Barriers are in place along the River Severn as the Environment Agency says water levels will peak on Saturday…. From BBC News. Full story

This site may contain information about: water depths. The blog is also related to: ocean.

In my last post, I used barriers and shared channels to write a simple program for performing force-based graph layout. There were two things lacking in that example: firstly, repulsion between unconnected nodes and secondly, some cool videos. I’ve now solved both of these shortcomings.

In order to implement repulsion between all nodes, each node must know the position of all other nodes at every time-step. I could implement this with message-passing, but that would be O(N^2) communications each phase, which isn’t very efficient, and at that point, message-passing is probably not the right design. We need some shared variables instead. Software Transactional Memory (STM) provides a nice way to do shared variable concurrency in Haskell, so we will use that.

The idea is that we will have a transactional variable (TVar) per node, holding the node’s position. Every time-step, each node will read from all the TVars to discover the positions of all the other nodes, and will then update its own position based on the other nodes. If we are not careful, we will have the same race hazard that I warned against last time: one node could update its position before all the other nodes have read its previous position. Or, if left totally unconstrained, one node could perform several updates before other nodes have even performed one. To prevent these problems, we will again use barriers to break up the simulation into two phases: discover, and act. This means that we will be combining CHP and STM: using barriers from CHP to regulate access to shared variables from STM.

All the code before the node process is exactly the same as last time, except that I now import Control.Concurrent.STM and Control.Parallel.Strategies (for some strict application) — and I’ve made the nodes a bit smaller. Our node process no longer takes channels as its parameters, but TVars instead. It takes a list corresponding to the other nodes, of pairs of booleans (indicating if the other node is connected to this one) and transactional variables (to read the other node’s position). It also takes a single transactional variable, which it should use to write its own position to:

node :: NodeInfo -> [(Bool, TVar NodeInfo)] -> TVar NodeInfo
     -> Enrolled PhasedBarrier Phase -> CHP ()
node start neighbours me bar = node' start
  where
    (connected, tvs) = unzip neighbours

    node' cur
      = do Discover <- syncBarrier bar
           neighbourPos <- liftSTM (mapM readTVar tvs)
           let new = updatePos (zip connected neighbourPos) cur
           Act <- syncBarrier bar
           liftSTM (writeTVar me (new `using` rnf))
           node' new

You can see that the node process is a bit simpler. In the Discover phase it reads from the TVars and in the Act phase it writes to its own. It is immediately apparent that the reads and writes are thus segregated. liftSTM is defined below (it is essentially the atomically function), and the `using` rnf is a strictness thing that you can ignore. The updatePos function is now more complicated as we have added repulsion (though it is all irrelevant to the concurrent logic):

    updatePos poss cur
      = sum [cur
            ,0.05 * sum [ let v = p - cur in v - ideal * normalise v
                        | (True, p) <- poss] -- connected nodes
            ,0.000002 * sum [let sc = (p `from` cur) ^^ (-2)
                             in NodeInfo sc sc * normalise (cur - p)
                            | (_, p) <- poss] -- all nodes
            ]
      where
        ideal = 0.02
        (NodeInfo x2 y2) `from` (NodeInfo x1 y1)
          = sqrt $ (x2-x1)^^2 + (y2-y1)^^2
        normalise (NodeInfo x y) = fmapNodeInfo (/ mag) (NodeInfo x y)
          where
            mag = sqrt (x*x + y*y)

instance NFData GLfloat
instance NFData NodeInfo where
  rnf (NodeInfo a b) = rnf a >| rnf b

liftSTM :: MonadIO m => STM a -> m a
liftSTM = liftIO . atomically

I’ve also had to add an instance of NFData to use rnf (strict evaluation) on NodeInfo, and the simple liftSTM function for executing STM transactions in the CHP monad. The draw process only has one change; instead of reading from a list of channels, we now read from a list of transactional variables:

drawProcess :: [TVar NodeInfo] -> Enrolled PhasedBarrier Phase -> CHP ()
drawProcess tvs bar
 = do displayIO <- embedCHP_ $ do syncAndWaitForPhase Discover bar
                                  xs <- liftSTM $ mapM readTVar tvs
                                  liftIO $ do startFrame
                                              mapM_ draw xs
                                              mapM_ (drawEdge xs) edges
                                              endFrame
...

Last time, I only had four nodes in my example, which was fairly boring. This time I have one hundred nodes, connected in a 10×10 grid. To make the demo interesting, all the nodes start with random positions:

startNodes :: IO [NodeInfo]
startNodes = replicateM (10*10) $ liftM2 NodeInfo r r
  where
    r = liftM (fromRational . toRational) $ randomRIO (0 :: Float, 1)

edges :: [(Int, Int)]
edges = across ++ down
  where
    across = [(x+10*y, (x+1)+10*y) | x <- [0..8], y <- [0..9]]
    down = [(x+10*y, x+10*(y+1)) | x <- [0..9], y <- [0..8]]

Finally, our slightly adjusted main process:

main :: IO ()
main = runCHP_ $
       do nodes <- liftIO startNodes
          vars <- liftSTM $ mapM newTVar nodes
          enrollAll_ (newPhasedBarrier Act)
           (drawProcess vars :
            [ let edgesOut = filter ((== i) . fst) edges
                  edgesIn = filter ((== i) . snd) edges
                  connectedNodes = map fst edgesIn ++ map snd edgesOut
                  conn = [(ind `elem` connectedNodes, tv)
                         | (tv, ind) <- zip vars [0..], tv /= mytv]
              in node n conn mytv
            | (n, mytv, i) <- zip3 nodes vars [0..]
            ])

There are not many changes from my previous shared channel-based version to use STM instead. My continued use of barriers alongside STM shows that you can mix STM and CHP quite nicely if you want to. Now, here’s what you’ve been waiting for — a video of the graph layout in action, first on my grid graph as defined above:

Choose from WMV format (suitable for Windows) or MPEG format (suitable for Linux, etc).

