chipKIT® Development Platform

Inspired by Arduino™

Refund?

Created Sat, 23 Jul 2011 18:21:20 +0000 by Mr_Fixit

Mr_Fixit

Sat, 23 Jul 2011 18:21:20 +0000

I am in the process of writing reviews on the Chipkit product lines for several publications. In this review, I am going to discuss the current state of the product line, and the support being provided. This article is going to try to be positive, but there are a large number of painful truths to be told here:

  1. 36.36% of the product works. No information provided as to when more functionality will be provided. ChipKit Library Status
  2. Support questions tend to go unanswered or simply ignored on the Chipkit site, and also when you contact Digilent directly.
  3. Advertisement states that this system is 100% compatible with the Arduino family, code, and most of the shields. Considering the first point above, this is completely false and misleading.

With all that said, I now have a last question: How do I get a refund for this paperweight? (Uno32)

ron.dunn

Sat, 23 Jul 2011 23:28:49 +0000

Interesting post. I share some of your concerns, but my result isn't as dire as yours.

I'm very unhappy with the lack of official support in the forum. I've received great diagnostic advice from a couple of users, I've muddled through a few problems, I've helped some others do the same, but getting the developers to comment on problems is like getting blood from a stone.

How hard can it be, for example, to provide a reference to the atan2() function, and/or a pointer to its source code? (self-interest here)

I do agree that marketing (or reviewers?) have run away with the truth. Everyone who worked on this platform knows that it is only partly compatible. Look at the recent discussion on the String class for evidence that deliberately non-compatible decisions have been made; as well as technically incompatible libraries. This perception needs to be corrected ASAP.

These problems aside, I'm quite happy with the board I purchased. It does work, my project didn't have dependencies on missing libraries, and is now running way faster than it ever did on Arduino.

My path from here, though, is probably to commit to Pic32 rather than to Max32. What I mean is that, given the support and library problems, I'm likely to convert any dependencies to true PIC code rather than using the Arduino compatibility layer.

I'm also likely to convert to the UBW32 board, but that is because its form factor is more easily converted to a production installation than a header-based interface like that shared between Max32 and Arduino. If I'd stayed with the Arduino platform, I'd probably have moved production to something like Teensy++ for similar reasons. Bonus ... EmbeddedMan (the UBW32) designer ... is actively supporting his product.

As I said in an earlier thread, a new product community can only be developed with active and enthusiastic support from its developers. Unless we start to see that, combined with some more accurate marketing, I'm afraid chipKit will die.

svofski

Sat, 23 Jul 2011 23:30:16 +0000

I'm such a happy camper, never having touched this entire Arduino mumbo jumbo, and I'm never going to. For me they're great boards to play with PIC32, the existing Arduino libraries are just a nice little bonus where you can peek around for code pieces. And there's a bootloader.

You could look at this from a positive angle, you're getting a chance to learn for real how the things work, what makes them tick, take a red pill.

Toley

Sun, 24 Jul 2011 02:06:19 +0000

I'm also very disapointed by the "so called" 100% arduino compatibility. Even if we all know that PIC32 core is more powerfull than AVR 8 bits, the actual results are very deceiving. Like Mr_Fixit said 36% of an Arduino is not an Arduino!

The real challenge are the key library like SD card FAT file system, TCPIP and USB host. I really think developpers have underrated the software part of the project and release hardware only for business reasons.

Personnally I'm willing to wait 1 or 2 months to see where the project is going, but at the end if the promises are not delivered be sure that the AVR enthusiam will blast the product and someone will have to carry the blame.

Mr_Fixit

Sun, 24 Jul 2011 04:43:00 +0000

In all reality, my needs are very simple. I need to control a single servo, and need to communicate with several I2C devices, and one serial device. Other than that, I only really require C++ as a programming language, a chunk of processing power, and I am an extremely happy camper. What I do not need is the hassle of debugging someone else's platform. I worked with Gumstix before. Love the platform, but programming it requires being mired in endless depths of cross-compiler and linux issues.

