Matrix user forums

Hi Jonny

On my pc,hold down the left schift key and dubble click annotation bubble >> no effect i can show ???


Regards Wolfgang

Hi Jonny

Stop it was my mistake,I test it on a one line bubble!!!!!

but
What mean exactly:

""and reducing the size of the annotation down to its smallest size will cause it to be >>>>>>>>>>hidden for that icon <<<<<<<<<<<<<<""

Regards Wolfgang

Hi Wolfgang,

If you drag the annotation to its smallest size then it will disappear completely.

You can also hide the annotation by right clicking the icon, selecting "annotations" and clicking "show".

You can then bring back any hidden annotations by again right clicking the icon, selecting "annotations" and clicking "show".

Hi Benj

One question more:

My bubble go up to "int_ist_flag[32];what is the most simple way to go back to "one line bubble" ?

mark show and unmark it is no change on the bubble-size.

Regards and thank´s wolfgang

Hi Wolfgang,

I'm not sure if this is possible to automatically switch back to a 1 line annotation. Johnny will know for sure but I think you may have to re-size the annotation manually for now.

Hello. Yes, there is no short-cut to show a single line - you have to drag the bottom line of that annotation upwards to the size you want it to be.

Jonny

Hi Jonny

Ok these i now,in my case it is a little bit uncomfortable,but i can live good with these;thankyou

Regards Wolfgang

Hello Matrix Team
While trying to solve one problem, I ran into another. The maximum baud rate listed in the RS232 properties is 38400.
I tried to load my flowchart at this rate and got this message, C:\Program Files\Flowcode\v5\CAL\PIC\PIC_CAL_UART.c(103): UART Baud Rate not supported in Software Mode.
I looked into PIC_CAL_UART.c(103): and found a maximum baud rate of 31250 on line 96. All my other equipment and software runs 38400. I just though you'd like to know. It' no problem so I'll run 19200 for now.
Mac

Hello Matrix Team,
I'm not making much headway past my post of Thu Jan 19, 2012 3:11 am. Now, in addition to the Serious Warnings I got before, I also have (Warning unreferenced functions removed:) and a long list of 72 functions, some which I think I need.
I still cannot communicate with the PIC18F4550 via RS232. As before, I can see default text coming from the chip and I can see commands going to the chip, but with no response. And I still see at lease one observable irregular function, so I know the flowchart as loaded is corrupt.
As a comparison, using V4.5 with the same flowchart, I only get 3 Serious Warnings and a list of 20 unreferenced functions removed. Also, everything still works when using V4.5.
Mac

Hello. One of the changes in v5 is we have tried to make the underlying code more portable via a Chip Adaption Layer (CAL). This means there are more functions available and you should only need a fraction of these. This is the reason for the increase in unreferenced functions warnings from the BoostC compiler.

Apart from this, you mention an irregular function? What do you mean by this, that the code inside it is garbaged? I appreciate that the project content may be sensitive, but the more you can post on this (or PM if you wish) the more we can help. Unfortunately the staff will have gone home for the weekend and Ben is best to help with any CAL issues, but I will do what I can.

With debugging, it is sometimes best to reduce your problem down to as small ammount of unknowns as possible. Can you recieve characters via the RS232 and display them with your HW with a small Flowcode program (without loads of interrupts, etc)?

Let me know how that works and we can go from there.

Jonny

I am having a similar issue to Simon( post # 23)with the free V5 upgrade to my recently purchased V4. After completing the form as shown by Simon, being shown the new key and clicking on "Download Link" I was simply returned to the blank form. I received an email message with the key as well and a line that reads "Download Link: " , but there is no link. Refilling the form yields the same error message shown in Simon's screenshot.

Regards,

Grant

Edit: The "Flowcode Downloads" box here worked with the new key for the upgrade. http://www.matrixmultimedia.com/flowcode.php
I now have V5 installed and activated.

Hello Jonny and/or Benj,
I'm slowly closing in on my problem areas. I have identified several things that need changing. I just have to figure out how.
In the "Component Macro" window for the "ReceiveRS232Char", under the parameters section it requires a valid expression. I don't know what expression it's looking for and suspect this might be one of my problems. In Flowcode V4, it accepted "1", but V5 says "1" is not valid.
Many Thanks, Mac

Hello. I dont know why it wont let you do this. Can you post the error message or a screenshot? This lets me add '1' fine for this. Jonny

Jonny, I made a mistake. Apparently I clicked on the drop down menu, didn't know what to chose, left the field blank and clicked on OK. This is when the message appeared, "Parameter 1 is invalid: You must provide a valid expression". I misread it, thinking it meant (expression = 1 is invalid).
Sorry bout dat.
Mac

Ah, thats a relief! You had me worried for a moment there!

