I wonder what does this mean for the prospects of OpenSTM32 and is ST going to make TrueStudio a free product.
Not sure what that means either.
But Atollic had a free version (as in free beer) of their toolchain for the last years already.
Both (TrueStudio and OpenSTM) are Eclipse, and I like neither ...
And I wonder why this thread is now "Marked For Action" ...
I guess it's a reminder for the moderators, they mark a thread for action when they are planning to reply in the future.
I hope that reply action isn't "plausible denial" ... ;-)
The overall "consolidation" process in the IT sector begins to frighten me.
Here's hoping they do away with that accursed "Buy-a-pro-license-now! It-Cures-World-Hunger!" nagware popup. "Modern" IDE's (i.e., bad clones of Microsoft's already-hideous Visual Stupidio) are annoying enough as it is but the nagware makes it unbearable.
The Visual Studio IDE definitely deserves more attention since version 2017. In a recent update they added support for mbed projects.
Agreed, it is a huge fat behemoth, but on a decent machine it is powerful and good looking. And kind of free.
but Windows only - which is a issue for some...
That is an issue for me, too.
Privately I use Linux only, a large fat Windows app is a no-go for me.
And hobbyist are much less bound to Windows, or not ?
BTW, I'm using Crossworks, which managed to have it's own GUI (non-Eclipse and non-VS) on Windows, Linux and Mac.
The Lite banner ads will be removed in next version. It is not free beer, nevertheless an improvement!
"Which IDE/Editor is the best" is a popular coffee break topic which typically ends with "let's agree to disagree" :-)
The choice of going for an Eclipse environment is inline with what the vast majority of silicon vendors are doing.
Additionally Eclipse also has its own plug-in ecosystem (Eclipse marketplace) and allows third-parties to extend the IDE with custom plug-ins. This can become key for a good ecosystem offering.
The TrueStudio Lite version has several limitations compared to the TrueStudio Pro version.
We'll see if ST will offer the Lite version for free, perhaps with some features more than the current version of Atollic, or if it will offer the PRO version for free.
Atollic's free version is (or should I say, was ?) not code size linited like the one's from IAR or Keil.
As noted, the free version clashes somehow with OpenSTM32.
And then, there was the TrueStudio removal from Cube's list of supported toolchains.
We are all in suspense for ST's next actions ...
It all makes perfect sense! LOL
(missing the 'rolls-eyes' emoticon...)
nothing more foolish to think up
eclipse is the most inhibitory development environment
full of s.h.i.t (same as site st.com)
and all product on base eclipse full s/hit
notepad ++ I think much better, why did not they buy it?
eto prosto pizdec blia
I find Eclipse quite useful & usable. There is certainly worse available - including some that you pay for!
Fortunately almost no one understands Russian cursing and swearing ... ;-)
But true, the opinions about Eclipse are divided, and I do not like as well.
And I still don't see the motivation to buy Atollic. Because competitors bought a toolchain supplier, too ?
Or for the other MCUs it supports (ARM and non-ARM) ?
I think ST.M. tries to offer a complete and well integrated solution and OpenSTM32 was not the best answer regarding the IDE part. It still suffers regarding LL support.
But my other question is, will ST.M. use this move to discard the generation of "Makefile" project in STM32CubeMX? I am afraid it will...
I hope it keeps a makefile approach. At the very least this exposes what settings are getting pulled from the project meta data that seem to get buried six ways to sunday on the different tool chains.
What there really needs to be some effort on is something to import or convert projects files between tools. Or some more generic project description which can then be synthesized to a specific target tool
Clive One wrote:What there really needs to be some effort on is something to import or convert projects files between tools. Or some more generic project description which can then be synthesized to a specific target tool
Clive One wrote:
Even though they're both Eclipse-based, neither the SW4STM32 project files not the Atollic project files are compatible with "vanilla" Eclipse!
And there is no way to document the Project settings other than taking screenshots of every configuration dialogue!
Strikes me this could also be an open plug-in stage at the back end of CubeMX to render the project. This could allow Tasking, Rowley, or whomever to generate projects and not be tied to the three most favoured tool of the day as chosen by ST.
Some clarifications that can be disclosed now:
You can expect better integration and more seamless ecosystem in 2018.
And can we also expect better STM32 documentation and more responsiveness on our questions here, or are you throwing all the money on the cuboids and ecosystems?
It may take more time than what you expect, but we manage your inputs and try to answer your questions.
Your USB related comments are an example. Khouloud G is working on them, and updates will come surely.
For sure, we cannot update the documents with each input, but we are trying to schedule regular updates of the documents based on relevant comments.