I also ran the algorithm on a simple loop graph:

Choose from WMV format (suitable for Windows) or MPEG format (suitable for Linux, etc).

The two videos have the algorithm running at 10 iterations per second, as that seems like a nice speed to view them at (and apologies for the slight garbage down the right-hand side of the videos). The grid example was inspired by an original example in the Java version of the Prefuse toolkit. Prefuse is an incredibly cool visualisation toolkit; some Java examples can be found online but more impressive are the more recent ActionScript examples. Try the Layouts tab, and also the dependency graph example if you have a spare minute.

Today’s post is a little different.  It looks at what things may mean, aside from their physical presence.  An earlier version first appeared as an emailed comment on October 21, 2009 to JStreet.org, an emerging interesting group noted for expanding the Middle Eastern conversation.  Here’s the item at JStreet/Settlements that got my attention…

Security Barrier & Two-state Sustainability:  “J Street supports the concept of a security barrier as an important element of Israel’s defense, but believes that the barrier must be located along an internationally recognized border.  Its present route has confiscated land and separated Palestinians from the jobs, health care and family. It will have to be relocated in many sections as part of a final status agreement.”

So let’s look at that for a moment.  Isn’t it possible that it’s exactly “the concept of a security barrier” that makes a barricaded border into a mutual membrane that filters out the directly accessible economic and individual interactions that define viable social and political life?  Can the term “barrier” end up turning the meaning of “security” upside down?

It’s not like Europe hasn’t torn itself apart twice in the past century, and numerous times prior to that (see History_of_Europe).  Yet, consider that the 1930-1939 barrier attempted between two European antagonists has become the signature term for a futile defensive crouch.  For details, look at the Maginot_Line polity and policy.  Just as the Line tried to substitute concrete for wisdom, mobile forces, and firepower, so an ongoing barrier says that an ongoing relationship will be grounded in a physical manifestation of an Existential disbelief in any shared positive reality between the people on the ground.

In relative terms, the Berlin_Wall outlived the Maginot Line by decades, but it didn’t buy durable security for the German Democratic Republic.  The poet Robert Frost wrote that “good fences make good neighbors” (see Mending_Wall), but it seems also true that solid walls make prisoners on both sides.  At least a fence lets air, light, and sight pass through, and provides a place to rest one’s forearms while discussing the day with a neighbor.  And unlike a solid wall or barrier, a simple fence is a mutual respect for a shared larger space.  That’s a security no barrier can provide.

Once the idea of “barrier” digs into the mind, other examples come into focus.  There’s Hadrian’s_Wall to keep Romans safe in Britannia.  There’s the Great_Wall_of_China  to keep out invaders from the 5th Century BC to the 16th Century Current Era.  Each kept people busy and delayed incursions, but their need for continuous work indicates neither succeeded in staving off depredations as the respective centers lost their way.  Makes one wonder about the ability to process lessons learned.  See current US efforts to paste walls of steel (South) and procedure (North) to cover policy shortfalls affecting both borders (U.S.-border-policy-question, United_States_border).

So there’s something seductive about a wall.  People keep building them to feel good by doing something, anything, to avoid doing something more insightful.  Their record says it’s more likely walls end up looking like a lunge for a security blanket, than like real clarity of mind.  Maybe simple fences to mark shared spaces aren’t so simple-minded after all.

A graph is a collection of nodes, and edges joining the nodes together. Automatically drawing them cleanly is an interesting problem. Some force-based graph layout algorithms view the edges between nodes as springs, and simulate the forces acting on each node in order to move the nodes into a good position. Connected nodes will pull towards each other until the edges are of an ideal length. We can implement such a layout algorithm in CHP, with a process per node.

To implement the algorithm, the nodes need to be able to find out the current positions of their neighbours (i.e. the nodes they are connected to) and update their own position accordingly. One approach would be to have a channel pair per edge, to enable sending of position information in both directions, as in this graph of three nodes:

I’m going to take an alternative approach of having one output channel per node, on which the node can send its position. The other end (i.e. the reading end) of these position channels will be shared, and these ends will be passed around to all the nodes that connect to the node with the output end. We can also give all the reading ends to a display process. So our wired up graph now looks like this:

A shared channel is represented by a small hollow circle. Each has one writer (the three nodes), and several readers (the connected nodes and the display process). Each iteration, our nodes will offer to send out their current position (as many times as they are asked for it) while also fetching the position of all their neighbours. Then they will all calculate a new position based on their neighbours and go again. One problem with this common discover-then-act design is that if you do not clearly separate the discovery of the neighbours’ positions and the updating of the positions, you can get nodes updating based on a mix of old positions (which is what you want) and the new updated positions — a race hazard. To prevent this, we divide each simulation step into two phases (discover and act) using a phased barrier.

A phased barrier is a synchronisation primitive. It allows processes to enroll on the barrier, to resign from the barrier, and to synchronise; processes only successfully synchronise on a barrier when all currently-enrolled processes synchronise. Each synchronisation, the phase of the barrier is moved on (and typically cycles around).

We will begin with some import statements, and declaring a NodeInfo type to hold the positions of nodes. We will also include a quick Num and Fractional instance for our NodeInfo that performs plus, minus, etc element-wise (NodeInfo 1 6 * NodeInfo 3 4 == NodeInfo 3 24):

import Control.Concurrent.CHP
import Control.Monad
import Control.Monad.Trans
import Graphics.Rendering.OpenGL
import Graphics.UI.GLUT hiding (alt)

data NodeInfo = NodeInfo GLfloat GLfloat deriving (Show, Eq)

instance Num NodeInfo where ...
instance Fractional NodeInfo where ...

