The KDE Improv Project Has Announced Its End

[kingu]

Always seemed to me like it was almost a success, to the point it just needed a little more, then ground to a halt, before it generated enough negative publicity and questions to keep it from working out.

[riklaunim]

Cubox-i exists with those SoCs and it's not a market hit. Most SoCs like this supports Android but when it comes to desktop Linux there are things missing, unsupported as not everything have been open-sourced etc. Comparing Celeron J1900 and all those ARMs there is a big difference if you want something for harder work on non-Android OS.

[0xBADCODE]

Where one fails, two new would appear.

This market is emerging and where one project fails, two new would appear and succeed. For example, aren't http://www.solid-run.com/products/hummingboard/ boards are neat in terms of size, price and features? Little, almost wrist-watch sized CPU module haves impressive set of interfaces, comparable to full-fledged desktops. And there're more boards like this. Its quite dynamic market and those who can do it fast, cheap and featured will be in advantage. So no need to cry - its just competition. Sorry, but these who can do it fast, featured and cheap about to take their places on the market. This is normal process for sure.

[xeekei]

Well he lives in Z?rich, which isn't exactly the cheapest place to live.

It's not like everyone in Z?rich have $200k to throw around.

Anyway, I hope this isn't a total loss. I like Aaron, and what the KDE guys are doing. I just prefer GTK applications.

[stiiixy]

I find this extremely hard to believe. The Raspberry Pi Foundation has sold over 3 million boards by now. If Olimex had sold more than that, we'd hear about their boards everywhere, but we don't. In fact there isn't really a community around these Olimex boards, they don't appear to be popular at all.

With the Raspberry-Pi getting all this unjustifiable attention, it's not hard to miss other products. You should look this project up.

It would seem that in our English-biased Western world, we miss a lot of the fun stuff that we used to do ('hand-made' software and hardware and not COTS) that the Eastern Euro's constantly cook up.

[genx]

Hi,

(disclaimer: the 3.3 mil is straight from olimex themselves, so take that as you will).

Can you provide a link or a quote for this ? I am quite surprised too.
I know Olimex for I own several smaller ?C boards from them and appreciate their products, but I cannot imagine such amounts of sales for this board.

As far as indivuals are concerned, there are (very roughly) 500 posts on Olimex A10 forum, whereas there are almost 500 thousands posts on Raspberry forum.
And as far as companies are concerned, I cannot imagine any company making a mass product (in the million range!) with this board (or any board). AFAIK, a quick redesign is always needed: add this port, remove those ports, modify the shape of the board, change the position of these connectors... and save the cost of the benefit of the original dev board maker (times n.10⁶).

[oliver]

I find this extremely hard to believe. The Raspberry Pi Foundation has sold over 3 million boards by now. If Olimex had sold more than that, we'd hear about their boards everywhere, but we don't. In fact there isn't really a community around these Olimex boards, they don't appear to be popular at all.

There actually IS a community around it, olimex has its own forums, linux-sunxi community also works with that community.

Appearantly, the lime has sold over 3.3 million allready. Probably not all to home users/makers though I guess?

And the r-pi has been around for 3 years now? So it took some momentum for it to be widespread and heavily used. Give it a while for makers to adapt a new board to be popular.

(disclaimer: the 3.3 mil is straight from olimex themselves, so take that as you will).

[brent]

Did you know that the Lime appearantly has sold more then the r-pi? ... Go figure.

I find this extremely hard to believe. The Raspberry Pi Foundation has sold over 3 million boards by now. If Olimex had sold more than that, we'd hear about their boards everywhere, but we don't. In fact there isn't really a community around these Olimex boards, they don't appear to be popular at all.

[oliver]

I really really do not understand how this could have failed. The Improv was a board around the EOMA68. The EOMA68 was based on the A10 and A20 SoC. The A10/A20 has an excellent and well supported community around it. The A10/A20 runs plain old linux (and android). There's Fedora for the A-series, there's debian, there's Ubuntu. Many many tablets (pengpod) that are really really cheap use A10/A20/A13 which are all very well supported by the linux-sunxi community and can run all the aforementioned distro's.

Yes they could use a lot of work in the driver department, but that aside, it works.

So they could have used the pengpod (and helped that project, which was really also just about re-selling cheap china tablets) or use any cheap china tablet based on these SoC's and be done with it.

But this was about opensource hardware? They where struggeling with hitting the numbers to do mass production? Did they even concider olimex? Olimex does opensource Hardware. They have several A10, A20 and A13 based dev boards, much like what the Improv+eoma68 are. The OLinuxino Lime is smaller then a r-pi and costs about the same AND opensource hardware! Yay.

So how could this have failed? How is this even possible? Why didn't the porting happen op a different A10 based platform (cubieboard/OLinuXino) and then source some tablet, transfer it to it, done. (I know it sounds easier then done )

With the previous post, about the eoma/improv thing going bust, I can think that's there where most of the resources got wasted.

How did this cost 200k from aaron alone? I don't understand. Did you know that the Lime appearantly has sold more then the r-pi? ... Go figure.

[riklaunim]

I do not understand why using ARM32 having now easier for GNU/LInux Intel x86 SoCs to make a competitive tablet?

When they started that project x86 didn't existed on tablets. They even used low end Allwinner A20 instead now minimal quad core Allwinner or Rockchip.

Also why do not them try to sell this project to Russia nad their new ARM64 "national SoC project" I am sure russians would be interested a lot in tablets that do not use Android MS WOS or IOS and are really open source.

There is nothing to sale. There is no patents or "intellectual properties". Mer is OpenSource.