I was about to come back to you with a 'there is a place to share code' rebuttal (http://gforge.chumby.com) but when I tried to go there, it's no longer online. Oh well, I was prepared to share code there, but it seems not enough others were for it to be worth continuing. In fairness to Chumby Industries, they have set up at least two systems in the past for developers to share code - loads of people shouted for it, but then when it (they ?) was made available no one (or not many) wanted to actually use it.

Wow - though I find the title a little harsh, I would have to mostly agree that there is something seriously wrong, but I think it goes beyond the website.

Where's the news?  What's been going on?  What's with the network, new devices, etc.?  I can't believe for a second that nothing has been happening.  And if so, well I assume that's not good news, or perhaps whatever was needed from the "free" developers has been fullfilled and there is no need for us anymore (speculating folks).  Oh, and no I don't really want to follow you on Twitter - I don't do Twitter - I do Chumby.  I want to come to the website and feel like I am involved in something, whether wearing my developer cap, or my Chumby fanatic one.  (You may want to update the featured developer as well ).

I personally don't think that "merging" the web systems together is really going to change any of that.  The problem is communication.  There is little, if any, outside of Duane helping people on the forums - which I assume he likes to do, but doesn't seem like something he should "have" to do.

I agree that another method of getting widgets on the Chumby would be a very good idea as well.  Of course I am partial to Flash, but I would like to see more depth, and variety.  I also think that the device is too "locked" down.  Yes, there are "baddies" out there.  But at least have some type of program for legitimate developers to push the hardware and functionality to the fullest.  There needs to be a system in place to reward good and dedicated developers if there is going to be no inherit ability to generate revenue.

The contests seem like a no-brainer.  It's not my place to speculate on Chumby business practices, but I know they do/did have funding, and even without a monetary reward, a free Chumby could have been enough to get people to really put some effort in.

Yeah - I used the word "been".  Unless there is some dramatic announcement, the bus has left in my opinion.  The widget output has been a trickle at best.  In the past Steve has mentioned that the device was a means to an end.  Well where is that end?  And does it include any of us developers?  The people who put the 1000+ widgets on the device to help give volume to the "network".

As I have stated before in the forums, if the idea is to become a "network", you're really going to need to invest (yes, money) in people who can do that for you.  And I don't mean just program - but come up with original compelling content.  I don't see that being a cheap or simple task.  If every television network played the same show, we would only need one station.

Another thing that has me very concerned is the amount of stories I get from Chumbyland users regarding the fact that without Chumbyland, the Chumby would have lived the rest of its days in a closet, or some other form of "burial".  I suppose this should be a compliment, but it worries me more than anything else.

Maintaining the Chumbyland community is insane (and I absolutely love it ) - there needs to be something similar in place for Chumby.  A community manager who can deal with more than just developer issues - and keep the site "alive".  I had offered to take on some type of role like that last year, but I suppose it wasn't in the cards.  But you should seriously consider someone who can do that for you.  Someone who is not wrapped up in code or hardware, but just a plain site-wide Chumby cheerleader (and no, I don't see Duane with a set of pom poms, shouting "give me a C", cutting it ).  And I also don't mean on a weekly or monthly basis, but everyday.

--

As far as some of the other items pointed out:

GForge - though it is down now, even when it was running it took quite a while to get things approved.  I lost my excitement pretty quickly and moved on.  Personally I think there isn't enough of a developer population to sustain it.  There probably never was in the first place.  But Chumby did the right thing by trying to do good by developers.  The people who were hammering for it should have been able to see this from the start.

Dev pages - there are "profile" pages but they don't offer much.  I have been asked if I have video, assets, etc. for a profile page.  I never provided any, so I assume the same is true for most others as well.  That is probably why they are not prominent.

Single log in - It certainly would make life nice, but it probably takes about .5 seconds to tap in my credentials.  Just sayin' .

--

In conclusion...I conclude that I don't actually have one .  Everything that I think currently is pure speculation.  It's about time for the Chumby folks to give us a real update on the state of things.

And as a final note, I have nothing but love for the Chumby and the people involved.  Users and developers alike.  I think the 1000's of hours that have been put into maintaining the Chumbyland community should be proof enough of that.  I only wish for longevity and success in the Chumby universe, in whatever form that may be.

If there is anything I can personally do to help in whatever manner, let me know.

Cheers.
cbreeze (develephant)

i have fixed the issue with gforge.chumby.com. 

/mdkail

The code repository issue is a bit of a sore subject for me.

I went to bat not once, but *twice*, of behalf of the rather vocal segment of the "community" to put up a community-supported source repository only to have all of the folks that advocated the most loudly suddenly vanish once it was made live.

cbreeze, I'm sorry that project approvals have been slow, however, we openly offered to allow members of the community be involved with that process and nobody stepped up.  For some reason, I wasn't getting email notifications, and I logged in to see what was up and approved the handful of projects that had accumulated.

