embeddedgurus.com: "An open letter to the developers of the MPLAB IDE"

[dannyf]

http://embeddedgurus.com/stack-overflow/2011/09/an-open-letter-to-the-developers-of-the-mplab-ide/

Nice reading,

I am so glade that emblocks is there.

[ElektroQuark]

I don't know if it has any meaning now. MPLAB is surpassed by MPLAB X. But we can talk about MPLAB X... 

[linux-works]

I avoided getting into controller work for a long time.  it was a field only for those who were very deep in it and hard to break in if you were just starting out.

arduino 100% changed all that.  made it accessible and even fun.  the tools were ALL FREE and ALL FUNCTIONAL.  no 'pay extra to remove no-ops' BS.

I know that microchip has some better hardware at the chip level, in some ways, but their braindead tool chains and the fact that they still don't 'get' the free software movement has me still being an arduino guy and next to zero pic experience.

PIC: atmel is totally winning over the new generation of kids and if you don't get your act together, you'll be the 'os/2' of chips (lol) in a few short years.

[dannyf]

the free software movement

I am too poor to use the "free" software.

atmel is totally winning over the new generation of kids

Dream on,

[linux-works]

I'm not dreaming; I see no new projects that are pic based compared to the numerous ones that are arduino based.

I would not even consider using a pic in an opensource project simply due to the fact that the tools are not free and not nearly as functional as they could/should be.

the ones using pic are the ones told to use it at school or industry.  I know of no one that starts out picking the pic anymore.  its used by pros (who have the expensive tools) but rarely by new guys starting out.

if you have info to show otherwise, bring it.  else, it seems pretty obvious if you just take a look at a place like hack-a-day (etc) and see how many pic submissions are there vs atmel.  pic is rarely even mentioned anymore and that's a fact.

[Dago]

I'm not dreaming; I see no new projects that are pic based compared to the numerous ones that are arduino based.

I would not even consider using a pic in an opensource project simply due to the fact that the tools are not free and not nearly as functional as they could/should be.

the ones using pic are the ones told to use it at school or industry.  I know of no one that starts out picking the pic anymore.  its used by pros (who have the expensive tools) but rarely by new guys starting out.

if you have info to show otherwise, bring it.  else, it seems pretty obvious if you just take a look at a place like hack-a-day (etc) and see how many pic submissions are there vs atmel.  pic is rarely even mentioned anymore and that's a fact.

I have to agree. I have seen very few PIC projects in the last few years.

I have used the xmegas in work projects and I have to say they are extremely nice and consistent and intuitive to use. Way better than atmegas.

[retrolefty]

I agree that the arduino project has brought microcontroller projects to a whole new classes of people that had not and would not otherwise have entered into the field. Lots of gray beards poo pood the whole concept early on, not unlike the mainframe computer types looked upon the first microprocessors in the 70s.

 And it's not that the original arduino developers were so brilliant or anything, they just had a good idea aimed at a unique audience (artist and non-techies) that needed simple tools and low cost way to enter the field.

 The arduino folks just took a bunch of existing open source projects already developed and avalible and integrated them together with a very very simple controller board based on a Atmel AVR 8 bit controller chip. The use of avr-gcc compiler/linker, Processing IDE, AVRDUDE, Wiring, etc, for the idea of using abstraction for things like using pin number instead of the port/pin addressing required at the chip level. It's actually really amazing how little new code the original arduino folks had to create, they just stitched together lots of open source code together and built a simple controller board which they then released as probably the very first highly successful open source hardware project.

 Soon even experienced software types found that the platform can be useful, as one is actually developing code (sketches) in C/C++ with all it's capabilities and history. So soon contributed libraries and projects quality and volumn was very impressive and has reached a critical mass that would be certainly hard for any other platform to surpase.

 Certainly there are more powerful controller chips available with tons more power at the same or even less price, but if one measure of power is the size of the existing user base it will be interesting to see how long the Arduino project keeps the simple 8 bit controller actually relevant long pass it's otherwise must use by expiration date.

 Indeed even the Arduino company has now released Arm based boards (Due) but it's not clear if it is being accepted and adopted in the huge numbers that the more simple 8 bit boards are. And being open source hardware one can now find simple Asian cloned E-bay sold AVR mega328P mini-boards (named Pro-mini) for like $3 each.

 And a whole new generation of small specialty vendors are now actively selling components, modules, sensors, etc to this population. Firms like SparkFun, Adafruit, etc started out small with this market and grew to multi-million size very quickly.

 For those of you who are as old as me it's very similar to the early growth of the microprocessor field in the mid 70s, where it was not the existing major computer and electronics companies the entered the market till later, it was very much a hobbyist driven market at first.