I dont want to deal with issues. Just write code for my project. I am sure many other users here are nodding their heads on that one. That is why we all chose Arduino, and got suckered into Digilent..... :evil:

Working in the medical device community, I understand far too well the importance of making a product that WORKS and is reliable. Providing customer support is absolutely vital as well. You are very correct in that the community provides no end of support, I thank all those wonderful people who help every day. Digilent....I want my money back. You have a lot to learn about quality assurance and customer service.

Andy123

Mon, 25 Jul 2011 14:30:42 +0000

Another vote for false advertising.

I understand that there is no way to reach 100% Arduino compatability, but at least I would expect basic examples and demos to work.

None of the examples except "blinking LED" will compile without errors.

While I really like design as development platform for PIC32, but claims that it is fully compatable with Arduino are false!

WestfW

Mon, 25 Jul 2011 17:26:29 +0000

None of the examples except "blinking LED" will compile without errors.

They do for me. In fact, In a quick random sampling of about a dozen of the included examples, I couldn't find ANY that didn't compile. Perhaps you have an install problem?

Sleepwalker

Tue, 26 Jul 2011 11:59:58 +0000

Kind of glad I found this post, having listened to some of the experiences, I may hold off purchasing the Max32 until this has progressed further. I understand it's a new board and new software, so I'm not criticising that part of it, but I was planning to buy the board with the specific intention of building a particular project (well a few, but one important one) and so similar to MrFixit, I wanted something that was relatively bug-free and to have all (or at least almost all) of the Arduino software's features. It does specially say that it's "Compatible with existing Arduino code examples, reference materials and other resources", so I made my decision based on that. If I was just buying it just to play around with, then it wouldn't be too much of an issue, but I have a specific task that I have to do and I don't want to be spending days/weeks trying to sort out whether a problem is with my code or the compiler being wonky - I'm sure there will be quite enough of my own bugs, I don't need more! ;) Likewise I may need a library that hasn't been sorted yet. Developer support is also a big factor for the same reasons and I have noticed that Brian is very active with the UBW32, though it probably suits my present needs a bit less than a Max32.

I very nearly went with (and still may) Pinguino pinguino.cc (sorry can't put URL, the site complains I look like a spammer!) but Max32 looked a bit nicer for me and I was under the impression (based on the advertising statements) that the software was much further along than it currently is. Has anybody here tried Pinguino? Any comments, good or bad or comparisons?

Again, this is not a criticism of the Chipkit setup or concept, it's more that it has been marketed as something that it currently isn't (and I base this on Chipkits own list, not just the posts in this thread). I really hope that they keep the development going and get it up to spec soon, it looks like it has great potential.

.

Andy123

Tue, 26 Jul 2011 14:37:27 +0000

They do for me. In fact, In a quick random sampling of about a dozen of the included examples, I couldn't find ANY that didn't compile. Perhaps you have an install problem?

Install? not sure what you mean? - Just download and unzip the file... and folder structure is there. Try to compile any Ethernet W5100 samples or DOGM128 library. Both will not complile without major changes

Mr_Fixit

Tue, 26 Jul 2011 16:23:39 +0000

Install? not sure what you mean? - Just download and unzip the file... and folder structure is there. Try to compile any Ethernet W5100 samples or DOGM128 library. Both will not complile without major changes

Absolutely correct. There are no complexities, install processes, or configurations that could be screwed up. Andy is correct. Most the stuff does not work at all. I2C works only partially, but many of the other features are toast.

At this stage, I am seriously wondering if Digilent is actually supporting this product at all. It has been well over a month since any form of a software update from them. :o

WestfW

Wed, 27 Jul 2011 00:42:25 +0000

wondering if Digilent is actually supporting this product at all.

Arduino is, to a large extent "community supported open source software", and I imagine that chipkit is trying to follow on in those footsteps. Neither the chipkit team, nor the arduino team is big enough to provide the level of support you seem to expect. So perhaps you'll need to look for a different product (at a different price.)

Or stop whining and start helping. You don't need to be able to actually FIX bugs in complex library code to help; merely creating a good bug report at the github site. The more information, the better, but pretty much anything is better than "hardly anything works."

I guess a lot of the Arduino community experts are off working on the actual arduino code, where there is no shortage of bugs either, and the impact and glory are better.

