Is Leaf and the Maple dead or dying?

LeafLabs Garden » Community » The Lounge

Is Leaf and the Maple dead or dying?

(158 posts) (25 voices)

Started 3 years ago by corguy
Latest reply from raymadigan

12…11 Next »

corguy
Member

There seems to be relatively little activity on the forums, most of the posts are days if not weeks old, there are only a handful of regular posters, there is NO support from and very little activity from the founders. I don't see any new derivatives being talked about by users. I'm working on a VET6 pcb but without support(and interest) I'm going no where fast.

I'm not being pessimistic but it seems like from the outside that Leaf has run into a brick wall. I've seen a lot of open source projects at this point and either they hit the afterburners and get new life from users or they just fade away. Is that where Leaf is or am I missing something?

Posted 3 years ago #
gbulmer
Moderator

corguy - I think I understand what you are saying.

I don't think it is fading yet, but that's just my opinion, I don't have any deep insight. I am not a member of LeafLabs staff.

One of the LeafLabs folks, poslathian, mentions their hardware plans on http://forums.leaflabs.com/topic.php?id=1204#post-7312

IMHO, Maple is competing with mbed, Arduino Duo, ChipKit, etc. as well as manufacturers products like ST's DISCOVERY boards, NXP expresso, Cypress, Atmel, TI etc.
IMHO, it is the software which differentiates those products.

There are some cheap 'semi-clones', based on Maple, so I assume there is interest and demand.

On the other-hand, Oak might offer capabilities which are several orders of magnitude more interesting (to me) than Maple and its competitors. I have been looking forward to Oak since 2010, so I will be pleased to see it materialise.

ala42 posts about a first cut at an F2/F4 bootloader and libmaple http://forums.leaflabs.com/topic.php?id=1268&page=2#post-7828

Several of us are working on new and alternative boards. Siy discusses one a little on threads including http://forums.leaflabs.com/topic.php?id=1268.

Several of us have made boards, my first Maple-ish thing is at http://ourduino.wordpress.com/2011/04/06/orone-cortex-m3-robot-controller-stm32f103/
so maybe we can help.

Poslathian has posted on a few threads that future LeafLabs 'maple-ish' boards will have a second MCU to give hardware debug support on board, so I am assuming the current boards are dead-ends.

For my part, I have started experimenting with STM32F4-DISCOVERY, which is an interesting piece of hardware. It includes hardware support for ARM's Serial Wire Debug (SWD) protocol, which is supported by 'proper' debugging software. This is the same (conceptual) approach as poslathian is proposing. Several other folks have posted that they are experimenting with the same board. (The STM32F4-DISCOVERY is currently stupidly cheap, I believe it is a marketing focused product.)

My goals are to make STM32F4/STM32F? boards to use to enable children and adults to enjoy programming. I can use the STM32F4-DISCOVERY for a while, and I've sketched out daughterboards to make it easier to use for experiments with groups of beginners.

... I've seen a lot of open source projects at this point and either they hit the afterburners and get new life from users or they just fade away. ...

Yes, I have seen the same patterns :-(

Maybe forum members would like to collaborate to help each other push on further? ala42 has certainly contributed a valuable step forward.

Posted 3 years ago #
siy
Member

Well, I'd like to be involved into collaboration. Perhaps it worth to try choose some common short/mid-terms goals which would present interest for most of members. This would help coordinate efforts.

Posted 3 years ago #
bubulindo
Member

I think the main thing missing in the Maple to get more people interested is the same libraries people use in the Arduino. However, actually creating libraries that make full use of Maples hardware would be a best goal. If there is a library that does something in Arduino, chances are that people won't need to use a Maple. But the things that the Maple can do that few others can, that is where the focus should lie.

I have created a library for the CRC unit inside the chip and together with Rod wrote a library to use the RTCs. However, having some set goals would definitely make things easier and faster to achieve and maybe revive the community/forum.

