Missing EEPROM.length() method

[ricardojlrufino]

Hardware

Hardware: ESP-12f
Core Version: 2.2.0

Description

The EEPROM.length () method is not available for the ESP
Ref: https://www.arduino.cc/en/Tutorial/EEPROMClear

Want to back this issue? Post a bounty on it! We accept bounties via Bountysource.

[MacroYau]

The current EEPROM Library implementation for ESP8266 requires you to initialize the size in the EEPROM.begin() function as you may notice in the example sketch.

@igrr I am afraid that the change is not as straightforward as adding the EEPROM.length() function. It seems that the function was introduced as a part of the EEPROM Library V2.0, which is largely different from our current design.

[ricardojlrufino]

Why the function begin (4096), creates a buffer of the same size?

In my view the EEPROM lib would be used to save RAM too, does not make sense I create a buffer of the same size.

I think what could be done is to have a smaller buffer when full, write the data in the EEPROM.

In my implementation I need to store a list of rfid tags , then I plan to allocate the maximum size: 4096 in this case I would finish with two buffers = 2 * 4096

It would be viable to improve this design?

[pgScorpio]

I think the EEPROMClass::begin should also have a version like EEPROMClass::begin(startaddress,size)
which would create a buffer of 'size' bytes and would apply 'startaddress' as an offset between addresses and the buffer index.

After that you can only read/write addresses between startaddress and (startaddress+size-1)
write should return a bool, false if the write is not within the buffer range.
read should return 0 if the read is not within the buffer range.

This way you can have a very large EEPROM, but never allocate more memory than needed for the current read/write actions.

PS: There should be functions to retrieve the size of the EEPROM and the current min- and max- read/write addresses (-1 if not between begin and end() ?)

PS 2: And EEPROMClass::end() should return the result of commit();

[DouglasBair]

Hello,
I was about to enter the missing function/method of EEPROM.length when I noticed theis issue.
It was in the Arduino.cc's EEPROM library up to and including 1.6.9, then in Arduino.org's IDE, it
is missing. No regression testing, no coordination. I've been testing to see what all works( or doesn't)
in the different versions of the IDE. I hope someone at Arduino.ORG will fix and TEST their software for
backward compatibility. July 17,2016

[igrr]

Excuse me, what does this repository have to do with arduino.org?

[DouglasBair]

Maybe I am wrong, but I thought you had a problem with missing a function of
'EEPROM.length'.
I use it in the Arduino.cc's IDE and it went missing in the version 1.7.0.
Are you using the IDE provided
by Arduino? Ignore my comments if they don't help you.
Doug

---

From: Ivan Grokhotkov [mailto:notifications@github.com]
Sent: Sunday, July 17, 2016 11:07 AM
To: esp8266/Arduino
Cc: DouglasBair; Comment
Subject: Re: [esp8266/Arduino] Missing EEPROM.length() method (#2066)

Excuse me, what does this repository have to do with arduino.org?

You are receiving this because you commented.
Reply to this email directly, view
#2066 (comment) it
on GitHub, or mute
<https://github.com/notifications/unsubscribe-auth/ARmIyPBWqSqOD__QilxTMykB8
KdS5OZEks5qWmElgaJpZM4IpSkm> the thread.
<https://github.com/notifications/beacon/ARmIyEhSjJj9hzrbVKdQP2OOBrK_Kg73ks5
qWmElgaJpZM4IpSkm.gif>

---

No virus found in this message.
Checked by AVG - www.avg.com
Version: 2016.0.7640 / Virus Database: 4627/12632 - Release Date: 07/17/16

[igrr]

In Arduino, EEPROM library is normally part of the core and not a standalone library. So there is one version for AVR core, one for ARM, one for ESP8266, etc. All these versions of EEPROM library are derived from the original AVR version as seen in Arduino.cc project. However as you may have noticed that original library (Arduino.cc one) has recently been updated with some new methods. The libraries which were derived from it may not have been updated by their maintainers yet to match the new API of Arduino.cc's library. This is the case here (we haven't updated our library to use the new API) and also the case with Arduino.org's AVR library (they haven't updated their library either).
So this doesn't really have anything to do with regression testing and backwards compatibility. Arduino.cc's libraries are evolving and everyone has to catch up :)