Mac wrote:Hello Matrix Team,
I'm not making much headway past my post of Thu Jan 19, 2012 3:11 am. Now, in addition to the Serious Warnings I got before, I also have (Warning unreferenced functions removed:) and a long list of 72 functions, some which I think I need.
I still cannot communicate with the PIC18F4550 via RS232. As before, I can see default text coming from the chip and I can see commands going to the chip, but with no response. And I still see at lease one observable irregular function, so I know the flowchart as loaded is corrupt.
As a comparison, using V4.5 with the same flowchart, I only get 3 Serious Warnings and a list of 20 unreferenced functions removed. Also, everything still works when using V4.5.
Mac

You should definitely recode to remove any 'serious warnings' in both V4 and V5 of Flowcode. Even when the program appears to work fine, these types of warnings point to a fundamental problem with the design of your program which could cause it to fail in unpredictable ways at some future time.

As for the 'unreferenced functions', these are simply information messages and can generally be ignored.

Steve wrote:You should definitely recode to remove any 'serious warnings' in both V4 and V5 of Flowcode. Even when the program appears to work fine, these types of warnings point to a fundamental problem with the design of your program which could cause it to fail in unpredictable ways at some future time.

As for the 'unreferenced functions', these are simply information messages and can generally be ignored.

From those i have tested so far, the most usual cause for warning messages from compiler is the use of more than 1 interrupts in the program code. Indeed, the use of a "global" variable resolves the problem in this case. To be clear, in my project i make use of both RC5 compo interrupt - applied on port RB0 by default - and the PortB interrupt. But this causes other issues, like the slow down in values update on LCD display. In actual hardware, i have checked that NUMBERS are "blanked" before their renewal! I am bothered on this. I tried the other way, to substitute the PortB interrupt with a continuous call - placed inside the main loop of program - of a macro e.g. "MANUAL" to check if the state of any switch connected in PortB has been changed. This also resolves the "call stack overflow" issue, while at the same time, does not affects the response speed of LCD display (i.e. numerical values are updated continuously without "blanking"). Though, still i receive 'possible sw stack overflow" for all LCD (FCD) compo macros. I can't do something for that, i use a P16F887 which of the stack is just 8 level deep.
By the way, using RC5 compo there is a device macro (FCD) for which i receive allways from BoostC compiler the "Possible SW stack overflow" message for the "FCD_RC50_RC5_Initialise_States". I can't solve it by no way.
On the other hand, i have many times thinked that there is no case to simultaneously operate both the remote control and the manual functions (EDIT: Trying to activate both Interrupts at the same time).
Although i purposively tried any abuse on it and never had a program corruption, i still believe that your suggestions are absolutelly right.

Hello Steve,
1. I'm trying to recode to remove the 'serious warnings' as you suggested. I think I have identified them. Two have to do with using multiple component icons of SendRS232String. It was easy to rewrite the smaller of the two macros to use the component just once. I could see how to rewrite the second "large" macro and use the component only once. This second macro uses the SendRS232String 32 times, all ending at a single point. All I would have to do is make some name changes in the string manipulation boxes and use one RS232 component at the single point. I changed one or two at a time and then tested. All went well up to around 8, then it quit working. Thats a long way from 32. But I'm going to keep trying. I can not see a way that this component could be called twice at the same time because in each case the accompanying code is called sequentially from a single source. I have attached an example flowchart to show what I'm trying to do. If you see a better way, I would like to know.
2. One of my problems was using a Custom USART interrupt in the main loop just before the ReceiveRS232Char. I guess it was not needed, (in V4.5) as it seems to work just fine without it.
3. I'm working on these problems, using V4.5. I still have not figured out what the problem is, in not being able to communicate with the HP488 via RS232 using V5. I've test V5 using a much simpler flowchart and it works fine.
4. I mentioned in an earlier post of at lease one observable irregular function. This is a regular function, but not called during the default mode. I feel sure once the RS232 communication problem is resolved, this can easily be fixed.
5. As for the 'unreferenced functions', I will ignore them. Thats a good thing.
Thank You Very Much. Mac

Hi Mac,

Looking at your file I noticed that there is no configuration settings for the device in the project options. Could this be the cause of the v5 program not working?

Also in your string manipulations you currently have code like this.
Should this instead be.
If so you can simply pop all of the string assignments once at the top of your program and then simply send out the appropriate string variable using the decision icons. You could also maybe replace the keep_track decisions with a single switch icon to tidy up the program a bit.

As an additional point, you can use the 'initialise' field of a variable to set this up how you would like, so for Blue set 'initial value' to "Blue".

You could even go further and set these as constants. To do this, remove the references in the file for Blue = "Blue", set them as the initial value and then select 'Convert to Constant' from the variables drop down menu. I hope this helps,

Jonny