 To be clear it's not a fan boy type of product or platform, as it is not really aimed or useful for the plug-and-play type users, one must learn some basic electronics and programming skills to actually be able
to convert an idea into a real project. But it does allow the quickest learning curve to just getting started of any other platform I had come across. It's fun and useful and cheap, what more does one need. 

[linux-works]

I choose arduinos for my open hw/sw audio projects because of wealth of examples (that jumpstarted me) and also so much online sharing of info.  it seemed like the ideal thing to get people to WANT to change the code and try new things out.

its not always the best chip for the task, but the fact that the devel tools work on all 3 platforms and are fully free (and pretty mature, too) seals to deal for me.

industry wise, its often about chip cost, chip size.  but for home projects, the size is not as much of a problem and cost is also not a problem since we don't build in thousands of units.  what does matter is the toolchain and community support and in that regards, nothing can touch the arduino, the raspi or the BBB.  those are the 3 top boards for home DIYers and nothing else comes close (for now, at least).

when I read gerry sweeney's write-up on the non-optimizing compilers on MPLAB adding no-ops unless you PAY to have them removed, that did it for me; I have nearly no desire to do anything PIC based if I can help it.  such things are hostile to users and I don't like helping any company that is this hostile to their userbase.

PIC is the microsoft and atmel/arduino is the linux, very very loosely speaking (if you catch my drift).  I avoid doing any microsoft programming for many of the same reasons I avoid pic.

[BloodyCactus]

Nice reading,

in 2011 when it was written... I'll stick to mplabx.

[janoc]

the free software movement

I am too poor to use the "free" software.

Yeah, it is better to spend gazillion dollars on a proprietary tool, surely is better than anything free, right? Except it uses GCC for compilation and GDB for debugging under the hood. Like 99% of ARM tools, the official Atmel's, even the new Microchip's ones, to name just the most common stuff.

I am not sure why are you taking ignorant jabs at free software here when, in most likelihood, you are dependent on it too. It isn't about free as in beer.

atmel is totally winning over the new generation of kids

Dream on,

I haven't heard about any projects using PICs that were built by kids recently, but a boatload of Arduinos, Raspberry PIs, Beaglebones and similar. Microchip missed the boat here big time, IMO.

[linux-works]

microchip REALLY missed the boat, that is so true!  as I read the story, the arduino guys approached both MCP and ATMEL and MCP turned them away.  they will regret that for the next 5 years, very likely.  their arrogance is also part of the problem and why people hate having to learn all the gory details of each fork of the mcp architecture.

arduino, while dumbed-down a bit too much (imho) did take the right approach toward an abstraction layer and it removes the need to care about which chip is below that layer, for the most part.  that was such a helpful and visionary design aspect.

...just wish the pin spacing on those 'shields' were reviewed by folks who understand that ".1 spacing is king" in the diy world.  sigh.

(pet peeve: people who design for DIY and yet avoid .1 spacing should be sent to a special form of hell when they die)

[dannyf]

I'm not dreaming; I see no new projects that are pic based compared to the numerous ones that are arduino based.

Projecting what you say to what the market says is probably not wise.

Unless you are the market.

[retrolefty]

...just wish the pin spacing on those 'shields' were reviewed by folks who understand that ".1 spacing is king" in t he diy world.  sigh.

 Did you ever read the back story about the strange spacing between just two of the four 'shield connectors'?

 It was simply an error that they didn't catch when their first boards were returned for their approval, they meant to have .1" spacing. But it was too late (and most likely too expensive) to correct by the time they did discover it. After that they were committed as there were already 3rd party suppliers shipping shield products that used the irregular spacing so it was turned into a "feature" as it makes installing shield boards backwards onto arduino controller boards impossible. 

[linux-works]

I'm not projecting.  why do you think so?

a simple count (if you want to do the work) on any open source hardware forum will show tiny numbers of PIC projects compared to atmel/arduino.

I'm not going to argue with you about this.  readers know what's what no matter how you want to try to spin things.

the fact is, MCP is not nearly as DIY friendly and the # of NEW projects using their gear has fallen to near noise level since the advent and popularization of the arduino architecture.

[dannyf]

a simple count (if you want to do the work) on any open source hardware forum will show tiny numbers of PIC projects compared to atmel/arduino.

Microchip (or Renesas or Freescale or ...) has zero market share on the Arduino forums, So they must be doomed.

I would suggest maybe a more robust way of reasoning?

[linux-works]

go to hack-a-day, then.  what percent of pic projects do you see vs arduino?

that forum is not locked to any one vendor, so its a fair place to start.  its a popular site for controller projects.

that's my 'easy' data point.  if you have one of your own that supports that pic is gaining popularity, I'm all ears.

[mazurov]

go to hack-a-day, then.  what percent of pic projects do you see vs arduino?