(I am in no way officially associated with microchip, digilent, or fubar. I am only peripherally associated with Arduino... These are my personal opinions.)

micronut

Wed, 27 Jul 2011 02:41:44 +0000

I totally agree with WestfW. Half the success of a electronic platform is its community. I frequent Parallax's Propeller forum where forum members write documents for things they feel lacking, have a very diverse code repository, and many have built and sold custom boards to fill many needs. The Uno32 and Max32 could be as successful if the community gets involved.

Sleepwalker

Wed, 27 Jul 2011 10:29:09 +0000

In their defence, it does appear that some changes have been made in the last month, the last 20 July according to the info on GitHub https : // github. com /chipKIT32/chipKIT32-MAX

sorry, I can't post the actual URL as it complains I'm being "Spamy" <sigh> You'll have to figure the above URL yourself. Note sure of the extent of these though. There is another post on the chipkit org board that says to check that GitHub link for updates and changes.

I'm not clear on whether the software development (or conversion from Arduino as the case may be) is actually a commercially developed/sponsored thing by Digilent or whether it's just a bunch of volunteers - I haven't seen anything that specifically says either way, but Digilent certainly promote it as if they are developing it (albeit Open-Source). Actually, they promote it that it has already been developed to the level of Arduino and I think that's what people here are a bit disappointed by. It does seem to be coming along slowly, but I expected a well rounded software product from what was advertised.

.

WestfW

Wed, 27 Jul 2011 22:27:04 +0000

developed to the level of Arduino and I think that's what people here are a bit disappointed

"To the level of Arduino" is pretty ambiguous. I'm pretty happy that the core arduino functions are supported; that's further than most would-be-arduino-competitors have gotten. Supporting (or providing equivalents to) the include arduino libraries is an obvious next level, especially since many of them are not very stable even on the AVR side. Supporting all Nth party libraries that happen to have been written is "obviously impossible."

Sleepwalker

Thu, 28 Jul 2011 11:56:42 +0000

Yes, that's fine, I'm not expecting it to have everything that everybody has made for an Arduino, I'm simply talking about what it states or implies in the Advertising.

As I said before, I'm not knocking the thing or the concept ONLY the fact that it advertises something that it is not (at least not yet). If they said "About 40% compatible and we are working to make it 99% by 2012", then I wouldn't have an issue, but that isn't that case. The other people in this thread aren't "Whining" about nothing, they are complaining because they bought it based on what it claimed and, at least at the moment, it doesn't actually meet what Digilent claim.

Perhaps you're missing that point, it's not the product or even the software they are complaining about, it's that it isn't (yet) what it claims and they aren't being reassured that it will.

There are various areas in the advertising where it heavily implies compatibility and the actual wording in one portion of the advertising is "Compatible with existing Arduino code examples, reference materials and other resources". That to me says if I get 99% of standard Arduino code examples and reference material (presumably tutorials and the like) it will work, and most other resources should also work.

I feel I'm probably a lot less critical than some who brought up points in this thread, but I believe they have a good point - The plain fact is that if you look at the list on the wiki, there are many things not working or with problems -eg. Ethernet, Firmata, Servo, Time, One_wire and various others with problems - Eg. I2C ('Semi OK'), SD ('Unknown'). There are other issues raised in other posts on this forum too.

The comment that it's "further than most would-be-Arduino-competitors have gotten" is subjective, obviously some people will place the worth of one library higher than another, but Pinguino seem to have a lot of libraries working. Having said that, I don't want to start up the debate of who is better, because I prefer the chipkit myself, at least for my current needs.

Again, I'm not knocking the software, the hardware or the concept or the efforts of the programmers, I feel this will be a great platform, if and when they get the main libraries finished off and the bugs ironed out, but they should just be upfront (in the advertising) about where it is at now and be clear abut what they intend and then we can make our decisions based on that before buying and nobody would have grounds for complaint. :)

svofski

Thu, 28 Jul 2011 14:18:11 +0000

When I just received the board, I tried a few examples and they worked like I expected. I can't really get in the pants of a typical person who buys an Arduino, but I don't think that the product deserves the bashing it gets here.

