Skip to content

Conversation

@dagger82
Copy link
Owner

update on 2020/3/22

Imroy and others added 30 commits December 5, 2016 14:55
To allow derived classes to have access to the protected variables and methods.
…to differentiate it from the other write()'s

The others first write their 16-bit length before the data.
_payload_callback should be initialized or we have crash (in case no _GLIBCXX_FUNCTIONAL available)
Fix not initialized field
…intable

Cut down on duplicated code for handling both JsonObject and JsonArray.
…hile we wait for an acknowledgment (in persistent session and QoS > 0)
Fix: Library remains blocked and PUBCOMP packet
Make library name uniqe by adding username in parentheses
Users on other platforms should include <functional> before <PubSubClient.h> if they want the added C++ features with callbacks.
Should I remove these lines and force all users to include the appropriate headers?
This gives a little extra type checking, at least with assignment.
Just continue the loop and see if available() or another read() fixes things?
Use C++ new/delete for all other dynamic allocation.
Should I use std::vector instead?
From a look on my system, it appears that std::vector is available on all ARM-based processors and ESP8266/ESP32. But not AVR. Is it safe to exclude AVR for this library?
Imroy added 27 commits May 16, 2017 03:21
Client object is first, callback is set with a method.
Some of the uses in connect_spec require constructing our own custom Connect object.
…brary has no such define

And sizes are 32-bit now.
TODO: Will have to construct the 'publish' packets with the correct size when this max size is greater than 127 bytes.
A better way to encapsulate the state variables.
Still untested!
…g length byte(s)

This allows us to parse the small, simple packets e.g connack, ping, etc. They're only four bytes, smaller than the five byte minimum in _read_header().
_read_length() is reenterable, able to shift a single available byte and continue later.
…e message

It returns nullptr instead of false.
…s for a response

The second one returns a message pointer of the message that came back, thanks to _wait_for().
CONNACK doesn't necessarily mean the connection was successful.
@dagger82 dagger82 merged commit 62663c1 into dagger82:master Mar 22, 2020
@dagger82
Copy link
Owner Author

update on 2020/3/22

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

6 participants