Then we will declare our phase data type, and a helper function to read from a list of shared channels in parallel:

data Phase = Discover | Act deriving (Eq, Show, Bounded, Ord, Enum)

readAll :: [Shared Chanin a] -> CHP [a]
readAll = runParMapM (flip claim readChannel)

Next, we will define our node process. The main body of the node process first begins the discover phase. It then acts as a sender and receiver in parallel: the receiver reads in the positions of all its neighbours, while the sender continually offers to send out its position. It finishes both of these once the phase changes. To facilitate this, we must enroll on the barrier again, and use one barrier end in the sender and one in the receiver. (If we did not enroll a second time, and tried to use the same single barrier end twice in parallel, this would be a mis-use of the library.) So here is most of the node process:

node :: NodeInfo -> [Shared Chanin NodeInfo] -> Chanout NodeInfo
     -> Enrolled PhasedBarrier Phase -> CHP ()
node start neighbourChans out bar = node' start
  where
    node' cur
      = do Discover <- syncBarrier bar
           (_, neighbourPos) <- furtherEnroll bar $ \bar2 ->
              giveOutPosUntilBar cur <||> do pos <- readAll neighbourChans
                                             Act <- syncBarrier bar2
                                             return pos
           node' (updatePos neighbourPos cur)

    giveOutPosUntilBar cur = (writeChannel out cur >> giveOutPosUntilBar cur)
                               <-> do Act <- syncBarrier bar
                                      return ()

The sender is the giveOutPosUntilBar process, and the receiver is on the right-hand side of the parallel. By making explicit the phase that we expect to begin with each barrier synchronisation, we both make our code clear (you can see which part is in the discover phase, and which part is in the act phase) and also effectively assert correctness; if the pattern-match fails, your code will produce an error.

Updating the position of the node based on its neighbours is all pure code. This is not a very sophisticated algorithm, but it will suffice for the purposes of illustration:

    updatePos poss cur = cur + (0.05 * average
      [let v = p - cur in v - ideal * normalise v | p <- poss])
      where
        ideal = 0.3
        normalise (NodeInfo x y) = NodeInfo (x / mag) (y / mag)
          where
            mag = sqrt (x*x + y*y)
        average xs = sum xs / fromIntegral (length xs)

The draw process is mainly irrelevant OpenGL logic (adapted from my boids example), but the interesting part is that it must act in the discover phase, partly because that’s the only time that the nodes will send their position, and partly because it’s actually the drawing that drives the frame-rate (a pull-based architecture).

drawProcess :: [Shared Chanin NodeInfo] -> Enrolled PhasedBarrier Phase -> CHP ()
drawProcess input bar
 = do displayIO <- embedCHP_ $ do syncAndWaitForPhase Discover bar
                                  xs <- readAll input
                                  liftIO $ do startFrame
                                              mapM_ draw xs
                                              mapM_ (drawEdge xs) edges
                                              endFrame
      liftIO (do setup
                 displayCallback $= glRunAs2D displayIO
                 let addTimer = addTimerCallback 500 timer
                     timer = addTimer >> postRedisplay Nothing
                 addTimer
                 mainLoop)
  where
    setup = do initialWindowSize $= Size 500 500
               getArgsAndInitialize
               initialDisplayMode $= [DoubleBuffered]
               createWindow "CHP Graph"

    startFrame = do clearColor $= Color4 0 0 0 0
                    clear [ColorBuffer, DepthBuffer]

    endFrame = do flush
                  swapBuffers

glRunAs2D :: IO () -> IO ()
glRunAs2D draw = do
  (matrixMode $= Modelview 0) >> loadIdentity
  (matrixMode $= Projection) >> loadIdentity
  ortho 0 1 0 1 (-1000) 1000
  preservingMatrix draw

draw :: NodeInfo -> IO ()
draw (NodeInfo x y) = renderPrimitive Polygon $ sequence_
  [ vertex $ Vertex2 (x + 0.05 * cos t) (y + 0.05 * sin t)
  | t <- map ((pi/10)*) [0..19]]

drawEdge :: [NodeInfo] -> (Int, Int) -> IO ()
drawEdge nodes (s, e) = renderPrimitive Lines $
  vertex (Vertex2 x1 y1) >> vertex (Vertex2 x2 y2)
  where
    (NodeInfo x1 y1, NodeInfo x2 y2) = (nodes !! s, nodes !! e)

Finally, we must initialise the nodes and wire up the simulation. For our barrier, we will use the enrollAll_ function that takes a barrier-creation function, a list of processes that take an enrolled barrier as a parameter, and runs them all in parallel with their own enrolled barrier ends (discarding the output). Crucially, enrollAll does the enrolling before any of the processes have begun. If you run your processes in parallel and get them to enroll themselves, you will create a race hazard in your program: one process might enroll and start synchronising by itself before the other processes have started executing. This is almost certainly not what you want. So here is the code:

startNodes :: [NodeInfo]
startNodes = [NodeInfo 0 0, NodeInfo 0 1, NodeInfo 1 0, NodeInfo 1 1]

edges :: [(Int, Int)]
edges = [(0,1), (1,2), (2,0), (1, 3)]

main :: IO ()
main = runCHP_ $
       do outChans <- replicateM numNodes oneToAnyChannel
          enrollAll_ (newPhasedBarrier Act)
           (drawProcess (readers outChans) :
            [ let edgesOut = filter ((== i) . fst) edges
                  edgesIn = filter ((== i) . snd) edges
                  connectedNodes = map fst edgesIn ++ map snd edgesOut
              in node n (readers (map (outChans !!) connectedNodes)) (writer c)
            | (n, c, i) <- zip3 startNodes outChans [0..]])
  where
    numNodes = length startNodes