It's a kind of a product for which it's rather hard to draw the line between "almost ready" and "ready". The original Arduino is probably never going to be "ready" either, it's just the community there is much larger and the problems tend to get more attention. Since I'm not the type who calls "Arduino" a "microcontroller", it's hard for me to understand how someone could expect for real 100% compatibility of all libraries on a completely different device. That just too naiive.

Actually, it can never be ready because there will always be someone who needs library X which hasn't yet been ported. If, by some freaky accident, this board becomes so popular that useful original libraries will be written for it, there will be complaints on the original Arduino forums claiming that Arduino is not 100% compatible because that XXX library for ChipKIT is not working.

"About 40% compatible and we are working to make it 99% by 2012" is just not something you write on a product packaging :D

elproducts

Thu, 08 Sep 2011 18:56:46 +0000

So if I understand correctly. The Pinquino can be used as a chipKIT compiler. For example, If I have a working design based on chipKIT and some shields, I can then use the Pinquino environment to convert the sketich into a pure .hex file that can be programmed into a custom PIC32 based board with your bootloader installed that has all my shield connection layed out a one circuit?

Darth Maker

Fri, 09 Sep 2011 03:26:21 +0000

When I just received the board, I tried a few examples and they worked like I expected. I can't really get in the pants of a typical person who buys an Arduino, but I don't think that the product deserves the bashing it gets here.

I will admit to having done a bit of bashing. You mostly stated my main problem with the board. It is advertised as 100% compatible with Arduino, which is not only currently a lie, but also cannot ever be realized (with this hardware), as the Arduino boards are generally 5V devices, and no low-level code will ever work on both.

It even more of a sore spot for me because I know someone who bought one expecting it to be very easy, and ChipKITs don't even have a Getting Started guide. He had never programmed and Arduino, but knew (partially from me) that Arduinos are easy, and thus expected a "100%" compatible device to be equally so.

I actually like the hardware, for what it is. But the software support is lacking. Obviously this can't be fixed in a short amount of time by just a couple people.

Regis, while I enjoy learning about the different available Arduino-like boards, I find it out of taste to advertise that way on this forum.

Perhaps the Pingduino Servo library could be used for ChipKIT?

davec

Fri, 09 Sep 2011 09:19:27 +0000

Perhaps the Pingduino Servo library could be used for ChipKIT?

We actually have two working servo libraries for ChipKit now, Servo (officially part of the build as of MPIDE version 20110822) and SoftPWMServo (user-contributed). So, these things take time but they are improving.

Toley

Wed, 21 Sep 2011 00:34:46 +0000

Hi Mr-Fixit, did you finally been refund ? I don't want to bash the Chipkit32 team but we must admit that the promises are still not delivered. Building and selling hardware is 1 thing but providing the software to support is another one. And an unhappy customer should be refund without any questions.

demineerija

Fri, 23 Sep 2011 00:27:06 +0000

Considering the price and package... I couldn't care less about the arduino functionality.

It is a great PIC32 development board. Using it with MPLAB C32/Pickit3 and could not be any happier with it.

Easy to connect to breadboards and you can basicaly test and implement any application you wish. All those I/O are just accessible so easily.

So why to rant? When you can unleash the true power of PIC32 :)

But if you are into arduino.. I guess it blows :D

les1943

Sat, 01 Oct 2011 16:23:38 +0000

I came to Uno32 from having a lot of ‘fun’ and reasonable results with MPLAB , asm files and 8 bit PICs , Having no C experience I have struggled a bit with implementing my designs , I seem to have spent more cash on programming books than hardware !. I have modified code examples found on the various Arduino sites, so far not used any libraries other than wire , the Uno32 works very well, :D I have a MCP23800 driving a 4x20 LCD and a 24LC256 EEPROM both functioning on I2C and now adding a DS1307 RTC , and then serial transceivers .

I have a couple of Qs

MPIDE compile errors (red) have line numbers , but no option to turn them on in the code frame ?! And what are the little blue lines and boxes that pop up for ? and could someone suggest a site to go for help with ‘defective’ code.

My very first processor project was a RCA CDP1802 so you may guess I am a bit past 50.

lloyddean

Sat, 01 Oct 2011 16:44:00 +0000

I say make the suggestion for a "Programming Questions" forum section.

But in the meanwhile post such thing in the "General Discussion" section. They'll get the hint!