Hopefully the following is constructive criticism:
1) Tech questions seem to take much longer than other vendors. I use TI parts pretty extensively. When asking questions on their forums (some questions pretty extensively), someone from TI responds pretty quickly. Sometimes same day (if early), but usually by the next morning. I have not had that experience at STM yet.
I have two open requests, one from Dec 11 and one from Dec 12. Both still unanswered (it is Dec 14 AM Eastern time now).
It seems to take overnight just to go from Submitted to Assigned status. I realize that STM support is likely in Europe somewhere. TI is in Texas (I think), and is 1-2 hours behind me.
My closed requests: TECH040106 Submitted Sept 7, answer given Sept 13. So 4 working days. TECH042014 was better, at 1 working day
2) Your web server seems very slow. I feel like I am back to dial up when I visit any STM web page.
I was using TI's forum as well - and the difference is, TI had manned it with technical engineers (besides volunteers).
In regard to specific hardware questions (MCU, peripherals), ST relies much more on volunteers.
I'd wish more of ST's hardware and software guys would be present here.
Is there a way to have this moved to its own discussion?
Although it is relevant to this community in general, It's pretty much off-topic to this particular thread.
As mentioned by Amel N, I'm already working on your USB comments. Our aim is to correct the reported issues on ALL concerned references manuals.
Thank you for your patience.
I think the absolute record is when ST replied to my bug submission after around 5 months fixing it after 6 months (SDMMC interface flag problem). Other people also indicated this as a clear bug shame no one came earlier to look into this...
Mergers and Acquisitions seem to excite executives but rarely deliver the the synergies promised, and frequently end up a disaster (ST-Ericsson) or the death of one of the companies (Buy to Kill).
Given that Eclipse and GNU/GCC are open source the "value" beyond the core team seems marginal. Strikes me you could hire key talent cheaper than buying the whole farm.
I'd expect support of competing chips to be removed, or self exclude rather than share competitive info, driving a large portion of the customer base away, and frankly being of no help to people who have to develop code for a product that might include chips from multiple vendors. Creating a whole bunch of "me too" walled gardens is a win for IAR and Keil (ARM)
Given that Eclipse and GNU/GCC are open source the "value" beyond the core team seems marginal.
The main value of this purchase is IMHO in marketing. Marketing works on promises (among other things). I've heard introduction of Cube/CubeMX attracted major customers, too. The big customers' management wants to hear the promise of quick and cheap development.
I'm sure it is appealing to the "big picture" management types where the details/practicality is less important, but the straw poll of old curmudgeons here it is not remotely fit for purpose in safety critical type applications.
With some suitable staffing I would have thought a GNU/GCC back-end could be integrated into CubeMX
As said in the blog, Atollic TrueStudio(ATS) Pro and Lite will be merged very soon and proposed completely free to all STM32 users.
Nice to hear.
But what does it mean for the other supported MCUs by TrueStudio ?
Are they gonna be removed, resulting in a ST-only tool ?
No reproach, just a question.
AvaTar wrote:But what does it mean for the other supported MCUs by TrueStudio ?Are they gonna be removed, resulting in a ST-only tool ?
I would also like an answer to that question!
Will it be like CodeRed, which started as a general toolchain, supporting multiple vendors' chips, until bought by NXP - who made it NXP-only.
That's still quite OK as they just customize gcc, so there's no impact on availability of the compiler as such.
The real nasty thing was when Microchip bought HiTech and killed their '51 line of their "omniscient compiler's" development, that might have seriously impacted the fate of '51.
Thanks Andi, I wanted to cite this example, but just couldn't remember the name.
I had been working with CodeRed at that time, too. And, ughh, with NXP controllers ...
Yes. The support for other MCUs will be removed. It will be tough for ST to get access to confidential information from competitors
Agreed, self exclusion by other vendors likely to occur. Still the compiler doesn't care if the CM7 is on an ST or competitors die.
Creating a whole bunch of fiefdoms and increasing part prices and overhead.
ST needs it's own IDE like all other major players. And it would be efficient to finally select a few options to focus on, probably:TrueStudio (free unlimited), Keil (for now free for L0, F0), IAR.But currently, not even all of the STM32Cube projects include examples "ready" for TrueStudio, only SW4STM32.They to finally select one and minimize fragmentation to improve quality.
They haven't needed one for the last decade, picking one and driving it off into a siding isn't the way forward.
This is all going be a huge distraction from the job of delivering robust and tested code, which really should be the focus.
Yeah, they killed off TrueStudio support in recent iterations with no explanation, presumably because Atollic was shopping itself to potential buyers.
Rendering a Project file isn't rocket science.
Should have read this first before posting ...
What about a MacOS version of Atollic? I like it a lot about AC6 to be able to work in my native environment...
