Budget to go to Practiscore

[wgnoyes]

Bill, the newest android versions can go out and find other devices. Doesn't work great with nooks (takes awhile to find devices), but works great on my Galaxy tablet.

Oh I know. As I wrote above, that feature works fine on the nooks when on my apple time capsule router here at the house, and the nooks find other devices pretty quickly. It doesn't work out in the field, however, on a linksys wrt54g and also still doesn't work here on a tp-link 3040 portable wireless router which I intend to use in place of the linksys.

[Graham Smith]

Oh I know. As I wrote above, that feature works fine on the nooks when on my apple time capsule router here at the house, and the nooks find other devices pretty quickly. It doesn't work out in the field, however, on a linksys wrt54g and also still doesn't work here on a tp-link 3040 portable wireless router which I intend to use in place of the linksys.

So it may be down to the router. I've tried this with a Linksys and a Trendnet router at home and office and with a Zyxel portable in the field.

[wgnoyes]

Sure. If it works with my apple time capsule but not with this other router, the variable is the router. Chris thinks it may have something to do with something called a ICMP setting somewhere in the router, but I don't know what that is.

[hitman]

Most of our Nooks get out of the wifi range at every match and when they come back in range they pick right back up.

Don't remember what wifi were using but its not an expensive one.

Yes if you want to reach out and grab scores you will need a list but syncing at the end takes me about 5 min for 8 stages.

[euxx]

Sure. If it works with my apple time capsule but not with this other router, the variable is the router. Chris thinks it may have something to do with something called a ICMP setting somewhere in the router, but I don't know what that is.

"Works" as in how? Are you saying iOS can see your android devices on the same router, but android devices can't see each other?

If no devices can see each other it is likely because your router has "client isolation" [1] feature enabled. You need to go into the router administration console and turn it off.

BTW, Practiscore is not using ICMP or broadcast protocols like Bojour and even iOS app falls back to the port scanning when it is looking for other iOS devices.

I can try to tweak the scanner used in android, but I need some more input from the users, e.g. what are the IP addresses of your devices and how long does it take to resolve them in the app.

Note that if your device IPs are closer to the end of the range of 1..254 it will take longer to resolve them.

[1] http://www.wirelessisolation.com/

[wgnoyes]

Sure. If it works with my apple time capsule but not with this other router, the variable is the router. Chris thinks it may have something to do with something called a ICMP setting somewhere in the router, but I don't know what that is.

"Works" as in how? Are you saying iOS can see your android devices on the same router, but android devices can't see each other?

...

[1] http://www.wirelessisolation.com/

Yes! "Works" in that iOS can ALWAYS see the android devices regardless of the router in use. (Which is why we insist upon an ios device being the master at a match.) When using my apple time capsule (airport extreme w a HD) , everything sees everything else under that router and does so fairly quickly.

Any other router I've tried, ios can see everything automatically, android can see nothing without a sync code.

apple time capsule addr range is 192.168.0.101-150. This is the one where everything works.

tp-link 3040 addr range is 192.168.0.101-150. Apple sees all, android sees nothing.

linksys router addr range is 192.168.1.101-150. Apple sees all, android sees nothing.

Under the latter 2 wifi networks, iOS sees the android devices immediately. Android (nook) never picks up iOS..

[euxx]

Bill what android version have you tried it with? I am asking because you usually never on last version.

[wgnoyes]

1.2.8. I'm usually within a point release.

[euxx]

Well the point releases is where most of the fixes happening. For instance I recall increasing connection timeouts and fixing some issues related to device discovery after 1.2.0. I will try to increase connection timeouts for 1.2.9, that will slow down scanning, so I will probably have to increase number of scanner threads. Right now there are 15 threads with connection timeout 0.6sec. So, in theory it should get to ip addresses in 100th range in about 5 seconds...

[wgnoyes]

(Actually, my mistake. 1.2.5.)

Okay, but (thread totally drifted off the budget topic) that still doesn't answer the question of why 1.2.8 sees ios on one router and not on another. On the tp-link router, there's a setting called ap isolation, which hides devices on the wireless network from one another. That setting is disabled by default. Just to be able to say I did, I turned it on, rebooted, turned it off, rebooted, and still no joy.