The list comprehension uses the edges list to pick out all the right channels for each node (i.e. it translates the connectivity expressed in the edges list into the channel topology). The code in this post forms a complete program. It is not completely effective as I have not added repulsion among non-connected nodes (an exercise for the reader perhaps), but here is a quick screenshot of the result:

My intention with this example was to illustrate the use of shared channels, and particularly barriers. The pattern shown here, of dividing simulations into phases, is one of their most common uses but they can be used elsewhere, sometimes in place of channels; from a more abstract perspective, channels in CHP offer synchronisation and communication, whereas barriers offer purely synchronisation. A one-to-one channel carrying the unit type is semantically equivalent to a two-party barrier with no phase information. The channel has the benefit of not needing explicit enrollment, but the disadvantage of being asymmetric in its use. For example, picking up and putting down forks in the dining philosophers example can be implemented using either two-party barriers or channels carrying the unit type.

Note: As often with my recent posts, writing them revealed that I lacked certain useful helper functions, so you will need the new CHP 1.7.0 (which also includes other changes I will discuss in future) for the above code.

by Anna Fenton

In the spring of 1996, I was heavily entrenched in a “life stage” transition from high school to college. My mom and I were traveling up and down the east coast for soccer recruiting trips to figure out where I would be spending the next four years of my life.

Penn State, UPenn, Tufts, St. Mary’s, Maryland, Lehigh, Colgate, Bucknell…these were a few of the places that were scouting me for soccer. They were a range of schools. Some were big (100,000 students), while some were small (1,200). Some were highly recognized for their academia, while others were acclaimed for their sports programs. Some were Division III programs, and some were Division I. Some were offering scholarships, and others were offering a warm hand. But what they all had in common was that they were claiming to want ME.

I had been accepted to every program (over a dozen) except for one. My heart was set on UPenn, but they were claiming that I needed 20 more points on my SAT’s in order for them to get me in. At the time, Penn State and Maryland were top tier Division I soccer programs, but I knew that if I went to play for them, I would likely sit the bench. They wouldn’t even guarantee that I would make the team.

In the end, it came down to two schools - Lehigh and Bucknell. It was a great feeling to know they were fighting over me. They had both claimed that I was their #1 recruit, and they were both Division I programs. They weren’t a top 20 program for soccer, but at least I would be starting as their goalkeeper. Academically speaking, they were on par with each other, and while they weren’t technically an Ivy League school, they pooled from the same caliber of students each year. So I had a tough decision to make. Bucknell or Lehigh?

What made the difference was their confidence in me. Bucknell wanted me more, and it felt good to know that I was what they wanted. But that didn’t last very long. The honeymoon was soon over after the season started 6 months later. I was their #1 recruit, and I did start over a senior goalkeeper my freshman year, but that quickly ended. Playing Division I college ball was bigger, tougher, faster, and meaner than I had ever imagined, and I was in for a rude awakening.

I went from starting and playing the whole game, to barely playing 10 minutes by the sixth game of the season. Some of it wasn’t my fault at all. Apparently the fourth year goalkeeper had a lot of political clout with the head coach, and despite the strong support of other players and even the assistant coach, it wasn’t enough to get me playing time. But if I were to be completely honest, I simply didn’t live up to the expectations of the Head coach. I was a disappointment…or at the very least, not quite as good as he had hoped for.

It was a fair assessment though. I certainly wasn’t any worse than the other goalie, but I wasn’t unbelievably better either. So when push came to shove, everyone was disappointed with the result – including myself. I wasn’t quite the superstar that everyone expected me to be, and I wasn’t the superstar that I was used to being.

But four years later…I was ranked #1 in the country for a Saves Per Game stat amongst the NCAA’s Division I goalkeepers, achieved Defensive Player of the Week in the Patriot League, and helped to lead Bucknell into unprecedented wins against rivals such as Lehigh and Colgate. At one point, we had even placed in the Top 25 NCAA Division I men’s soccer rankings, upsetting a couple Top 10 programs in the country.

The reason this all came to mind is because I feel as though I’ve been experiencing a bit of Déjà vu. The past 17 months of my career at Immersion Active have been a steady uphill battle of finding my mojo, chi, or in layman’s terms, my confidence. And although everyone has been extremely supportive and patient (as well as invested in my success) at Immersion Active, I haven’t quite lived up to the hype of someone with a Director of Business Relationships title.

So, it has been decided (and I can’t say I fault them for it) that I am no longer a “Director”.

So last week, I didn’t want to continue reflecting myself as this position, despite their insistence that the changing of my title could wait until business cards had arrived, so I logged onto my account on LinkedIN and edited my title to “Business Relationships Manager”. Within 24 hours, I had received a warm note from a client which I found to be quite humorous and ironic:

Lol. Little did she know, I had just been DEmoted, not PROmoted (in title only).

Even though this story doesn’t have a happy ending (yet), I know that the story is still waiting to unfold. The book of my life is still being written, and I’m anxious to see the story that God will write for me.

As we near Thanksgiving this year, I am reminded of how thankful I am, of how far I’ve come, and how much I’ve accomplished. I may have had my share of disappointments and unexpected outcomes, but I’ve also had my share of things to be thankful for as well. Pastor Steve wrote something that “stuck” with me a couple of weeks ago,

When you face an enemy, obstacle, or setback, you can either see it as a wall or a door.

As he explained, the Israelites chose to look at Goliath as a wall…and they waited for 40 days hoping it would go away. David, on the other hand, saw things differently – and God blessed him for it.

Every champ has his/her chump moments. But they’re a champ because they never stay a chump.

This chump is moving on.

At the risk of coming across as naive… I’m still pondering about ‘objectivity’ in the research element of the project.

