Re: End of Chumby as we know it...

Well, I'm really in a race for time at this point.  My options for doing things to the existing system are somewhat limited.  Unfortunately, there's no easy way to put a message on all chumbys without some deep hacking on the Control Panels.  Perhaps I'll get to that when I'm confident I'm ready to take over the systems.

Here's where I'm currently at with the new system - I have CC, C1, and I3.5 devices booting and running the "Space Clock" (nee LCARS Clock) widget, and playing ShoutCAST, iheartradio, blue octy radio, and Sleep Sounds.  I still need to do CBS Podcasts (easy), and Pandora (hard).  All of the music sources that do not rely on the chumby servers run as before (NOAA Radio, iPod, My Streams, FM Radio, Squeezebox Server, My Music Files).

There were actually several other support services that needed to be implemented first to get this much running.

My next step is getting the "xapi" devices working (C8, I8).  They use a more complex and more extensive protocol.  It's *very* likely that I'll have to make some minor Control Panel modifications for these two.

Most of the old systems have been backed up - source code for the servers, Chumby-authored widgets and Control Panels, public GPL source, toolchains, and various support files used by the service.

Much of this stuff requires some level of archaeology - most of the internal systems that held all of the documentation are offline, and some of them were never transferred to Amazon.  In some cases, what I'm having to do is network traces to see what systems are going where for what, then tracking down the source for each end to see what's expected.  The other issue is that some of the systems are quite old and I'm having to rewrite a few of them to work with the latest versions of various frameworks.

My hope is to switch over to the new systems in mid-February, so we'll still have a couple of weeks with the old systems available in case I miss something crucial.  My biggest worry is that I miss some keys or certificates that are just sitting on a server and aren't backed up anywhere anymore.

Right now, I'm doing "git'er done" programming, and including the various third-party API keys directly into the code.  After the crunch, my plan is to factor out these keys, then I should be able to open source the new system, and people will be able to help if they want.

As for testing, I would certainly appreciate folks to assist when I'm ready - all you have to do is place a "hosts" file on /psp to redirect the various domains to the new systems.

I hope to have the new corporate entity in place this week, in which case I can start signing up for new hosting services to hold this stuff.

Re: End of Chumby as we know it...

That is amazing work.  Any you picked the clock I love and use  smile

Are any of the weather items or stock items included or available?

Re: End of Chumby as we know it...

The problem with the current stocks and weather widgets is that they require user configuration, which is stored on the server associated with the "widget instance" in a particular "channel" for a particular "user".  The means that I either have to restore all that initially, or alter the widgets to stored the data locally.

In addition, the NOAA-based weather widget requires an additional service to be working - the zipcode->weather station lookup.  Once the widget knows the weather station, it then goes to NOAA directly for the data.  Unfortunately, the NOAA servers are *very* slow to respond, which you can see in the NOAA Weather widget when it seems to take such a long time in a "Loading..." state.  The widget attempts to cache the data in Mobile Shared Objects, but due to a quirk in the Flash Player's MSO logic for widgets loaded from local file storage, it doesn't work a well as it should.

The stock widgets use a third-party web service for the stock data, so there's no additional service required.  Again, the issue is the storage of the configuration.

I certainly see the value in supporting these and other widgets and I'll undoubtedly get to those at some point.

However, we need to crawl, before we walk, before we run.

Re: End of Chumby as we know it...

Duane, Crawl before walk.....I totally understand that!   Just switching it all to work as you are is pretty amazing!

Re: End of Chumby as we know it...

Sounds like you are making some progress.
Is there anything we have to do to make the crossover to the new system or will it (hopefully) be automatic when logging on? Duane's new system will just take over the older server system?

On another note: My I8s are doing great but my I3s are having problems logging on to the servers. They want me to go through the wifi signin everytime I reboot. Every third time or so it finally goes through. It is supposed to logon automatically. Well things may clear up when Duane does the changeover.

Re: End of Chumby as we know it...

The goal is that you should not have to do anything, or at most, reboot.  The main change will be to the IPs to which the various domains point (www.chumby.com, xml.chumby.com, widgets.chumby.com, etc).  What will happen is that as the new IPs propagate, some devices will switch automatically, probably within 24 hours of the domain change.  Some devices might hold onto the old IPs until the original servers go offline, at which point they'll revert to the standard "off-line" clock, since they'll treat it as an outage.  Once they're rebooted, they'll come up using the new servers.

Similarly, in the case of C8 and I8, if there are some required Control Panel changes, it's possible the devices will get confused until they're rebooted.

We can actually force reboots of the CC, C1 and I3.5 devices from the old server, which we might do once we deploy.

Some other side effects - any alarms using "blue octy radio" or "Sleep Sounds" will probably have to be reconfigured, since the newly deployed stations will have different IDs than the ones from the old service.  One the other hand, I also took the time to fix or remove the stations that weren't working, and added a couple more.

Re: End of Chumby as we know it...

Hmmm, it looks like The Verge picked up the story.  Seems a bit more pessimistic than I'd prefer...

I'm going to open up forum registrations so new people can comment.  Hopefully the spam won't be too bad.

Re: End of Chumby as we know it...

is there any way individuals could host the download of the control panel locally?

Re: End of Chumby as we know it...

I *believe* that if you ssh into a working chumby, and copy /tmp/controlpanel.swf to a USB dongle, it should use that one instead of downloading a new one upon reboot.  However, that also means that if I change the Control Panel on the server you won't get a new one - so I wouldn't do that just yet, since I may have to put something out for this new system to work properly.