Posted 3 years ago #
siy
Member

Completely agree with both items. It somewhat painful to move from Arduino to Maple just because lots of popular hardware requires porting of libraries. It's doable, but requires time, efforts and some knowledge. What is worse is that such a library designed for limited hardware capabilities of AVR MCU's, so while it was completely acceptable for Arduino, start looking somewhat ... lets say 'limited' on Maple.
At present I'm playing with different kinds of graphic displays (OLED and LCD). When all displays which I have start working properly I plan prepare library (probably set of libraries) for them. If these plans will be successful, I'm going to publish results.

By the way. Perhaps as one of the first short term goals we need to create some kind of community code repository. I see at least two things which already can be published there - F4 port of libmaple+bootloader by ala42 and RTC support by bubulindo. Hopefully RTC support can be ported to F4 and merged into libmaple. It would be good starting point for further development.

Posted 3 years ago #
bubulindo
Member

Yeap... there were some changes done by me yesterday to allow the definition of the prescaler value. I still need to implement the destructor (stop clock) and at least document and clean the files.

One thing I will also add is the daylight saving time and maybe timezone support. One extra thing would be a recurring alarm that you can set with a given period (DAILY, WEEKLY...). Although this last one is probably of no interest to most.

By the way, I have the CRC32 ready, but I'm still reading on the types of CRC32 to complement the library.

Posted 3 years ago #
LarryP
Member

Greetings all,

Not that this fact will save them from competition in the market, from what I understand, the leafLabs' crew are all MIT students (whether undergrad or grad students, I don't know.) Given this, it's not unreasonable to expect them to be busy studying at a rather demanding school.

However, that's a reason, not an excuse that the market will care much about. IMHO, leafLabs has to go long, or go home. Otherwise alternatives, both those already out there, and the upcoming ARM-based Arduino Due, will garner most of the mind share and the market share -- to say nothing of the Rasperry Pi.

Personally, I'd like to see leafLabs succeed, though I think they need to focus on usability of their hardware and software, especially usability on Windows (miserable an O/S as it is), because the great majority of their prospective customers are not LINUX savvy, and probably won't try to climb that learning curve -- especially if one of leafLabs' competitors offers a solution that works -- and works consistently and reliably -- on the *currently-sold* version(s) of Windows. Once the ARM-based Arduino-Duo comes out, IMHO leafLabs had better have their competitive strategy well honed, in place, and executed very well indeed. Anything less may be their death knell.

Posted 3 years ago #
gbulmer
Moderator

I'd like to start by doing something short term. Partly, I'd like to understand how we can collaborate.

How many folks have STM32F4-DISCOVERY boards? Those are cheap, straightforward to get, and have some fun hardware. If folks want to design their own boards, that's fine too.

We could set some goals to help ala42 test, fix, extend and document the STM32F4 startup and library code.

We could start off with small aims. For example, agree a subset of Wiring, and a couple of other peripherals beyond the base Wiring 'language'.

This needn't detract from LeafLabs work, and it might help.
When LeafLabs get there hardware out, the library will be tested and hopefully stable.

Maybe start a github repository so that we don't impact on ala42?

Would we need a 'paid for' github repository, or would a free repo be fine? If it helps us, I'd be willing to pay something for six months to see how it goes.

I'm not sure how LeafLabs would feel about us posting on this forum. Maybe they can comment?

Posted 3 years ago #
gbulmer
Moderator

LarryP - I can agree with much of what you say.
I think *WE* can choose what to do.

I think there is enough variety of features, capabilities and costs between manufacturers products, that there is interest for easy to use libraries for each major Cortex-Mx MCU family from each major manufacturer.

Currently STM32F is one of the dominant Cortex-M3 product families, so I think we are in a viable place. STM32F4 is significantly more powerful than Cortex-M3. I am focusing on Cortex-M4 because it appears to let me do even more interesting things.

IMHO, we could build some Easy To Use STM32F4 libraries to complement Arduino Duo. I don't think this has to be a competition with a single winner. I think several viable alternatives will encourage folks to progress beyond ATmega Arduino's.

We could initially focus on STM32F4-DISCOVERY boards, so we don't need to wait for any new hardware.

