be found by searching and filtering for text within images
* Clip (via bookmarklet) or email entire webpages into your account
* Can import html files (you’ll see why this was important for me)
See also: Wired Review and TUAW Review.
Migrating between applications has never been an easy task. In this case I need to migrate from a WordPress blog to evernote. I could manually click “Clip to Evernote” for each post on that blog or I could’ve written a simple AppleScript to do it or I could probably have found a way to do it in Javascript or I could’ve taken advantage of the “clip” thing in another way. But off course I choose the hardest way possible – I wrote a python script to convert the WordPress XML Export File to multiple HTML notes and then dragged those files to evernote. At least it was fun if a colossal waste of time…
Anyway here’s the python script in case you ever want to convert a wordpress blog (or more accurately a WordPress XML Export File) to html files.
wpdepress.py
# Copyright (c) 2008 Luis Rei
#
# Permission is hereby granted, free of charge, to any person obtaining a copy
# of this software and associated documentation files (the "Software"), to deal
# in the Software without restriction, including without limitation the rights
# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
# copies of the Software, and to permit persons to whom the Software is
# furnished to do so, subject to the following conditions:
#
# The above copyright notice and this permission notice shall be included in
# all copies or substantial portions of the Software.
#
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
# THE SOFTWARE.
# Notes:
# - currently does not handle images, attachments or comments
# - was only tested on MacOS X (10.5)
# - not "carefully" developed e.g. poor exception handling, little testing, ...
# - see also http://wordpress.com/blog/2006/06/12/xml-import-export/
import string, os, sys, getopt
from xml.dom import minidom
__author__ = 'Luis Rei (luis.rei@gmail.com)'
__homepage__ = 'http://luisrei.com'
__version__ = '1.0'
__date__ = '2008/03/23'
def convert(infile, outdir, authorDirs, categoryDirs):
"""Convert WordPress Export File to multiple html files.
Keyword arguments:
infile -- the location of the WordPress Export File
outdir -- the directory where the files will be created
authorDirs -- if true, create different directories for each author
categoryDirs -- if true, create directories for each category
"""
# First we parse the XML file into a list of posts.
# Each post is a dictionary
dom = minidom.parse(infile)
blog = [] # list that will contain all posts
for node in dom.getElementsByTagName('item'):
post = dict()
post["title"] = node.getElementsByTagName('title')[0].firstChild.data
post["date"] = node.getElementsByTagName('pubDate')[0].firstChild.data
post["author"] = node.getElementsByTagName(
'dc:creator')[0].firstChild.data
post["id"] = node.getElementsByTagName('wp:post_id')[0].firstChild.data
if node.getElementsByTagName('content:encoded')[0].firstChild != None:
post["text"] = node.getElementsByTagName(
'content:encoded')[0].firstChild.data
else:
post["text"] = ""
# wp:attachment_url could be use to download attachments
# Get the categories
tempCategories = []
for subnode in node.getElementsByTagName('category'):
tempCategories.append(subnode.getAttribute('nicename'))
categories = [x for x in tempCategories if x != '']
post["categories"] = categories
# Add post to the list of all posts
blog.append(post)
# Then we create the directories and HTML files from the list of posts.
# The "base" directory
outdir += "/wordpress/"
if os.path.exists(outdir) == False:
os.makedirs(outdir)
os.chdir(outdir)
for post in blog:
# The "category" directories
path = ""
if authorDirs == True:
path += post["author"].encode('utf-8') + "/"
# This creates a path for the file in the format
# category1/category2/category3/file. Note that the category list was
# sorted.
if categoryDirs == True:
if (post["categories"] != None):
path += string.join(post["categories"],"/")
if os.path.exists(path) == False and path != "":
os.makedirs(path)
# And finally the file itself
path = outdir + path
title = post["title"].encode('utf-8')
filename = path + "/" + post["id"] + ' - ' + title \
+ '.html'
# Add a meta tag to specify charset (UTF-8) in the HTML file
meta = """"""
f = open(filename, 'w')
f.write(meta+"\n")
# Add "HTML header"
start = "\n\n\n\n\n"
f.write(start)
# Convert the unicode object to a string that can be written to a file
# with the proper encoding (UTF-8)
text = post["text"].encode('utf-8')
# Replace simple newlines with
+ newline so that the HTML file
# represents the original post more accuratelly
text = text.replace("\n", "
\n")
f.write(text)
# Finalize HTML
end = "\n\n"
f.write(end)
f.close()
def usage(pname):
"""Displays usage information
keyword arguments:
pname -- program name (e.g. obtained as argv[0])
"""
print """python %s [-hac] [-o outdir] infile
Converts a WordPress Export File to multiple html files.
Options:
-h,--help\tDisplays this information.
-a,--authors\tCreate different directories for each author.
-c,--categories\tCreate directory structure from post categories.
-o,--outdir\tSpecify a directory for the output.
Example:
python %s -c -o ~/TEMP ~/wordpress.2008-03-20.xml
""" % (pname, pname)
def main(argv):
outdir = ""
authors = False
categories = False
try:
opts, args = getopt.getopt(
argv[1:], "ha:o:c", ["help", "authors", "outdir", "categories"])
except getopt.GetoptError, err:
print str(err)
usage(argv[0])
sys.exit(2)
for opt, arg in opts:
if opt in ("-h", "--help"):
usage(argv[0])
sys.exit()
elif opt in ("-a", "--authors"):
authors = True
elif opt in ("-c", "--categories"):
categories = True
elif opt in ("-o", "--outdir"):
outdir = arg
infile = "".join(args)
if infile == "":
print "Error: Missing Argument: missing wordpress export file."
usage(argv[0])
sys.exit(3)
if outdir == "":
# Use the current directory
outdir = os.getcwd()
convert(infile, outdir, authors, categories)
if __name__ == "__main__":
main(sys.argv)
Posted: March 7th, 2008
It’s been months since I was insulted by Apple fanboys for saying that the iPhone needed an SDK and now Apple has officially released a beta version of the iPhone/iPod Touch SDK.
I can’t access developer.apple.com at the moment but Levi Figueira was kind enough to upload both screenshots of the pages and the PDFs of agreements involved during a twitter discussion with me and David Rodrigues (also read his post on the iPhone SDK in Portuguese).
The good news is apple provides the App Store to distribute your applications with the iPhone Developer Program and all you have to do is pay them $99 and 30% of what your customers pay for your app.
The bad news is that, from what I understood, there’s no other way to get apps on an iPhone, even if it’s your own apps on your own iPhone or if they’re Open Source apps.
The other good news is that if they’re free apps, they don’t charge the 30% (30% of 0 is 0 so…). So as long as someone paid $99 to be part of the Developer Program, they can put free apps on the Store at no additional charges.
The other bad news is there won’t be porn games on the iPhone 
I wonder if they consider hentai (Japanese eroge visual novel) games “porn”.
I think the iPhone Developer Program is a really Great Thing (TM), think Steam for the iPhone/iPod Touch but more open. I just wish it were “optional”.