At this point, I feel like a complete fool for advocating internally on this issue.

A good idea for how more people that don't know flash could make something for the chumby easily:

Canvas. You could make the chumby's screen like a canvas and have a way to upload the javascript canvas code instead of a .swf to chumby and it is pretty much the same as flash. Chumby would only need to interpret the javascript functions that have to do with the canvas.

Canvas: https://developer.mozilla.org/en/HTML/canvas

Thanks for all the updates!  I assumed the added stress of the economy is probably a huge factor for you all right now.

@Duane - I certainly wasn't meaning to knock the "forge" work, if that's how it came across.

Excuse the shameless plug, but the upcoming version of Chumbyland has more options like this.  And if you would like to be on the beta team - please let me know. 

Looking forward to the future.

Cheers.

One thing about single sign-on that I think warrants discussion - one problem with it is that if your credentials are compromised in one place, then they're compromised everywhere.

Given that we run a lot of third-party software for which we haven't run complete security audits (such as this forum), then does it really make sense to have all of your security riding on the quality of the weakest link?  In other words, if this forum software has a bug, then should your credit card details in the store potentially be in jeopardy?

This issue isn't isolated to chumby - in fact, the recent general trend is to move *away* from single sign-on in many high-profile hacking targets for this very reason.  Some online game companies now actually *want* you to use completely different credentials on their forums from the ones used by the games themselves so that actual game accounts don't get hijacked.

There are some possible solutions to these issues - using trusted authentication services and using systems such as oauth to manage credentials in a manner that reduces exposure.  Then, of course, your systems become slaves to a third-party, and unfortunately, there have been several high-profile incidents where such centralized services have failed for some period, disabling otherwise unaffiliated sites.

Then, of course, there's  the matter of trust with the central authority - that's what doomed Microsoft Passport.  Sites and users simply didn't want to trust them as a central authentication authority.  And trust comes in many forms - Facebook seems generally trustworthy (though Beacon raised a few eyebrows), but appears to burn through $200M a year, and that train runs out of steam at *some* point - I'd hate to lose all of our users' ability to log in if/when that happens.

As far as having a common header/navigation UI across the various sites, that's reasonable, though it's *far* more complicated than simply cutting and pasting HTML.  Chumby is currently expanding globally, and working with various partners to present customized, branded versions of the site for various products.  That means this stuff is increasingly dynamic, and sensitive at any given moment to lot of variables.

There's also an overall code management issue - when your upstream provider supplies a new version of the software, you need to now repatch it with your modifications and regress it to make sure you haven't broken anything.  In our current deployments of the forum and wiki, they're run more-or-less in stock condition, which makes us far less sensitive to version incompatibilities with patches.  Since going live, we've updated the forum and wiki several times each, with virtually no downtime.

As cbreeze (aka develephant) noted, the effort to create accounts on these services takes very little additional time, and personally, I think the firewalling of these systems has security advantages that I feel outweigh convenience.  One thing that's interesting is that our early users, who were often hackers, when to some effort to create custom credentials for each service, some with customized email addresses - probably to try to catch us in privacy violations or whatnot.  Hackers in general are very suspicious of centralize authentication services.

Please feel free to comment and/or disagree - we're open on this issue.

It is 3796 bytes when i compressed it. I am new to javascript so it is probably not that efficient. To try it, you have to include it in a page with "<canvas id="canvas" width="320" height="240"></canvas>". It does everything the other one does. Here is a screenshot in firefox: http://www.flickr.com/photos/33883927@N04/3681549901/ Here is the script uncompressed:

var canvas = document.getElementById('canvas');
if (canvas.getContext) {
  var ctx = canvas.getContext('2d');
  //colonStatus = if the colon is visible
  var colonStatus = false;
  //segPos = positions of segments in a digit. 0=x,1=y.
  var segPos = [[-1,-12,13,0,-13,12,-2],[-29,-15,-18,0,13,16,29]];
  //neededsegs = segments needed to build digit or letter (1-9,a,p,-)
  var neededSegs = [[0,1,2,4,5,6],[2,5],[0,2,3,4,6],[0,2,3,5,6],[1,2,3,5],[0,1,3,5,6],[0,1,3,4,5,6],[0,2,5],[0,1,2,3,4,5,6],[0,1,2,3,5,6],[0,1,2,3,4,5],[0,1,2,3,4],[3]];
  //hourmode = current mode (12/24). true = 12, false = 24.
  var hourMode = true;
  function changeMode() {
    if (hourMode == true) hourMode = false;
    else hourMode = true;
  }
  function clearCanvas() {
    ctx.fillStyle = 'rgb(153,153,102)';
    ctx.fillRect(0,0,320,240);
    ctx.fillStyle = '#000000';
  } clearCanvas();
  function drawSegment(segment,x,y) {
    ctx.beginPath();
    if (segment == 0) {
      ctx.moveTo(x+15,y-5);
      ctx.lineTo(x+8,y+5);
      ctx.lineTo(x-4,y+5);
      ctx.lineTo(x-15,y-2);
      ctx.lineTo(x-15,y-1.7);
      ctx.lineTo(x-14.8,y-2.2);
      ctx.lineTo(x-14.6,y-2.7);
      ctx.lineTo(x-14.3,y-3.2);
      ctx.lineTo(x-14,y-3.6);
      ctx.lineTo(x-13.8,y-3.8);
      ctx.lineTo(x-13.2,y-4.3);
      ctx.lineTo(x-12.5,y-4.7);
      ctx.lineTo(x-11,y-5);
      ctx.closePath();
    } else if (segment == 1) {
      ctx.moveTo(x-4.2,y-14);
      ctx.lineTo(x+5,y-8.5);
      ctx.lineTo(x+4.5,y+8.6);
      ctx.lineTo(x-2.5,y+14);
      ctx.lineTo(x-5,y+11.8);
      ctx.closePath();
    } else if (segment == 2) {
      ctx.moveTo(x-4.4,y-5.2);
      ctx.lineTo(x+3.3,y-16);
      ctx.lineTo(x+4,y-15.6);
      ctx.lineTo(x+4.7,y-15);
      ctx.lineTo(x+5.2,y-14.4);
      ctx.lineTo(x+5.6,y-13.7);
      ctx.lineTo(x+5.8,y-13);
      ctx.lineTo(x+6,y-12.2);
      ctx.lineTo(x+5.1,y+13.3);
      ctx.lineTo(x+1.3,y+16);
      ctx.lineTo(x-5,y+10.9);
      ctx.closePath();
    } else if (segment == 3) {
      ctx.moveTo(x-13,y-.7);
      ctx.lineTo(x-7,y-5);
      ctx.lineTo(x+7.5,y-5);
      ctx.lineTo(x+13,y-.5);
      ctx.lineTo(x+6.8,y+4);
      ctx.lineTo(x-7.3,y+4);
      ctx.closePath();
    } else if(segment == 4) {
      ctx.moveTo(x-1.7,y-13);
      ctx.lineTo(x+5,y-7.3);
      ctx.lineTo(x+4.5,y+9.8);
      ctx.lineTo(x-5,y+13);
      ctx.lineTo(x-4.2,y-11);
      ctx.closePath();
    } else if (segment == 5) {
      ctx.moveTo(x+1.7,y-16);
      ctx.lineTo(x+5,y-13);
      ctx.lineTo(x+4.3,y+13.7);
      ctx.lineTo(x+3,y+16.5);
      ctx.lineTo(x+2.6,y+16.8);
      ctx.lineTo(x-5,y+7.2);
      ctx.lineTo(x-4.4,y-11.2);
      ctx.closePath();
    } else if (segment == 6) {
      ctx.moveTo(x+7.6,y-5);
      ctx.lineTo(x+15,y+3.8);
      ctx.lineTo(x-12.1,y+4);
      ctx.lineTo(x-13,y+4);
      ctx.lineTo(x-14.3,y+3.3);
      ctx.lineTo(x-15.2,y+2.5);
      ctx.lineTo(x-15.6,y+1.8);
      ctx.lineTo(x-15.9,y+1.8);
      ctx.lineTo(x-16,y+.4);
      ctx.lineTo(x-15.9,y-1.7);
      ctx.lineTo(x-6,y-5);
      ctx.closePath();
    }
    ctx.fill();
  }
  function drawDigit(digit,x,y) {
    var len = neededSegs[digit].length;
    var segment;
    for (var i=0;i<=len;i++) {
      segment = neededSegs[digit][i];
      drawSegment(segment,x+segPos[0][segment],y+segPos[1][segment]);
    }
  }
  function getDigits(number) {
    var tens = 0;
    while (number >= 10) {
      number -= 10;
      tens++;
    }
    //returns the digits of the two-digit number.
    return [tens,number];
  }
  getDigits(76);
  function setTime() {
    clearCanvas();
    //colon
    if (colonStatus == false) {
      ctx.beginPath();
      ctx.arc(141, 97, 4, 0, Math.PI * 2, false);
      ctx.arc(139, 117, 4, 0, Math.PI * 2, false);
      ctx.fill();
      colonStatus = true;
    } else {
      ctx.fillStyle = 'rgb(153,153,102)';
      ctx.fillRect(134,92,12,30);
      ctx.fillStyle = '#000000';
      colonStatus = false;
    }
    //time
    var d = new Date();
    var h = d.getHours();
    if (hourMode == true) {
      ctx.scale(.4,.4);
      if (h > 12) {
        h -= 12;
        drawDigit(11,486,164);
      } else {
        drawDigit(10,607.5,205);
      }
      ctx.scale(2.5,2.5);
    }
    var hours = getDigits(h);
    if (hours[0] != 0) { drawDigit(hours[0],74,103); }
    drawDigit(hours[1],114,103);
    var minutes = getDigits(d.getMinutes());
    drawDigit(minutes[0],168,103);
    drawDigit(minutes[1],208,103);
    ctx.scale(.4,.4);
    var seconds = getDigits(d.getSeconds());
    drawDigit(seconds[0],592.5,310);
    drawDigit(seconds[1],637.5,310);
    var month = getDigits(d.getMonth());
    if (month[0] != 0) { drawDigit(month[0],200,402.5); }
    drawDigit(month[1],245,402.5);
    var day = getDigits(d.getDate());
    drawDigit(day[0],330,402.5);
    drawDigit(day[1],375,402.5);
    var year = getDigits(d.getFullYear());
    year[0] = getDigits(year[0])[1];
    drawDigit(year[0],460,402.5);
    drawDigit(year[1],505,402.5);
    //draw hyphens in date
    drawDigit(12,287.5,402.5);
    drawDigit(12,417.5,402.5);
    ctx.scale(2.5,2.5);
  }
  setTime();
  setInterval(setTime,500); //must be 500 for colon
  canvas.addEventListener('click',changeMode,false);
  [(-5),{0:5}];
} else { alert('Error: browser doesn\'t support canvas'); }