Go to any rec.center. Over a period of a year, how many kid's paintings do you see on it's walls comparing to works of professional painters?

[mariush]

microchip REALLY missed the boat, that is so true!  as I read the story, the arduino guys approached both MCP and ATMEL and MCP turned them away.  they will regret that for the next 5 years, very likely.  their arrogance is also part of the problem and why people hate having to learn all the gory details of each fork of the mcp architecture.

I actually think it was a smart move, in a weird way.

Look at Atmel now... yes, Arduino is very popular, but it mainly uses 2-3 chips from Atmel's line and now Atmel has to continue to make those chips and support them otherwise their reputation will suffer.  But how about newer, better parts? Do you think any reasonably large percent of the persons buying Arduino even bother to check what other chips Atmel does, do you think they even care to download the datasheet and work around the wrappers supplied to them by the Arduino guys? Hell, I'd bet the majority of the people buying Arduino don't even know that the microcontroller is made by Atmel.

It's almost like a chicken and egg problem - if the Arduino guys don't make a board with a newer chip and support it in the software with wrappers, for the majority of those amateur developers that chip doesn't exist...
 

[dannyf]

go to hack-a-day

I said this earlier so let me be blunt: if you want to know how a product does in its market, go look at how it does in its market, not how it does in *****some segments***** of its market.

Hope it helps.

[ecat]

I like the Arduino, it's ideal for quick prototypes where some driver code is already available and it certainly hit a nice target market. With that said...

a simple count (if you want to do the work) on any open source hardware forum will show tiny numbers of PIC projects compared to atmel/arduino.

Microchip (or Renesas or Freescale or ...) has zero market share on the Arduino forums, So they must be doomed.

I would suggest maybe a more robust way of reasoning?

Indeed.

How many types and variations of mcus does ATMEL make? Would you suggest that ATMEL cease production of all but those used in the Arduino projects? After all, Arduino is so popular making all those other chips is a waste of time?

How many Arduinos are sold each year? Beside me I have a carbon monoxide alarm containing a PIC16, tens of thousands sold perhaps? My two Logitech remote controls  though different models both use PICS, hundreds of thousands sold? I have an AllForOne remote around here somewhere iirc PIC based too, again hundreds of thousands sold?

These are just the devices that come immediately to mind, Microchip may have missed one boat but there are many, many more in the harbour.

[dannyf]

now Atmel has to continue to make those chips

I don't think Atmel cares the least bit about it. I did this calculation awhile back: sales of chips related to Arduino wouldn't even amount to a rounding error for Atmel.

[Rufus]

microchip REALLY missed the boat, that is so true!  as I read the story, the arduino guys approached both MCP and ATMEL and MCP turned them away.  they will regret that for the next 5 years, very likely.

Microchip shipped their 13 billionth PIC last year, and continue to ship more than 1 billion PICs a year.

Here http://www.adafruit.com/blog/2011/05/15/how-many-arduinos-are-in-the-wild-about-300000/ they think there are 300k Arduinos shipped.

Microchip ship more PICs every 3 hours than have been used in Arduinos total. Really missed the boat and regret? The Arduino chip market is completely trivial and the generally clueless makers using Ardunos are not the ones who are going to get Microchip design wins in volume products.   

[linux-works]

the arduino architecture extends much more than the 328 style chips.

in fact, its now more than atmel, as has been mentioned (arm and even MCP versions of 'arduino' that I've seen).

there was a short project called galago (I knew the 2 guys working on it, back a year or 2 ago) and that had the promise of using arm chips but in an arduino-like environment with a full hardware on-chip debugger, as well.  they didn't get too far (business wise) but they did have a cool idea and it was all because of the arduino 'movement'.

[dannyf]

The Arduino chip market is completely trivial

Unfortunately, facts like that are very difficult to comprehend by the all-too-important Arduino folks.

[linux-works]

atmel understands something mcp missed: planting seeds.

apple knows this, too (lol: apple seeds, ha!)

they (apple) give huge discounts to schools so that kids learn on apple gear.  MS is also trying that.

once you get someone trained on some tech, they tend to use it over stuff they don't know.

analogy: why is cisco the undisputed king of networking gear?  my answer: because their CLI is everywhere, they get companies (and schools) to value the certifications from cisco and once you know their gear, you tend to stay with it and choose it for new projects.

this is a very smart business approach.  and I would also likely choose an atmel chip over a pic chip in a commercial setting since I know one over the other much better.

the commercial side of sales is not what we are talking about HERE, though; the controllers forum on THIS site is more about diy and not at all geared towards those that do this for a living.  folks like that are here, but this forum is not about them, its about the learner and home diy'er.  and for them, PIC nearly does not even exist on their radar.