Engineering Thread

TheAmateur

Boston wrote: »

int index_of_largest(int unsorted_a_int*,int size)
{
if(size==1)
return 0;

index = index_of_largest(a,size-1);
if a[size] > a[index]);

return size;
else
return index;

}

aah, I was trying to do it without your variable "index" and I kept getting errors. Any idea why this didn't work? (technically it should... i think)

int index_largest (double a[], int size){
    if (size == 0)
       return size;
    if (a[size-1] < a[(index_largest(a, (size-1)])
         return index_largest (a, (size-1));
    else
        return size;
}

it didn't like the
"if (a[size-1] < a[(index_largest(a, (size-1)])"
line.
Also, we technically don't have to make the index_largest function recursive (she just says "it could be" in the sheet) but I get the feeling if i go in there and use iteration I'll hear "If you want the last mark you gotta use recursion in both functions" or something...:rolleyes:

Boston

Its not going to work because you're stuck in a recursive loop with no way out. When you find the largest index you need to return it, instead you make another recursive call which in turn again finds the largest index with again calls another recursion.

The recursive swapping is easy compared to the recursive max index.

Brian

I hate functions, and I hate recursion even more.

Boston

It's simply a new way of thinking about something. Recursive programming one of the simpler paradigms.

TheAmateur

Boston wrote: »

Its not going to work because you're stuck in a recursive loop with no way out. When you find the largest index you need to return it, instead you make another recursive call which in turn again finds the largest index with again calls another recursion.

The recursive swapping is easy compared to the recursive max index.

I see. Got another problem though. I get an error message at the "else" bit of your method. It goes:

In function `int index_largest(double*, int)': 
expected primary-expression before "else" 
expected `;' before "else"

Any idea why?

Jonathan

TheAmateur wrote: »

I see. Got another problem though. I get an error message at the "else" bit of your method. It goes:

In function `int index_largest(double*, int)': 
expected primary-expression before "else" 
expected `;' before "else"

Any idea why?

Look at your code closely.

You are missing some parenthesis.

(a[size-1] < a[(index_largest(a, (size-1)))])

TheAmateur

Well I actually meant Boston's method, but you're absolutely right, and it compiles perfectly now. I'm not so sure anymore it does what I thought it did though, so I took it out. I'm still getting that error message with Boston's code.

Brian

I, too, will be looking at your code closely.

Boston

I probably have a ; or } in the wrong place, I didn't intend for you to copy and past it.

TheAmateur

Never said I did...:pac: but if someone were to, they'd end up with a crazy value in the first slot of their function. No idea why though. I've pretty much given up on the whole recursive index_largest function anyways, seeing as we don't specifically "have" to make it recursive.
EDIT: Got the feckin thing. Stupid swap function was swapping a value outside the array...

Brian

Need some help.

Want to share my laptop's LAN connection over Wifi to my phone. This works with my 3G modem on Mac OS X, but I want to do it in Win7 with my Halls TCD LAN. Clearly the proxy is going to present some problems. I set up an ad-hoc connection with the Network Manager to broadcast, which it does, but my phone can't connect to the internet through it.

Giz a clue.

Nietzschean

Check if you can reach local tcd sites on the phone, if you can then just configure the phone to use the tcd proxy, it should have settings pages for it. Though be aware you are breaking all sorts of college regulations by having a rogue access point......

Jonathan

Two year hiatus??

Nietzschean

Yeah good times, i figured i was doing enough nerdy things

Brian

Nope, local pages don't work. There doesn't seem to be a proxy settings option in the S60V5 browser/network manager either.

Hmm. Maybe if I got Metakall, it would see my computer as a TCDwifi thing...

Jonathan

Baza210 wrote: »

Nope, local pages don't work. There doesn't seem to be a proxy settings option in the S60V5 browser/network manager either.

Is is under advanced options in access points manager in S60V3.

devinejay

Slightly unrelated, but you've just reminded me;

Has anyone tried/managed to get the Last FM plugin/app thing to work on TCDwifi? I would be inclined to think the NAC wouldn't let anything but the usual browsers on to use the proxy, so I never expected the app to be allowed, hence never really trying too hard.

Jonathan

devinejay wrote: »

Has anyone tried/managed to get the Last FM plugin/app thing to work on TCDwifi? I would be inclined to think the NAC wouldn't let anything but the usual browsers on to use the proxy, so I never expected the app to be allowed, hence never really trying too hard.