Yes, but most people don't look to see where it redirects to. All it does is make a UI element behave more predictably.

Duane and other chumby official people,

I understand you are tired and exhausted by users' endless complaint, and you have much to say why a chumby is a chumby in current way, however please don't try to negotiate or persuade customers, because you can't.

As a chumby fan and an IT engineer like you, please let me say this.
They didn't vanish, but they left, becuase their problem was not solved.
Please don't think their demand was a fake, but ask yourself if it wasn't another "faster horse" of Henry Ford.

In my personal opinion, chumby.com needs an official Yahoo-Pipes-like service (don't have to be so rich, but more flash-developer friendly and authentification supported), because crossdomain mess is a compensation for chumby's good security and Yahoo Pipes is not made for that. (I'm sorry, if it's *another* "faster horse" complaint.)

And Duane, you work too hard, seriously.

Thank you for the response.

I totally agree with you, but I am afraid it is also a reason why I feel uncomfortable developing a
widgets depending on Yahoo Pipes. If it's a such burden, can we expect Yahoo or other outer commercial or individual proxies to carry it when chumby users increase more and more? (And yes, I think Chumby One is very hopeful.)

Anyway, if it's only a matter of traffic, can't you just start from closed-beta? You can restrict widgets by API keys, limits target sites starting from ten or twenty major RSS sites, cache them on server for minutes. You may think about extending the service after killer application is created.

Of course, it might be another deserted project and I would be a only one to use it (I don't think it will happen), or too many people apply for it and complain about it anyway(this is more likely).
Either way, I think it is not meaningless for you to show us how do you recognize current security/usability balance of your service, and do you have a will to do something about it sometime in the future. It will be very relieving for widget developers including me.

Hello cbreez:

I am glad to hear from such a good widget maker as you.

I just thought it might be great if many flash designers out there would participate in chumby developers.
For me, creating a widget is a fun, writing PHP is OK, but establishing and maintaining proxy server for chumby users is a bit too much.
I'm afraid most of flash designers think the same way, because, in my impression, there are not very many people who can do both writing actionscript and server management like you.
(And, 'design' is very important part of chumby, isn't it? Chumby widgets are destined to be a part of a scene).
I also think practical and useful widgets (such as your gmail reader) should not rely on individual server if chumby is meant to be a popular equipment, no matter whether it is official or user created.

You may be right, I may be wishing too much, but I am not so sure about developing a widget is optional.
When I first met chumby, I got an impression that it is not just a hardware, but it is more like a service trying to realize a greater vision; "small but always-open window to the internet".
If my impression was right, user created widgets may be a main part of chumby's service, and opening
a door to more developers may benefit both developers and chumby industries. (Who would buy an iPhone without App Store?)

Ah... Am I saying this to such a good contributor like you?
Please excuse me if I sound too rude, but please understand that I want chumby to be a popular gadget like iPod so that I can recommend it to non-geek friends.
And I apologize, I should not have continued this stimulatingly titled post, but I could not ignore the gravity, you know.