James Brock
James Brock
> 1. The checksum we do not verify ( We have the option to verify the checksum though. We leave it to the library user to decide if they want...
Hi @rossb34 , I'm open to this, but I don't use Cmake. Do you want to make a PR? We would also want a .nix shell file which provides a...
"zero-copy serialization" https://gitlab.haskell.org/ghc/ghc/-/wikis/linear-types
Yeah, good point. That's a rare case but it could occur. That should be fixed. Do you want to make a PR?
Thanks for the PR! I can't look at this right now; will look in a few weeks.
(These lines are from the example programs `writer01.cpp` and `reader01.cpp`.) On reflection, this is fine: ```c++ for (; reader.is_complete(); reader = reader.next_message_reader()) { ``` But maybe this ```c++ hffix::message_writer new_order(logon.message_end(),...
I suspect that `reader = reader.next_message_reader` is causing #27 . What we really need is a whole `message_reader_buffer` abstraction with iterators which dereference to `message_reader`s. That would be a massive...
Also the docs for this function have a typo, missing a period.
Hi, thanks for opening this issue. > 1. On Windows, there is no 'ssize_t' type. You're right, `ssize_t` is not standard and we shouldn't use it. I'll fix that. >...
I tagged v1.1.1 for the `ssize_t` bugfix. https://github.com/jamesdbrock/hffix/releases/tag/v1.1.1