Last.fm on what? windows or mobile phone? It works on both anyways.

tcdproxy.tcd.ie:8080

devinejay

Windows, I'll give that a go. The extents of my efforts consisted of sticking the proxy address from Chrome into the plugin, didn't work anyway!

Cheers

Jonathan

devinejay wrote: »

Windows, I'll give that a go. The extents of my efforts consisted of sticking the proxy address from Chrome into the plugin, didn't work anyway!

Cheers

The proxy for the browsers is http://proxypac.tcd.ie/accelerated_pac_base.pac

If you download the file and actually look at what it contains you will see this:

// The last octet of the IP address is examined
// Even TCD IP addresses will be pointed to proxya, odd addresses proxyb
// PAC file format official documentation: http://wp.netscape.com./eng/mozilla/2.0/relnotes/demo/proxy-live.html

// Variables:
//              host            - the hostname extracted from the URL.
//              url             - the full URL being accessed.
//              ip              - IP address of client, can resolve to loopback if used from non-TCD network.
//              lastoctet       - last octet of IP address of client.
// removed dnsDomainIs(host, ".tcd.ie") || dm 21-5-07

function FindProxyForURL(url, host)
{
        var ip = myIpAddress();
        var lastoctet = ip.split('.');

        //if requested URL is local; go direct else via proxy
        if (isPlainHostName(host) ||
            dnsDomainIs(host, "localhost") ||
            dnsDomainIs(host, "127.0.0.1") ||
            url.substring(0, 5) == "news:" ||
            isInNet(host, "134.226.0.0", "255.255.0.0") ||
            isInNet(host, "10.1.0.0", "255.255.0.0") ||
            isInNet(host, "10.2.0.0", "255.255.0.0") ||
            isInNet(host, "10.5.0.0", "255.255.0.0") ||
            isInNet(host, "193.1.64.0", "255.255.248.0"))
            return "DIRECT";
        else {
                //loadbalance even ending IP's to proxya, odd to proxyb
                if ( (lastoctet[3] % 2) == 0 )
                        { return "PROXY proxyA.tcd.ie:8080 ; PROXY proxyB.tcd.ie:8080" ; DIRECT }
                else    { return "PROXY proxyB.tcd.ie:8080 ; PROXY proxyA.tcd.ie:8080" ; DIRECT }
             }
}

A different proxy is used for third party apps.

devinejay

Fair nuff.

Also, this new last.fm subscription thing is BS. I Heard about it at the end of March but when no "free trial" appearing I figured they had included us in their definition of the UK (thankfully for once). Then it don't use it for a while, come back and I've got 30 tracks left...

Still, 3 euro a month isn't bad at all. Just an inconvenience.

Jonathan

devinejay wrote: »

Also, this new last.fm subscription thing is BS. I Heard about it at the end of March but when no "free trial" appearing I figured they had included us in their definition of the UK (thankfully for once). Then it don't use it for a while, come back and I've got 30 tracks left...

I've never subbed to last.fm. I dont see any advantages.

I'm happy with my current setup of scrobbling and listening to streams from foobar and mobber and then making funky graphs with last.fm extra stats

devinejay

Funky indeed...

Brian

jmccrohan wrote: »

Is is under advanced options in access points manager in S60V3.

Finns inte i v5.

TheAmateur

Was going through the 2006-7 1e5 exam and q.4 reads

A car of mass 500kg rounds a bend on a horizontal road at a speed of 10m/s. The radius of curvature of the bend is 40m. The fictional force (in Newtons) exerted on the car by the road is:
a = 5000, b = 250, c = 4900, d = 1250, e = 125"

Is that supposed to read "frictional"?
On a more general note, anyone got any study tips for these exams?
Also, this thread has slipped to page 3!?!
EDIT: just realised it probably isn't frictional... there's no coefficient of friction given in the question. The question just looks the same as one he gave in a tutorial about friction stopping the car from sliding as it's going around the corner. So.... I'm guessing the answer is not C?

Jonathan

So (500*(10)^2)/40 = 1250 so yea, D is correct.

Boston

Jonathan wrote: »

So (500*(10)^2)/40 = 1250 so yea, C is correct.

D is correct no?

TheAmateur

Oh yeah... I meant d too... yep. Em, just out of curiosity, why would the centripetal force come from the road? And why is it "fictional"?

Jonathan

Boston wrote: »

D is correct no?

Duly edited.

Boston

It's been a long time since i did mech, but afair you need to find frictional force and do up a force balancing equation.