Re: End of Chumby as we know it...

but there'd be no way to netboot that way?

i realize any updates to the control panel would not be handled automatically that way, but i'm sure that could be dealt w/ in a variety of ways.

Re: End of Chumby as we know it...

I'm in for support/donation (sub model).

A weather widget would be nice in addition to the clock. I'm OK with only one weather widget.... same as you are looking to provide a single clock.

Good luck and thanks for your help.

John.

Re: End of Chumby as we know it...

Appreciate the continuing efforts Duane!  I originally came to Chumby as a lower cost alternative to Internet/WiFi radios and have never regretted my choice. I'm glad to hear that there may be continuing shoutcast support.

Re: End of Chumby as we know it...

Duane,

Thanks SO much for your help.  Have you set up a paypay account for donations for server time/etc.?  And/or could you set up a Kickstarter project?

Re: End of Chumby as we know it...

Just throwing an idea out: What about replacing the OS with Android?

Re: End of Chumby as we know it...

Let's rally together and keep the little guy alive!
Chum

Seriously though, there are nearly 4,000 Likes on the Chumby Facebook page.
Surely some of these people are willing to donate $10 or more to keep it going.

I have a lot of great memories with this little box and would hate to see it die.
Sure I could run something off of a USB stick, but it's just not the same.

Once you have something in place please let me know and I'll spread the news as best I can.

Long live Chumby!

Re: End of Chumby as we know it...

Duane, I couldn't be more impressed with your efforts so far.  Thank you SO much.   I hope your dedication will be noticed by a possible future employer.   We'd certainly all be references, though rather unusual ones!   Thanks Duane.

Re: End of Chumby as we know it...

jasonh wrote:

Just throwing an idea out: What about replacing the OS with Android?

Unfortunately, the devices do not have sufficient memory and CPU power to support any recent version of Android.  The closest we ever got was a version of Android 1.6 that kinda booted to the desktop but little else.

However, if anyone wants to make another attempt, I'm all for it and will help as best I can.

Really, my goal is to hold back the abyss - I'm hoping the community will engage to move forward.  I can't do it alone.

Re: End of Chumby as we know it...

Count me in as a Chumby fan!  I loved the little device, and I'm willing to donate to help.  Is there a link or paypal account you are using for donations? 

Keep up the good work and thanks for keeping Chumby alive!

Re: End of Chumby as we know it...

We're willing to donate as well. We will never remember the upcoming birthdays or wedding anniversaries without our Chumby!

Re: End of Chumby as we know it...

I would have thought that 40,000 devices could be handled by a rented host in a data centre.  As you say, a couple of hundred dollars a month.

Re: End of Chumby as we know it...

Duane wrote:
jasonh wrote:

Just throwing an idea out: What about replacing the OS with Android?

Unfortunately, the devices do not have sufficient memory and CPU power to support any recent version of Android.  The closest we ever got was a version of Android 1.6 that kinda booted to the desktop but little else.

However, if anyone wants to make another attempt, I'm all for it and will help as best I can.

Really, my goal is to hold back the abyss - I'm hoping the community will engage to move forward.  I can't do it alone.

what about local download of the control panel and content? kind of like a mirror? that could help save a lot on your bandwidth costs.

you'd need a way to specify an alternative boot host though.

Re: End of Chumby as we know it...

aris wrote:

I would have thought that 40,000 devices could be handled by a rented host in a data centre.  As you say, a couple of hundred dollars a month.

The current system has about 4TB of stored data and serves about 5TB of data per month to the current active 40K devices.  You can see the finances above.

As I said, I'm just trying to hold off the darkness - we'll revisit the rest once we've survived the Chumpocalyse.

Re: End of Chumby as we know it...

It would be sad to have Chumby go. It's a cosy, nice device, I'm used to having it near my bed.

What about the following plan:

1. Run a donation campaign to gather enough funds to run Chumby in the current form for a few more months.
2. This would give enough breathing space to implement a lightweight hosting platform that wouldn't need $4K to run.
3. Fund ways to reduce costs - perhaps somebody can volunteer VMs and bandwidth. Actually, if Chumby was running on physical hardware, how many servers would it need? What are Chumby's bandwidth needs? Some of us work in places which have lots of unused outgoing bandwidth and/or CPU resources. For example I can consider bandwidth donation, as long as it's not in gigabits.

Re: End of Chumby as we know it...

Duane wrote:
aris wrote:

I would have thought that 40,000 devices could be handled by a rented host in a data centre.  As you say, a couple of hundred dollars a month.

The current system has about 4TB of stored data and serves about 5TB of data per month to the current active 40K devices.  You can see the finances above.

As I said, I'm just trying to hold off the darkness - we'll revisit the rest once we've survived the Chumpocalyse.

5TB data per month is about 17 megabits/sec on average, if I'm not mistaken ... So a 200 mbps link might be enough, considering bandwidth bursts?

What is the percentage of static content ? Can it be hosted somewhere else ?

Re: End of Chumby as we know it...

You can see the current server structure from the second post:

4 Amazon EC2 "micro"
1 Amazon EC2 "small"
3 Amazon EC2 "large"
3 Amazon EC2 "x-large"

2 Amazon RDS "large"
1 Amazon RDS "small"

Storage: 4TB
Outgoing bandwidth - 5.1TB/month