Can we actually be truly objective about our project and research?  We were involved in the bid writing and thin.clusion was our idea – so are we the best people to actually implement and research it?  Surely we already have some kind of bias toward positive outcomes?

Perhaps I’m missing the point a little?  I feel objective enough about the project – I can anticipate the potential barriers (is this also being biased? – having pre-conceived ideas of ‘barriers’)

Do I just step back a little from the project and allow it to take its own course, with little or no involvement – except for carrying out the research elements?  Will that give me a more objective viewpoint?

Pondering the meaning of life, for sure.

Flickr: Uploaded by ben pollard on 18 Mar 08, 5.36PM GMT.

Susana Iranzo (Universitat Rovira Virgili) and Giovanni Peri (University of California Davis, CESifo and NBER) have a new paper that looks into the relationship between migration and trade:

The remarkable increase in trade flows and in migratory flows of highly educated people are two important features of globalization of the last decades. This paper extends a two-country model of inter- and intra-industry trade to a rich environment featuring technological differences, skill differences and the possibility of international labor mobility. The model is used to explain the patterns of trade and migration as countries remove barriers to trade and to labor mobility. We calibrate the model to match the features of the Western and Eastern European members of the EU and analyze first the effects of the trade liberalization which occurred between 1989 and 2004, and then the gains and losses from migration which would occur if barriers to labor mobility are reduced. The lower barriers to migration result in significant migration of skilled workers from Eastern European countries. Interestingly, this would not only benefit the migrants and most Western European workers but, via trade, it would also benefit the workers remaining in Eastern Europe.

All of us want to be liked, but most of us feel it’s a bit of a random process of who likes us and who doesn’t.  We worry and obsess about whether or not someone might like us or not, but we often feel that it’s not something we can control.  That’s totally normal and understandable.

So, let me share a secret with you…and this is the honest truth: I was voted class introvert in high school.  Painfully shy and insecure, most days I felt like nobody liked me.   What I discovered over time was that feeling came from not fundamentally liking myself very much.  In that I discovered a bigger secret that can benefit us all.

So, want to know how to make people like you?  It works almost every single time…with all people and all situtations.  And, if it doesn’t, you will know that it wasn’t you.  That you’ve done everything you can do.  It’s a powerful tool and it’s very simple.

There are two ways to make people like you:

1. Like them first
2. Be helpful

There it is.  So, whether it’s a social gathering, a job interview, or first date….if you want someone to like you (and that is completely up to YOU) then start by just liking them first.

Find something about them that you can genuinely like. For some people that might be that you like their shoes, for others it may be that you have a shared interest in a cause or field of study, and for still others it may be that you like what they do.   The like has to be genuine for it to work or you will do more damage than good by being fake.

Then, if you want to build on that initial sense of like and create a solid relationship, focus on being helpful.

And, by helpful, I don’t necessarily mean you need to go tromping though someone’s life doing stuff for them.  Often there are things people do because they mean to be helpful, but aren’t because they are intrusive, bossy, or irritating.  That’s not what I’m suggesting at all.  Rather, adopting a spirit of helpfulness or an openness to being asked for help is more on target.

For example, after chatting with someone,  a simple “it was so very nice to talk with you today.  If there is ever any way I can be helpful to you, let me know”  can do wonders in building long term relationships.  It doesn’t matter who the other person is or who you are.   This offer of help resonates with people as deeply likable.   You don’t commit to anything in particular and of course reserve the right to say no if what they ask isn’t something you can do, but you hold out an openness to helping them but don’t assume you know what might be helpful.

All of that communicated very simply, honestly, and directly.  Likably.

So, just try it to see.  Smile at a stranger.  Talk with someone sitting by you at a meeting or in class.   Take a leap of faith and know that you are deeply likable and there is absolutely no reason why someone wouldn’t like you if you like them first and radiate helpfulness.   Life’s so much more fun when you like people. …  Starting with yourself.

Any goal can be achieved with difficulty and for a long time or in easy and fast way.

It is very important for women to understand why we need to reach goals in easy and fast way.

Constant stress, predominance of negative emotions, illnesses, and problems with men influence on women’s well-being – physical, emotional, and mental. Through women it influences on our future children. It influences on our future generation. Therefore, it has the main impact on the whole world.

Take a moment to think of it. Take a moment to realize the connection between the well-being of women and the well-being of the whole planet. It depends on us who our children will be. It depends on us what the future generation will be.

Now we will learn how to achieve goals simply and fast.

Take responsibility

First of all, we need to realize that everything what we have or do not have is the result of our behaviour and our patterns of thinking. We create any situations ourselves. Therefore, there is no one to blame. Instead of criticizing, blaming and resenting we need to take responsibility.

The acknowledgment of responsibility makes our life easy, interesting, and fulfilling.  We learn how to reach goals on our own and not to depend on external circumstances. On the contrary, these so-called circumstances start to benefit us.

Focus on the most important

Then it is crucial to be aware what the main thing is for you and always to be focused on it. Most of us do not take time to think of it. Many people pay attention only on the secondary things. That is the reason why they never achieve what they want or they do it slowly and go through a lot of problems.

For example, what is most important in preparation for the childbirth? Giving love and care to your baby or buying pretty infant clothes and furniture?

All we need to do is to sit down and think what is the most important.

Take actions

After we find out what the most important is, we need to act upon it. If we do not make steps, we do not move. You might find it silly that I pay attention to this, but for most of us taking actions is the very difficult thing to do. We usually find a variety of excuses why we can not do it right now. Instead of looking for possibilities – we look for excuses.

We all know these excuses: “I can’t do it now,” “I don’t know how to do it,” or “I will do it later.”

My answer is “if we do not make steps, we do not move.”

Reveal barriers

When we start doing something, we meet barriers which do not allow us to reach goals. Usually they are ridiculous and self-invented. For example, false opinions, illusions, fears, and many others.

