BBB ships more than 100K units -- and soon with Debian!

Hi John,

What good news (all of it!). I think the "simplified programming" may refer to a C compiler that was completed a while back (I have not tried it), but I hope it also includes a framework for communication between the PRU code and the user program(s) on top of Linux. The JavaScript tutorials thing is great, since Node.js makes prototyping so fast (I've not used Python so I can't compare - but at least with JavaScript people can reuse the learnt knowledge in html files too).

Excellent news, and yes as shabaz wrote, every part of it is good!

And very well done Jason Kridner for making the necessary adjustments in software support.  I'm sure the message was arriving loud and clear from multiple quarters.

Morgaine.

Interesting thread here https://groups.google.com/d/msg/beagleboard/WYrk-JUAkbM/EdXVSEiU6oYJ where Geralg Coley confirms it.

For me, it says a lot when Jason doesn't participate - seemingly at all - in the community surrounding his creation, announces Debian on a third party website and leaves it to Gerald to confirm it after someone posted the same link as John did. I'm beginning to wonder what part Jason actually plays in Beagle* these days as it seems to be left to Gerald, the hardware guy, to take all of the flak. It also seems that Robert Nelson has become the defacto software leader and the move to debian is likely largely down to his influence.

The rest of that thread is worth a quick read, we're clearly not the only ones seeing problems with angstrom.  Anguel's reply made me laugh, and before someone suggests it, no it's not me posting in disguise

selsinork wrote:

 

For me, it says a lot when Jason doesn't participate - seemingly at all - in the community surrounding his creation, announces Debian on a third party website and leaves it to Gerald to confirm it

I don't actually understand what's going on with Jason.  I don't even have any wild theories. (*)  If I scratched my head every time that Drew talks to Jason yet Jason ignores E14 almost entirely, I'd have no scalp left.  I have to guess that we're missing the kingpin entirely, otherwise some of this would have fallen into place by itself.  It hasn't.

Coder27 is good at piecing together seemingly disconnected pieces of a complex puzzle.  I wish he were here.  And conversely, the value of us being here has decreased in his absence.  Hopefully there's going to be some movement on that when Sagar and colleagues return from their holidays, otherwise things are looking bleak.

Anyway, good for Beagleboard.org and BBB.  Despite lack of explanations, they actually do the right thing for the community in the end, and hence for themselves too.  E14 could learn from that.

Morgaine.

(*) Maybe it's nothing more complex than simply not being talkative.  It's unlikely that he's doing "the right thing" just by accident --- he must be doing the listening part of dialog very effectively.

As we reach the end of the year and consider New Year Resolutions and what things are likely to change, it occurs to me that one area which has been shrouded in total silence is the future of the BeagleBone family.  There hasn't been even a hint of where Beagleboard.org would like to take it.

Viewed commercially, BBB is clearly a huge success and the beancounters probably want to milk it for all its worth, but 2014 is going to be a time of massive competition from numerous parties using the A20, and hopefully Jason is aware of this.

I'd like to see additions at both ends of the range:

• A high-end OMAP5-based BeagleBone with SATA and gigabit Ethernet.
• An even more paired-down BeagleBone to compete in the sub-$25 niche.

Neither of the above would (nor should) replace BBB, which I think dominates an excellent hilltop of its own.

Morgaine.

selsinork wrote:

 

Interesting thread here https://groups.google.com/d/msg/beagleboard/WYrk-JUAkbM/EdXVSEiU6oYJ where Gerald Coley confirms it...

 

The rest of that thread is worth a quick read, we're clearly not the only ones seeing problems with angstrom.  Anguel's reply made me laugh, and before someone suggests it, no it's not me posting in disguise

Thank you for the "State of BBB Software" link.  It pretty much confirms what I've been saying: having a reliable OS infrastructure is a problem for Beagles, and the thread has lots of comments about "it doesn't work" versus "it works fine", but I didn't see any reference to a "blessed" stable version that is recommended, like RasPi does.  This is something RasPi does do better.  It's not that hard to do, but does require someone with the knowledge and will to set it up and maintain it.

