Casio fx-9860G SD- How to read files in the sd card?

[SimonLothar]

What will be realistic speed of copying data from sd to storage memory please - much worse than the current speed of saving main memory to storage memory using your insight add-in?

Pure writing or reading of a file to or from the sparkfun takes 20-25 s for a 200000 byte file. Now I copied the file from the sparkfun to the SMEM. Bfile_CreateEntry takes about 30 s on a freshly optimized storage mem (nearly double the time, if the storage mem is not freshly optimized). Writing the file (Bfile_WriteFile) using 514 byte blocks takes about two minutes (25 s max. for reading from the sparkfun). Most of the time is consumed by the Bfile-syscalls.

[TeamFX Profile]

Oh, you have uploaded a new Insight.g3a version...
But I had to remove that funny "function disabled" thing...

[SimonLothar]

Oh, you have uploaded a new Insight.g3a version...
But I had to remove that funny "function disabled" thing...

I hope I have all eliminated the flash erase commands, which I experimentally inserted in some of those functions. The Prizm is such a cissy...

[SimonLothar]

Oh, you have uploaded a new Insight.g3a version...
But I had to remove that funny "function disabled" thing...

I hope I have all eliminated the flash erase commands, which I experimentally inserted in some of those functions. The Prizm is such a cissy...

Just kidding...

[AmazoNKA Profile]

Thanks for the updated insight- my soldering help was postponed till the coming Sunday so hopefully i will be able to tell soon what firmware version my openlog is and if it is any better. Thanks again

[SimonLothar]

The performance of the syscall Bfile_CreateEntry as well as Bile_WriteFile is very poor.
If I copy the file in the windows file explorer, the transfer is completed in a few seconds.
I think there are some prerequisites to match, which I do not know yet.

[AmazoNKA Profile]

Is there any improvement with bfile functions at all if you were to manually run memory manager optimisation before going into your add-in please?

[SimonLothar]

Is there any improvement with bfile functions at all if you were to manually run memory manager optimisation before going into your add-in please?

Bfile_CreateEntry (200000 byte) takes about 30 s on a freshly optimized storage mem (nearly double the time, if the storage mem is not freshly optimized). The very next Bfile_CreateEntry takes 50 s again.
I fear there is no other way, than tracing down the way of the LINK-function or any other system-function which does a fast SMEM-create/write.
Perhaps there is some flag, which has to be set or preliminary function, which has to be run.
That work will consume some time.

[SimonLothar]

I think I found the reason of the bad SMEM performance.
There is a difference between fx-9860- and fx-CG-calculators.
With fx-9860G/GII-calculators Bfile_CreateEntry must be called with the final filesize (This is what I remembered and I thought it has to be applied to the fx-CG, too).
With fx-CG-calculators Bfile_CreateEntry must not be called with the final filesize, but with size==0.
Above that I now use a secondary SMEM-write-buffer as large as possible. This reduces the count of individual SMEM-writes.

Now the SMEM-thing is finished within the time, which is needed to transfer the data from the sparkfun, t. i. 20-25 s instead of three minutes in case of a 200000 byte file.

[AmazoNKA Profile]

That's what i like about this calculator - almost every time we need a work around for something you discover more things helpful with something else and the whole satisfaction of working with it feels even more rewarding... New ideas emerge

Thank you so much for all your work - I hope you will release another insight revision for faster tests soon.

Thanks again