Avid user of AC6 and have used Atollic in the past. The cost of Atollic had always been salt in an open wound, so this news is welcomed. Getting a good quality tool like this in the hands of ST developers will go a very long way to expanding the brand and consumption of ST products. I just hope they focus on the monetization of their hardware and don't try to make a quick buck with software.
I would imagine the tool teams overhead will now be buried in everyone's part costs. Expect the headcount at Atollic to be trimmed to reflect the focus of a single customer.
I've been using Atollic for the past year and been paying for it for 5 months now. I think it was a fantastic IDE. Sure you can set up Eclipse yourself and use GDB and other tools around it. But the Atollic TrueStudio just has everything under one roof. The free version is already pretty good. The premium is worth the price.
I like the trueStudio too, looks pretty good and many advanced functions. I hope ST will get this ball running in the future.
I started off with Atollic as well and I still like it. The only reason why I changed to Eclipse + GnuARM + GDB is because it supported PrettyPrinting for C++, I could do FreeRTOS debugging without paying for TrueStudio Pro and I could update the toolchain freely.
So if ST now adds Pretty Printing, TrueStudio would be the perfect package.
Hmmm, just looked at the latest release, V9.0, (Welcome to the TrueSTUDIO Family | Atollic ) and the website says:
Atollic TrueSTUDIO is the ideal tool environment for prototyping, proof-of-concept, and new hardware evaluation. It is also widely used by students and hobbyists.
Having dealt with "business speak" for decades, the phrase "It is ..." means it is not what it does not explicitly say.
So it is not the tool for robust, optimized code, ready for commercial applications.
Anyway, waiting for the next release that...
will include Pro features at no charge.
As robust and optimized as gcc can (and LL driver). And debugging works... but where ST-Link fails, it should allow J-Link - I read some topics that it can't, don't know the actual situation...
Anyway, much better than SW4STM32 IDE...
Such deductions from marketing speak (a.k.a "politics speak" a.k.a. "diplomatic speak") had been failry accurate for me the last times ..
I guess that's true, gave SW4STM32 one try more than a year ago, and quickly deinstalled it.
However, I never get use to the dumb default settings of Eclipse (and TrueStudio).
I hate editors that believe they know better what I want to do.
Every action in a variable/register view window spawns a useless console message "Read 4 bytes @ address 0xxxxxxxxx (Data = 0xxxxxxxxx)".
And if closed, it reopens a spam console.
And why an extra icon "Instruction step mode" ?
Just setting focus into the disassembly window and "Step over" would do.
IMHO the focus seems on fancy looks, not robust functionality - no wonder things went as they did ...
AvaTar wrote: So it is not the tool for robust, optimized code, ready for commercial applications.
"It's not the brush, it's the artist..." You know as well as I that the only actual tool for such sits between your ears.
AvaTar wrote: Anyway, waiting for the next release that...will include Pro features at no charge.
If I understand correctly, 9.0 includes all the Pro features. That was the point (no pun intended, sorta ;-) of this release.
"It's not the brush, it's the artist..." You know as well as I that the only actual tool for such sits between your ears.
I don't exactly agree.
One can build the very same source code with different compilers/toolchains, and compare the results according to above mentioned criteria.
That would amount to the "scientific method".
TrueStudio for STM32 supports Pro features for free.
... and there is CubeMX plug-in, ready for the integration... which didn't worked well, but now here is the TrueStudio team, ready for action. Maybe, there won't be any obstacles in the way to performance, but I won't hold my breath.
I guess, there will be modifications to the project wizard and CubeMX will be dismantled in useful Eclipse views.
Dear STM32 users,
If you would like to improve or learn how to work with STM32 MCUs in Atollic TrueSTUDIO IDE, please join our webinar "Debugging and profiling your STM32 device using TrueSTUDIO® for STM32" on Tuesday, May 22.
So the main question that comes to mind that I do not think I saw in this discussion is:
If we incur issues using Atollic which ST has acquired do we post for help in the STM32 forum or Atollic?
I might add at this point the Atollic Forum is very unresponsive.
That's the thing with "user" forums, really rely on a) having users and b) ones that care enough to respond or help.
Having decided they have to own a tool-chain, I'd expect at some point for the forum to move here as a sub-forum.
In the meantime you should probably cross-post and cross-link to both.
Also ponder that support is expensive, free software will not carry support burdens. I suspect Atollic has shed a lot of staff.
for the moment being, the Atollic forum will remain independent from ST Community.
If you are experiencing issues with the Atollic forum (e.g. lack of answers, as I saw mentioned above) please provide here the URL where you didn't receive an answer and we'll follow up internally.
c.c. Brian Kling Thomas Borowczak
Retrieving data ...