I used to spend a lot of time at Google Groups, especially in the BeagleBoard group.  That all ended it 2012 when Google updated Google Groups, and reading the BeagleBoard group became more bother than it was worth.  I don't think users were asking for any major updates, but Google got the idea that Google Groups needed to be updated to look more "modern" (or something) to attract more users (or something).  So they added a bunch of flashy stuff that made it harder to use and made it really hard to find previous comments, which is pretty ironic for the world's leading Search company.  I suspect lots of users went elsewhere.  Stop me if this sounds familiar. 

I think the problems with New Google Groups helped element14, which has had some terrific BBB content added over the last year, especially from shabaz.  This is in spite of BBB content being hard to find and follow because of the baffling Knode structure, about which Morgaine has commented at length.  I don't think the recent element14 update has helped this any.  I'm hoping that we'll recover from the Update, because element14 is a great resource and even with the latest changes it's a lot better IMO than New Google Groups, particularly because you can add other kinds of content like 'blogs.

Happy New Year to our friends in Australia and New Zealand!  Still 12 hours to go here on The Coast.

John Beetem wrote:

 

I don't think users were asking for any major updates, but Google got the idea that Google Groups needed to be updated to look more "modern" (or something) to attract more users (or something).  So they added a bunch of flashy stuff that made it harder to use and made it really hard to find previous comments, which is pretty ironic for the world's leading Search company.  I suspect lots of users went elsewhere.  Stop me if this sounds familiar. 

I'm a sucker for clarity and stating the obvious, so I'll grab my crayons and underline yours:

• Google Groups "upgrade": FAIL
• Element14 Jive "upgrade":  FAIL

Companies won't stop doing what their webbies think is "cool", that is the nature of the universe.  Not all the combined might of the PhD legions at Google and their billions of petty cash can alter this fundamental truth:  webbies are out of control, because they are governed by "cool". (**)

Just doing things because they seem nice to you is an EPIC FAIL at the very first step.

There may be a few people in webbie teams recognizing that there is some truth in the above, and wondering how to improve things.  It's easy, just follow the most basic concepts of engineering:  check requirements, choose a solution that ticks the most boxes, implement, measure results against requirements, and repeat with new requirements addressing the observed deficiences.

Let me repeat again: measure results against requirements.

Let me repeat again: measure results against requirements.
Let me repeat again: measure results against requirements.

You may notice the similarity with the Scientific Method here.  It has brought us down from the trees.  Use it.

Morgaine.

(**) This is just shorthand and hopefully it's obvious, but perhaps I ought to explain the intended meaning anyway.  Working towards "cool" means heading like a moth directly towards the brightest light, working without a process, just doing things because they seem like a good idea at the time.  It is the opposite of engineering, which evaluates everything at every stage --- measurement is the single most important word in the vocabulary, as it enables the negative feedback that corrects errors.  We all delude ourselves if we try to work towards complex goals from the seat of our pants --- the end result is always brown pants.  The problem with software is that it's so damn easy to do simple things with it that we're deluded into thinking that complex things can be handled that way too.  Wrong.

John Beetem wrote:

 

I used to spend a lot of time at Google Groups, especially in the BeagleBoard group.  That all ended it 2012 when Google updated Google Groups, and reading the BeagleBoard group became more bother than it was worth.

I'll just say that I hate the web interface to googlegroups too.  However, there is another way. They really make it hard to find the info, but all of the googlegroups are setup as mailing lists, so you can send an email to groupname+subscribe@googlegroups.com and have the discussions delivered to an email address of your choice. That's what I do and just read them in Thunderbird in threaded mode.