The opinion that “all men are egoists” can not let a woman to build good relationship with her husband. Or the fear of meeting people can not let a woman to get a decent job.

In this case everyone has two choices: to discern barriers and then to achieve the goal easily and fast; or to give up and forget about dreams.

Self-education

The last thing I want to mention is a self-education. Self-education is a key to fulfilling life. The one who does not educate himself is like a donkey that goes where others lead him.

In order not to be a donkey, the one needs to question everything. Maybe I am just lying to you right now?

Share your opinion!

Getting around Ryerson campus is not always the easiest experience in my power chair.

I have Cerebral Palsy which limits my ability to walk long distances, affects the usefulness of my hands, and impairs my speech.  I use a power chair and I am accompanied by a guide dog, Lacey, to get around.  I need to be constantly aware of space.  Every day, I manoeuvre through tight spaces, heavy doors, busy elevators and creaky lifts.   In going through the obstacle course that is my day, every day, I have learned to overcome the embarrassment that comes with getting stuck on Ryerson’s dated wheelchair lifts, caught in snow banks and to continue moving forward.

Exercising:

I really should go to the gym more often.  I bet you have said this too.  However, my trouble is not only finding the time to work out, but also figuring out how to get into the RAC.

In my power wheelchair, I travel through the maze of Kerr Hall and two accessible doors to the gym entrance.  After swiping my card and going through an automatic door, I have to fight with a heavy door that is not automatic.  A student without upper body strength need’s someone to open the door for them.

Bikes and treadmills are accessible on the main floor that I can use, but weight equipment is down three flights of stairs.  In a tucked away corner by the stairs to the running track, there is the RAC’s solution- an old, finicky lift.  Operated by a key, and barely large enough for a wheelchair, the lift feels unsafe, like riding a wooden roller-coaster would.   I cannot use this lift by myself so I have to pay an attendant $20 to accompany me for weekly workouts.

Once, my worker and I had to ask a gym employee to operate the lift from outside because we could not get it to respond.  At one point, we were stuck halfway between floors.  The lift made a creaking sound every few seconds that scared me, my worker and Lacey.   From then on, we chose to walk down the stairs – it actually is the safer solution.  Going to the gym is crucial for my physical and emotional health, but finding the time and money limits my number of visits.

Elevators – Waiting in line…

Moving around other buildings poses challenges too.  In the library building there is an elevator reserved for people with disabilities.  Despite this, there is often a crowd of able-bodied students surrounding the elevator doors.  With the stairs directly behind the elevator, able-bodied students line up causing disabled people to wait longer and squeeze into tighter areas.   It amazes me to see a crowd in front of the elevator and how, if it’s taking too long, they decide to take their stairs.  I think, “If only I had that option.”

Test Centre for Students with Disabilities … really?

The test centre for students with disabilities is in the basement of the Victoria building.  Yes, it’s in the basement and it’s not accessible via the main elevator.  To access this service, students with physical disabilities must be able to call from a phone on the first floor and ask for someone to come get them with the lift.  They may wait a while until a staff person answers the phone or is able to leave the office.    Because of this arrangement, most students with physical disabilities write tests and exams in another location in the POD building.  Needless to say, I have only been in the test centre once.

Dreadful winter

It’s not only some buildings that are inaccessible, it’s the path to them.  That’s why I am dreading the winter.  I’m not a fan of the freezing temperatures, but it’s the snow that worries me.  Moving around the crowds on campus in my power wheelchair can be an ordeal, and snow-covered sidewalks make my day more challenging.  In past winters, I have missed class because I cannot get past the ramp outside of the ILLC.  Campus security guards have declined to assist students with disabilities to class as it is deemed unsafe.  Yet, the campus remains open and I miss class.  Returning from weekend trips home I have gotten stuck on un-ploughed campus sidewalks and depended on people passing to give me a hand.  It is embarrassing, yet seems to be an inevitable part of living on campus.  When the sidewalks are cleared, I am grateful for the ease of moving around campus.

Finding easier ways

These extra steps throughout my day have become routine.  Most of the time I don’t pay attention to the time or effort it takes in moving around campus.  I’ve learned to speak up about barriers or figure out alternative paths.  I take full advantage of tunnels and know where the best elevators are.  I’ve scouted out the accessible bathrooms.  I am grateful for the food service providers (especially at Tim Horton’s and Maggie’s) who make sure I have my meals and daily caffeine fix.  And though I don’t get to the gym as often as I’d like, I’m still getting there.

As we know from our high school science classes, sound waves are created when an object moves or vibrates. When these waves reach our ears, they cause our eardrums to vibrate, sending signals to the brain that we interpret as sound. 

 | Frequency 

  [number of wave crests per unit time that pass a fixed location | measured in Hz- the number of waves per second]

 measures the tone or pitch of a sound.  
 For example, a bass guitar plays lower frequencies than a violin. 
 Human beings usually hear 15 to 20,000 [Hz] frequency sound.  

    | Amplitude  

 [the wave height | measured in dB] used to measure the intensity of a sound

 A measurement of the wave traveling through the air is used as an indication of the intensity of sound or its volume, and is described in terms of a scale called the decibel (dB). Noise measurements made by filtering high- and low-pitched sounds—approximating the way an average person hears sounds—is called the A-weighted level or dBA [1924].

   0 dB     |          the faintest sound we can hear;
 30 dB   |          a quiet library or in a quiet location in the country; 
  45 dB   |          typical office space or ambience in the city at night;  
 60 dB   |          a restaurant at lunch time; 
70 dB   |          the sound of a car passing on the street; 
80 dB   |          loud music played at home;  
 90 dB   |          the sound of a truck passing on the street;  
 100 dB |          the sound of a rock band; 
115 dB |          limit of sound permitted in industry; and  
 120 dB |          deafening. 