[euxx]

Well, my guess is as good as yours. Maybe those routers are routing wireless waves slover than your third one. If I am not mistaken, you could try to bump up signal boost setting on linksys and for the kicks try to turn off all ios/ apple devices around and see if that will make any difference (I am not kidding).

[jcwren]

Eugene,

I've got 1.5.10 on my iPad, 1.2.8 on my Galaxy S4, and 1.2.7 on the NOOK. The router (a TP-LINK TL-MR3040) is handing out addresses 192.168.0.100 (iPad), .101 (S4), and .102 (NOOK).

Here's the time it takes for them to be found by each other,

iPad - ~2 seconds

Galaxy S4 - ~5 seconds (~12 seconds to scan all 254 addresses, call it 21 addresses per second)

NOOK - ~3 minutes, 22 seconds (~8 minutes, 12 seconds to scan all 254 addresses, call it 0.5 addresses per second)

Maybe those routers are routing wireless waves slover [sic] than your third one. If I am not mistaken, you could try to bump up signal boost setting on linksys and for the kicks try to turn off all ios/ apple devices around and see if that will make any difference (I am not kidding).

I'm not sure what this means, but it makes no sense. While a given router might be slower than another, we'd be talking multiple orders of magnitude, perhaps the equivalent of comparing the first wireless router ever built against the the latest top-of-the line router made today. Factually, that's just wrong.

As for "bump up signal boost", it's more useful to know what the signal-to-noise ratio is (generally available from the router, rarely available from client devices). Given that in both mine and Bill's case the router is about 3 feet away from the devices we're using, changing the signal strength isn't going to make a damn bit of difference. In addition, most routers are already configured for the maximum *legal* output power for the country of operation. While it may be possible to increase it, it's not recommended.

1. It's illegal.
2. It can interference with other wireless networks.
3. It can result in adjacent channel interference because of "splatter".

In other words, unless one seriously knows what one is doing, one should not do that.

[euxx]

Here's the time it takes for them to be found by each other, iPad - ~2 seconds

Galaxy S4 - ~5 seconds (~12 seconds to scan all 254 addresses, call it 21 addresses per second)

NOOK - ~3 minutes, 22 seconds (~8 minutes, 12 seconds to scan all 254 addresses, call it 0.5 addresses per second)

Such data I can work with. It is not "newer", just too long. So, numbers for S4 are matching my own math (15 threads in parallel with 0.6 seconds per connection attempt). But the only way I can interpret your data from Nooks is that they are running 2 threads in parallel instead of requested 15.

I will have to play with my setup here it has smaller IPs and unlike Bill's setup I always do get devices discovered.

[Steve_P]

Would it make sense to assign the nooks static ip addresses at the bottom of the range? <br />I don'thave any real data, but jc's search times sound similar to what i was observing with nooks and a galaxy note tablet

[wgnoyes]

It might (have to figure out how; doesn't appear to even be an option in the graphical nook manager/rootnscoreit package), but even if available, that again is the kind of propeller-head stuff that normal average people aren't going to know the first thing in the world about how to do, and it'll just further damage the perception (beyond the hassle of rooting) that nooks are a readily-available platform to use for practiscore.

[jcwren]

Would it make sense to assign the nooks static ip addresses at the bottom of the range? <br />I don'thave any real data, but jc's search times sound similar to what i was observing with nooks and a galaxy note tablet

You can't assign NOOKs a static IP address. However, if you have control of the router, you can redefine the DHCP address range to be .1 to .253 and put the router at .254.

I would advocate that what PractiScore should do is take the IP address of the device itself, subtract 32, and start scanning from there. Given a typical range environment, it's unlikely that there's going to be that many devices on the network to start with, and that you'd have 32 Android PractiScore devices present. By doing this, you still scan the entire address range, but you pick a more logical starting point.

[Steve_P]

Ahh, got it. I hadn't looked yet to see if static ip was even possible. <br /><br />Bill, i think you are right about not making the nooks seem any more difficult than necessary.

[wgnoyes]

I tried putting in a lower ip addr range to the dhcp server specs of the tp-link router and it wouldn't take ANYTHING other than what was already there. When you say .1 to .253, is that the low-order 3 digits, as in 192.168.0.1 to 192.168.0.253? How would setting the router's addr to .254 help things?