As you probably know, I'm not a big fan of web forums, be that e14/jive, the RPF forums, or google groups, it's my belief that mailing lists are better and also much easier to deal with. The beagle* stuff goes wrong even there by having a single google group that covers all of the boards and everything from user support to kernel development. It doesn't work and it shows in the chaos there currently where most of the noise is little to do with the BBB, mostly basic "I don't know how to use a computer at all" or "I don't know how to use application X". The RPF has been forced to outgrow that stage by 2million+ users and at least they have some sensibly segregated areas for different sorts of discussion.

selsinork wrote:

 

Interesting thread here https://groups.google.com/d/msg/beagleboard/WYrk-JUAkbM/EdXVSEiU6oYJ where Geralg Coley confirms it.

 

For me, it says a lot when Jason doesn't participate - seemingly at all - in the community surrounding his creation, announces Debian on a third party website and leaves it to Gerald to confirm it after someone posted the same link as John did. I'm beginning to wonder what part Jason actually plays in Beagle* these days as it seems to be left to Gerald, the hardware guy, to take all of the flak. It also seems that Robert Nelson has become the defacto software leader and the move to debian is likely largely down to his influence.

 

The rest of that thread is worth a quick read, we're clearly not the only ones seeing problems with angstrom.  Anguel's reply made me laugh, and before someone suggests it, no it's not me posting in disguise

Gerald is a monster at replying to the mailing list. If he answers the question, then why do I need to chime in? I'm frequently poking around for unanswered questions and trying to help out rock-star developers like Robert help the Beagle community move ahead. It has always been my objective for the community to survive on its own, but this is my life and my passion, so you I'm not running from the flak. If there is something you see I'm not answering, it is fine to call me out specifically. If anything isn't running right, I'm most certainly at least partially to blame.

selsinork wrote:

 

John Beetem wrote:

 

I used to spend a lot of time at Google Groups, especially in the BeagleBoard group.  That all ended it 2012 when Google updated Google Groups, and reading the BeagleBoard group became more bother than it was worth.

I'll just say that I hate the web interface to googlegroups too.  However, there is another way. They really make it hard to find the info, but all of the googlegroups are setup as mailing lists, so you can send an email to groupname+subscribe@googlegroups.com and have the discussions delivered to an email address of your choice. That's what I do and just read them in Thunderbird in threaded mode.

 

As you probably know, I'm not a big fan of web forums, be that e14/jive, the RPF forums, or google groups, it's my belief that mailing lists are better and also much easier to deal with. The beagle* stuff goes wrong even there by having a single google group that covers all of the boards and everything from user support to kernel development. It doesn't work and it shows in the chaos there currently where most of the noise is little to do with the BBB, mostly basic "I don't know how to use a computer at all" or "I don't know how to use application X". The RPF has been forced to outgrow that stage by 2million+ users and at least they have some sensibly segregated areas for different sorts of discussion.

I've personally always used as an email list with beagleboard at googlegroups.com being where I send the e-mail. You have to set your Google Groups account to send you the mail.

We miss you John. I wish Google didn't do what they did, but I haven't seen an honestly good replacement that keeps people who prefer to interact with everything as e-mail happy.

That said, my inbox has been flooded for months. If the message doesn't specifically CC me, I probably miss it. I look for beagleboard e-mails that don't have replies first or ones that clearly mention a topic in the subject where I feel I'm an expert (or at least relatively knowledgable).

Some of these things I feel must organize themselves. I find that very few of the problems people have with BeagleBoards are specific to Beagle at all---they are more to do with Linux, or real-time, or programming, or electronics in general. I feel that Beagle has a huge opportunity to make it easier for people to learn about all of those topics and more in a friendly manner, but I feel some other solutions than forums, such as books, interactive websites and in-the-box-software make better long-term impacts. I'm very interested in feedback on that and would encourage you to call me out in a discussion on that in the Beagle mailing list (or less preferably here).

Community-sourced efforts like element14 forums and http://beaglefu.com/ are very important in my mind to get answers to people in the ways they feel the most comfortable. We've tried to make the Google Group solve as many people's needs as is possible, but I see these other efforts as a very natural part of a large and growing ecosystem of developers and users.