 The dB(A) scale begins at zero, which represents the faintest sound that can be heard by humans with very good hearing. Conversations take place in the 50 dB(A) range and a chainsaw whines at about 100 dB(A). Normal highway traffic sounds rank about 75 dB(A) and jet airliners around 90 dB(A). For most people, discomfort starts in the 70 to 80 dB(A) range, with the threshold of pain around 140 dB(A). The Federal Highway Administration (FHWA) has chosen 67 decibels as the point where state and federal agencies must consider reducing the noise level.

   | Sound behavior

 Because the sound has the properties as wave, it has properties of “reflection” and “transmission”, and attenuates in accordance with the distance. These properties are illustrated below for reference.

When a wave reaches the boundary between one medium another medium, a portion of the wave undergoes reflection and a portion of the wave undergoes transmission across the boundary. The reflected wave may or may not undergo a phase change (i.e., be inverted) depending on the relative densities of the two media. It was also mentioned that the amount of reflection is dependent upon the dissimilarity of the two medium. For this reason, acousticly minded builders of auditoriums and concert halls avoid the use of hard, smooth materials in the construction of their inside halls. A hard material such as concrete is as dissimilar as can be to the air through which the sound moves; subsequently, most of the sound wave is reflected by the walls and little is absorbed. 

Reflection of sound waves off of surfaces can lead to one of two phenomenon – an echo or a reverberation. Reverberation |  often occurs in a small room with height, width, and length dimensions of approximately 17 meters or less. Why the magical 17 meters? The effect of a particular sound wave upon the brain endures for more than a tiny fraction of a second; the human brain keeps a sound in memory for up to 0.1 seconds. If a reflected sound wave reaches the ear within 0.1 seconds of the initial sound, then it seems to the person that the sound is prolonged. The reception of multiple reflections off of walls and ceilings within 0.1 seconds of each other causes reverberations – the prolonging of a sound. Since sound waves travel at about 340 m/s at room temperature, it will take approximately 0.1 s for a sound to travel the length of a 17 meter room and back, thus causing a reverberation (t = v/d = (340 m/s)/(34 m) = 0.1 s). This is why reverberations is common in rooms with dimensions of approximately 17 meters or less. Perhaps you have observed reverberations when talking in an empty room, when honking the horn while driving through a highway tunnel or underpass, or when singing in the shower. In auditoriums and concert halls, reverberations occasionally occur and lead to the displeasing garbling of a sound.

Echo | Reflection of sound waves also lead to echoes. Echoes are different than reverberations. Echoes occur when a reflected sound wave reaches the ear more than 0.1 seconds after the original sound wave was heard. If the elapsed time between the arrival of the two sound waves is more than 0.1 seconds, then the sensation of the first sound will have died out . In this case, the arrival of the second sound wave will be perceived as a second sound rather than the prolonging of the first sound. There will be an echo instead of a reverberation. 

| Geometry & Reflection

The way sound rays reflect off plane (flat) mirrors or surfaces is very simple: the angle of incidence is equal to the angle of reflection.

The angles in question are measured with respect to the normal to the reflecting surface. The angle of incidence is the angle the incident light ray makes with the normal, and the angle of reflection is the angle the reflected ray makes with the normal.

This is exactly the same behavior seen in billiard balls bouncing off a rail on a pool table. (We neglect friction and other minor effects.) Even sound waves bouncing off a wall at an angle follow this same law of reflection.

| Atmospheric effects

Winds will increase sounds downwind from a source and reduce them upwind. This is not solely a result of the velocity effect, but also because the spherical wave-front is deformed by the prevailing wind.   

As discussed previously, the speed of sound is dependant on temperature, the higher the temperature, the higher the speed. This means that when the temperature near the ground is higher than that of the upper air, sound rays tend to arc upwards slightly. Thus less energy will reach a listener some distance away at ground level. (For a given amount of sound energy, the distribution area is increased).

At night, when the ground surface is cooler than the upper air, the inverse occurs: sound energy tends to arc downwards. (For a given amount of sound energy, the distribution area is reduced).

| Barrier Design

Barriers, such as walls or screens, will act to create an acoustic shadow. The reduction in sound level within this shadow zone is dedendant on frequency (as we discussed earlier). At high frequencies the effect of the barrier is most pronounced whereas at low frequencies much diffraction occurs at the edges, so the shadow effect is diminished.

| Type of Road and Average Speed

Rural roads – 110 km/h speed limit – use 1
Urban Freeway – 90 km/h speed limit – use 92
Urban Highway – 70 km/h speed limit – use 65
Urban Street – Dual Carriageway – use 60
Urban Street – Single Carriageway – use 55
Urban Street – Single Congested – use 50 

Background: There is currently a paucity of research investigating what speech and language therapists, in particular, perceive are the greatest barriers to implementing evidence-based practice.

Aims: The purpose of this study was to investigate the perceived barriers that are faced by speech and language therapists in southern Ireland when attempting to implement evidence-based practice.

Methods & Procedures: A 34-item questionnaire was sent to 39 therapists working in several counties in southern Ireland. The survey received an 82.1% (n = 32) response rate.

Outcomes & Results: The results of the study indicated that certain barriers are perceived to prevent evidence-based practice being implemented successfully. The most significant barrier affecting evidence-based practice implementation was reported to be a lack of time to read research (71.9%). Additional barriers that were found to be the most significant were the research having methodological inadequacies (62.5%) and insufficient time to implement new ideas (59.4%). Other important factors identified as being significant barriers to the implementation of evidence-based practice were those associated with the quality and presentation of the research, workplace setting, and lack of skills of the therapist. Associations between specific barriers and workplace setting or grade were also investigated. Some possible reasons for these barriers and the implications for clinical practice are also discussed.

