Fellow Travelers,
I've stated clearly, my starting point is the C DLL cdecl calling
convention entry points.
You can test the "C" code all day long, but that doesn't prove whether
the DLL is re-entrant.
I have plenty of test cases that demonstrate the *it is not.*
Warning or not, from one professional to another, the community
deserves to know about this in order to make an informed decisions
how/whether to proceed.
Cheers,
Michael Powell
On Mon, Nov 6, 2017 at 3:17 PM, Garrett D'Amore <garrett@xxxxxxxxxx> wrote:
Michael,
(I fully expect you to repost this to the public list, but I advise you not
to. See more below. You should at least read it before you repost it.)
Your actions are incredibly disruptive, not just to me, but to the nanomsg
community at large. I’ve repeatedly told you that I need a test case in *C*
for the complaints you have with multithreaded operation. I’ve also
“proven” that nng is multithreaded by supplying a test program (and you’re
free to modify that if you think it will help you build a reproduction
scenario.) You’ve refused to provide anything in the way of meaningful
characterization, demanding that I debug your program and your test
environment, which as I’ve repeatedly told you is *wildly* different from
what I designed the tests for. (The tests were designed to be run in
isolation, in separate processes.) You insist that the *only* possible
cause for your problems is a hypothetical bug in my software. (There
certainly are bugs in my software — but I’ve yet to uncover any evidence of
the particular defect you claim exists, despite spending several hours of my
Sunday — when I wasn’t planning to code at all — attempting to do so.)
You need to stop engaging in this issue with the community, and take a break
from nng, immediately. I’ve recommended you do this, and I’ve asked that
if you are going to continue to participate that you do so in a constructive
manner. You refuse and stubbornly persist.
The thing is, I have only one other option left. That is to ban you from
all things nanomsg. (I can’t ban you just from “nng” because too much is
shared — including the nanomsg list which you’ve hijacked for your ends.)
Hijacking the mailing list, or the github repo, WILL result in a ban. I
recommend you cool off, and go do something else for a while. Stay away
from nanomsg and nng.
Reposting this message to the nanomsg lists will be treated as continuing to
hijack, and also result in a ban. (Feel free to repost it somewhere *else*,
like a blog, where you don’t hijack people’s email; I don’t care about that.
If you do repost it *anywhere* public, I will be sure to post the entire
history somewhere, so you might want to keep that in mind when airing dirty
laundry.)
This is your last warning.
- Garrett