Edit: I think the CMSIS libraries are helpful for engineers who want to get at every capability of the underlying hardware, but IMHO CMSIS is a bit daunting for someone who is new to programming, or new to embedded programming. So something which makes it easier to get started would likely be helpful.

Posted 3 years ago #
Rod
Member

I, for one, am seriously into the Maple concept and would really like to see it move forward. My focus is low low power power (like really low, as in coin cells and 10 year product life). So a second chip for debugging support is a bad idea to me. I do like the FPGA idea in Oak, but with the F4 I no longer have a need for an external FPGA anywhere in my system.

(Actually, I'm driving 8 digital lines each at a 6 mHz data rate with a non-repetative pattern with a Maple Mini now. I would have thought I would need an FPGA for that, but no, an STM32 at 72 mHz does just fine and doesn't even introduce much for clock jitter.)

What I like with Maple is that the software is clean and simple (as opposed to opaque and battery eating as the ST stuff is) and offers me the opportunity of making my technology widely accessible via the Arduino model. The software is clean and simple, and the hardware is as well. Saying no to hardware clutter is a big deal.

And then, of course, I came here in the first place because I'd already done my home work and the the ST chip and ARM architecture just blew away anything else (such as the TMS430, Atmel, PIC, and on and on and particularly including other ARM implementations).

I'm for a community repository.

I did the low level RTC support for the F1 and I'm definitely interested in doing it again for the L1/F4. The big thing for me there is to make the deep sleep modes of the ST chip accessible. Its been brilliant to have bubulindo build the time of day stuff on top of that.

From a Leaf Labs point of view, I've done the start up thing (like 5 times) and I know there is a point where you just need to pay your bills. I don't have anything negative to say about people putting a focus on that. I was interpreting the reduced level of chatter as Leaf Labs pulling back a bit with the plan to release code closer to the time they could actually ship product using it. That way they can pay bills better.

It is time for community contributions if we are going to make them....

I'd actually like to include hardware designs as well as software. For example, I really like the Maple Mini. It lets me prototype systems very quickly without having to worry about bring the processor up on each new board design. I'd like to see an F4 version and an L1 version. Certainly the F4 one would need to be wider than a 40 pin dip, but that's no big deal with mill max SIP board connectors. I would be very interested in contributing tested designs and have Leaf Labs profit from manufacturing and shipping them.

Profit would let them spend more time contributing to the community themselves.

Posted 3 years ago #
gbulmer
Moderator

I feel both hardware and software contributions would be good.

Can we agree that, by default, everything will be the same license as the current Maple libraries?

Can we agree it should be git based, on github, to stay compatible with LeafLabs, and skip past the 'SCM religious wars'?-)

(Can we agree to use 'Zulu' (UTC, suffix Z) when we talk about time?)
Can we provisionally aim to get set up by end of day, 24:00Z, Friday 27th January 2012 ?

Posted 3 years ago #
corguy
Member

I really like the STM32 platform and I think the Maple/Leaf is in a great place to be a leader in the market. Yes they have a lot of competition with the RasPi and Duo but I really think the STM32 will make a difference and I think Leaf is well positioned.

But as LarryP said Leaf needs to step up. I can understand the need to pay the bills and finish school but the users did not ask them to create Leaf they chose to start Leaf and they will be ultimately responsible for the success or failure. The question is will they step up?

I think I know the key to successful open source projects because I've seen it happen over and over. Support from the project founders is key because it gets users excited and builds a community. Users can only do so much. I was working for a company in the mid nineties and we were open sourcing a lot of it’s processes. We started using MySql shortly after it’s release and we could email Monty (the MySql founder) and have answers at all hours of the day and night. These response’s continued well after Monty and MySql were a success and he had a team of people around him. We used Mysql for mission critical things for the sole reason we could count on the founder to provide support. This has been true of other projects I’ve seen as well.

Founders drive the community. I’m not seeing that here, Leaf put out a great product yes, but they have pretty much disappeared. If there working on the next great thing that’s fine but it’s really shooting themselves in the foot because there won’t be any users left when they release it. The users will have moved on to projects that are supported.