Conclusions & Implications: This small study suggests that therapists agreed that evidence-based practice is essential to the practice of speech and language therapy. There are, however, barriers in place that are perceived to prevent its successful implementation. It is hoped that because these barriers have been identified, individual clinicians and organizations can be proactive in aiming to provide an evidence-based service to their clients.

from the International Journal of Language and Communication Disorders

East Side Gallery - Associated Press Photo

November the 9th will be the 20th anniversary when East Germany opened its borders after the challenge from President Ronald Reagan, “Tear down these walls!”

I listened this morning to a radio commentary about the event and how this radio host had kept a souvenir of the wall as a reminder of the event. But he went further…he talked about what it was like today, after 20 years. Had things changed?

To some degree they had, however what people had envisioned would happen didn’t. East Germany still remains the poor cousin of the West and there differently are still barriers, although not so visiable.

I started to reflect then on what barriers we put in front us that stop us from change – obtaining our goals and visions. The reasons of course are in our past and solution lies with us.

A good friend up dated his status on Facebook, with “I must…must take more pictures?” Or something like that – in reference to his lacking of getting out and getting some shots. His lifstyle is “busy”. Job, kids, hockey schedules, “honey-do’s” and a merry range of other items fill up the limited hours in his and our days.

The East Side Gallery is getting a face lift today, November 6th. The artists that painted the murals are back putting fresh paint on their master pieces. It is the worlds longest open-air art gallery. To see the wall’s gallery visit:

http://www.eastsidegallery-berlin.de/

So what do you do with your barriers. Put a fresh paint on them? Maybe and why not? But put on the paint, not to hid the fact that it is a wall, but to celebrate that it is there. Acknowledge and acceptence are keys to happiness. So what if you are too busy to take photos, or too busy to do whatever…accept it AND go out and take photos. You ARE a busy person and don’t let that define you and restrict you from your vision or goals.

“They” the committee of they say “something has to give.” You can’t do everything. The great human experiment is to find how much one can do with his life. You don’t have to do anything, but even the act of doing nothing is something – so this all gets very confusing…

I am going to leave you with a few words from a favorite philospher “Bucky” Fuller:

• The Things to do are: the things that need doing, that you see need to be done, and that no one else seems to see need to be done. Then you will conceive your own way of doing that which needs to be done — that no one else has told you to do or how to do it. This will bring out the real you that often gets buried inside a character that has acquired a superficial array of behaviors induced or imposed by others on the individual.

He also said, “Don’t fight forces, use them.” I apply that to mean, don’t fight the fact that you have a certain lifestyle – use it to accomplish your vision or goals – work with it.

Now in closing to my friend…is it that you want to take great pictures or is just pictures stopping you. You can always take pictures anywhere, anytime and while doing anything…question is, can you do all that and take great pictures too?

Cheers

(Photo: AP Photo/Gero Breloer)

How the Brazilian city of Rio is walling in its favelas… From BBC News. Full story

This site may contain information about: city lights. The blog is also related to: newyork city.

Just to be clear, I am a huge fan of lean. I am a huge fan of Toyota and have seen some great results from projects over the years. Lately it seems, though, that lean in the United States and Europe is turning into just another overused, misunderstood concept that everybody does but few actually do well.

Read more at http://corporatedeathspiral.blogspot.com

“If it’s very painful for you to criticize your friends – you’re safe in doing it.  But if you take the slightest pleasure in it, that’s the time to hold your tongue.”  ~ Alice Duer Miller

“Other people’s opinion of you does not have to become your reality.”  ~ Les Brown

“Preconceived notions are the locks on the door to wisdom.”  ~ Merry Browne

 ”Judgements prevent us from seeing the good that lies beyond appearances.”  ~ Wayne W. Dyer

 ”Our thoughts are unseen hands shaping the people we meet.  Whatever we truly think them to be, that’s what they’ll become for us.”  ~ Richard Cowper

 ”If only closed minds came with closed mouths.”

 ”I am an invisible man…. I am a man of substance, of flesh and bone, fiber and liquids – and I might even be said to possess a mind.  I am invisible, understand, simply because people refuse to see me.”  ~ Ralph Ellison, The Invisible Man, 1952

Sometimes Love Hurts ~ by Carol Wiebe

Sometimes, love hurts. We might even question if exposing ourselves to love is worth the potential pain.  

But love is not just a surface emotion: it embodies a wide spectrum of feelings. So wide, that poets, writers, artists, saviours, politicians, advertisers and countless others have tried to define and describe it through the ages. Various holy writings declare that God is Love.

I do not think it is possible to live on this earth and miss the experience of love. Simply being born is a gift of infinite love: we are given access to all the possibilities (including challenges) that life has to offer. The more we open ourselves, the deeper our experience can go.

Your task is not to seek for love, but merely to seek and find all the barriers within yourself that you have built against it.    ~ Rumi

The authors were invited to present at the international conference, Metropolitan Desires: Cultural Reconfigurations of the European City Space, in Manchester, UK on 8th September. The paper, ‘The City and the Control of Space’, discusses the relationship between the city and space [public and private]. More specifically, it focuses on how the city and ownership of space is demarcated, enclosed and implied. The roles of governance and security upon civic, urban and personal space call into question the true nature of that which we consider public. Frequently the design of public space is concerned with the control of that space, rather than its appropriation, [de-re-mis]use and legitimate occupation. Increasingly it is the proximity of digital and real space that is testing these realities and challenging the convention of behavioural patterns established incrementally by the accumulation of policy and technological change during the last 20 years. The question of what constitutes community, networked and residual space is of concern here as are devices of appropriation, enclosure, severance, fragmentation, and cultural identification of space.