[jcwren]

I tried putting in a lower ip addr range to the dhcp server specs of the tp-link router and it wouldn't take ANYTHING other than what was already there. When you say .1 to .253, is that the low-order 3 digits, as in 192.168.0.1 to 192.168.0.253? How would setting the router's addr to .254 help things?

Yea, the router has to have a static address, and it needs to be outside the DHCP pool. I think the TL-MR3040 comes up as 192.168.0.1, so you could set the DHCP pool to .2 to .254, or something like that. Although there's really no need to have 253 addresses available for devices, so you could reduce the pool to .2 to .33 or something like that.

[wgnoyes]

Well his gateway addr is 192.168.0.1, and the pool is .100 to .199. Lease time is 120 (1~2880), no domain specific, primary and secondary dns 0.0.0.0

Changed gateway addr to .254 and lan ip addr also to .254. Still handing out .100-.199. 2 nooks attached with addrs of .100 and .101. In practiscore 1.2.5 device sync, spinning circle..... 5 minutes and counting, nothing. I reset it all back to factory.

Edited October 26, 2013 by wgnoyes

[Graham Smith]

OK, I may be completely missing something here, but the first problem I run across is that once the Nook is out of the range of the router, it disconnects and won't reconnect when it's back into range without going into the settings and telling it to look again.

As far as IOS seeing the Nooks, I've had no luck with that either using two different routers and two different IOS devices.

There's either something missing here in my setup or others are doing something I'm not doing.

[euxx]

Graham, could you share with me what are IP addresses on your Nooks and on master device?

Also, how long did you wait on the sync screen? There is an indicator on the application action bar that scan is still in progress.

As for wifi re-connect. Have you tried to tap "battery" icon on the system titlebar on top? It should bring up quick wifi settings where you can turn wifi off and then back on. It should reconnect.

[Graham Smith]

Graham, could you share with me what are IP addresses on your Nooks and on master device?

The router is 192.168.10.200 and the DHCP is 192.168.10.201-225

As for wifi re-connect. Have you tried to tap "battery" icon on the system titlebar on top? It should bring up quick wifi settings where you can turn wifi off and then back on. It should reconnect.

Yes, I know that works, but then that defeats the entire idea of being able to walk into a bay where people are shooting and pull the scores from the Nook to the Master without having to fiddle around with things. If I've got to take the Nook from the scorekeeper and fuss with it, then taking a minute more to key in the Sync code is no big deal.

What I've been trying to figure out is how Bill is able to just walk into any bay with an IOS and have it automatically find the Nook without him having to fiddle with anything on either machine. He says it works but I'll be darned if I can make it work.

[wgnoyes]

Dunno no. I've previously been using a old linksys wrt54g router set to stone cold factory settings; I didn't even change the SSID, and I see nooks fall off the air when they're taken deep into a bay where the berms get in the way, and automatically pop back on air when brought back uprange. When we ran our first two practiscore matches this year, Chris Wren brought out his outdoor-grade wireless routers and extenders and ran them up flag poles for range-wide coverage, and as you might imagine, everything stayed online then.

The rooting system I've been using for awhile is the graphical nook manager, and now I'm using a rooting system that Chris Wren has developed called RootnScoreit, which is based on the same system; it just does everything for you.

What amazes me is that I can just centrally locate the old linksys router along a row of stages and it picks up most of the nooks up and down the range.

Keying in the sync code all the time would be a non-starter for me. (Having to do it once per device to provision the nooks at the start of the match is inconvenient, but not a disaster.) So would having to manually reconnect devices to wifi all the time. The latter would be a complete show-stopper in my opinion and leave me with ios as the only viable option for a scoring platform. Put it this way: if I were told that my RO's had to put up with repeatedly stopping what they were doing to manually reconnect devices to the network in order to sync scores during a major match, then that match would be scored instead with paper and ezwinscore.

[wgnoyes]

...

Also, how long did you wait on the sync screen? There is an indicator on the application action bar that scan is still in progress.

...

I don't know about Graham, but I can answer that question from my own testing. 5, sometimes 10.... MINUTES! (Hitting the refresh "button" doesn't help.)