I asked a question that I think is easily answered over a month ago by one of the Maple guys. I got one response from a helpful user but was not able to get anywhere. I'm sure I could spend a day or two and come up with an answer but at this point I’m ready to move on unless I see some changes in activity here and I’d wager that others are thinking the same thing or have already moved on. Because while I might be able to resolve this issue I'm sure there will be more and I'm not interested in doing that with zero support from Leaf.

Posted 3 years ago #
gbulmer
Moderator

corguy - Founders drive many projects, but sometimes changes happen.

What are the alternative options today? What are your criteria?

I agree, it would be great if LeafLabs spent more time on the forums. But they haven't. I feel folks should raise issues, or it won't be clear that there are issues to be solved.

The lack of visibility of Leaflabs work is an issue for me. It would be easier for me to plan my activities if I knew what there development schedule is, and what they are working on. It would be even better if there were a well understood and open development process. There isn't. But I don't intend to give up because of that.

I don't feel my relationship with LeafLabs is a standard commercial relationship. I don't think "LeafLabs are the supplier, and LeafLabs should spend time supporting me".

Instead I think "Wow! This has moved me on a big distance; way further than I could have got alone. This has many more of the ingredients I need than anything else I can find. The license gives me enough freedom that I can use those ingredients to help me."

I am focused on people learning to program, and creating things for themselves. I've used Arduino, tinkered with mbed, and investigated several others (ChipKit, NXPresso, etc.). Leaflabs provide a lot more of what I need than any other project I am aware of, and less baggage.

At this moment, we can chose to move forward and build on what already exists.

ala42 and others have produced something to move us forward a significant step. If a few people would like to collaborate around STM32F, then we can move forward together, which might be much more effective, and more fun, than alone. We can hope that LeafLabs will bring the hardware, but in reality we can move forward based on the ST Micro Discovery boards and our own hardware design skills. I know of several STM32F4 designs in development, so I have little doubt that it'd get somewhere interesting. We may even help Leaflabs progress faster, which I think would be a benefit.

I have no problem with folks wanting a different support approach, and a different commercial arrangement. I am only saying there are alternatives available to us, and for some choices we may have enough control that success is in our hands.

For my $0.02, I like STM32F, and I am working to make some of the stuff work anyway. So I would welcome collaboration.

Posted 3 years ago #
StephenFromNYC
Moderator

Hello all,

corguy, I am very glad you started this thread. It is an interesting read, and I hope the resulting discussion increases the public visibility of the LeafLabs staff. My assumption and hope is that they are working quietly on the next "great" Maple/Oak board or software version.

The question you started a few weeks ago ("OpenOCD and JTAG") described a problem experienced with an "O*" Maple clone:

http://forums.leaflabs.com/topic.php?id=1231

Have you (or other O* Maple clone users) been able to get any support from the O* board seller?

I do not develop for Maple using Linux, but did you try kqr2's suggestion to update your entire OpenOCD config file?

I am guessing the LeafLabs team has purchased samples of the Maple clones (it is a good business practice to know your "competition"), but I am not surprised they did not share an OpenOCD config file (assuming they have one). I may be wrong, but I do not think creating an open source project obligates one to supporting competing projects/products.

As Rod said, it is important "to pay your bills" and by buying a clone you are not helping the LeafLabs group pay their bills. Many people want to stretch their purchases today by buying generic food (I like some "store" brand products), but there is also the saying "you get what you pay for".

It is not clear if your problem has been replicated using an authentic LeafLabs board. If the problem had been replicated with a LeafLabs board then I would have increased expectations of help from the LeafLabs staff.

Just my two cents.

Stephen from NYC (full disclosure: I am not a member of the LeafLabs staff)

PS- Lately, I have been very busy (with non-microcontroller) projects. No end in sight, the tunnel is dark.

Posted 3 years ago #
gbulmer
Moderator

StephenFromNYC - Lovely to hear from you. Constructive as ever :-)

Edit: Ooops - Happy New Year!

Posted 3 years ago #

RSS feed for this topic

12…11 Next »

Reply »

You must log in to post.