|
Hey Sebastian,
I think that’s the only big thing that needs done.
Cheers,
Aaron
On Apr 23, 2015, at 11:02 AM, Sebastian Kostuch <> wrote:
Hi Aaron
it seems like current version of WebSocket client is performing quite well (as of beta version) and other mentioned things are done. Is there anything else what blocks us from making final release?
Regards
Sebastian
On 14.04.2015 16:44, Aaron Brown wrote:
Hey Sebastian,
I think the bulk of the integration work is now done. I’ve added some tickets to cover other things that need done before the final release:
https://github.com/ndt-project/ndt/issues/178
https://github.com/ndt-project/ndt/issues/179
https://github.com/ndt-project/ndt/issues/180
I think the big thing that needs done is just testing.
Cheers,
Aaron
On Apr 14, 2015, at 4:44 AM, Sebastian Kostuch <> wrote:
Hi Aaron,
would it be possible to create tickets with information what should be done before NDT final release? Especially info about how the integration with websocket client should look like?
Regards
Sebastian
On 13.04.2015 21:51, Aaron Brown wrote:
Hey Folks,
I merged in the websockets branch, and did some other (crappy) integration work with the d3-based websocket client. You can see an example at http://desk179.internet2.edu:7123/
I’ve got RPMs and the source tarball of a tentative -rc4 at http://ndb1.internet2.edu/~aaron/ndt-3.7.0-5.el6.x86_64/
There’s still stuff that needs done before a final release:
- Place copious warnings about the flash client
- Use the flash client’s definition of “good environment”
- Improve the integration a bit with the websocket client
- Place copious warnings in the websocket client
But this is probably a reasonable -rc4 to get folks testing out the websocket client. If folks would be willing to do some testing, I’ll announce -rc4 onto ndt-users.
Cheers,
Aaron
On Apr 13, 2015, at 4:32 AM, Sebastian Kostuch <> wrote:
Hi,
that sounds reasonable. When can we expect then websocket changes being merged into master and release of rc4 version?
Regards
Sebastian
On 11.04.2015 00:55, Jordan McCarthy wrote:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512
I entirely agree with Peter: getting web-sockets support into the next
official release appears to be very much a win-win. Both because the
code sounds like it's completely ready to go, and because all other
GUI-friendly tests are proving to be pretty hideously unreliable.
I had previously thought that we could make do by just releasing a
properly-qualified and environment-limited flash client, at least for
the time being. But given the increased reports of bizarre numbers
coming out of the flash client (coupled with the fact that we know it
doesn't work properly on Mac OS X or Linux), it seems like we would be
remiss in not at least providing the option of running a
WebSockets-based test.
Jordan
Jordan McCarthy
Open Technology Institute @ New America
Public Key: 0xC08D8042 | 4A61 3D39 4125 127D 65EA DDC2 BFBD A2E9 C08D 80
42
On 04/10/2015 03:03 PM, Peter Boothe wrote:
MLab is very excited about websockets and the _javascript_-only
network tests that they enable, so if we can get websocket support
merged into mainline NDT sooner rather than later that would be
great. We have found that, for an increasing number of users,
Flash and Applets are no longer an acceptable solution - applets
now cause horrendous security warnings and flash is disabled on an
increasingly large number of platforms. This leaves us a bit in
the lurch in terms of how to get measurements from people running a
web-browser until we get websocket support on the NDT server.
Also on the positive side, the websockets branch is fully
unit-tested and end-to-end tested ('make check' should run the
tests on your local machine), and it looks like it can be merged
with no conflicts.
-Peter
On Fri, Apr 10, 2015 at 6:36 AM, Sebastian Kostuch
< <mailto:>> wrote:
Hi all, it has been long time since last NDT final release and
also preparations for this one are taking few months so far. Isn't
then better approach to release new NDT 3.7.0 final version ASAP
with minimal changes/fixes required to do so and after that within
few following months make new release (3.7.1?) which would contain
new features (websockets + there are also changes on MultiplePorts
branch which would be good to integrate with it as well) ? Merging
websockets now will probably make release process last longer so
what do you think about such approach?
Kind regards Sebastian
On 09.04.2015 22:54, Nathan Kinkade wrote:
Aaron,
I really like this plan. I think that by the time RC4 is out
we'll have a pretty good _javascript_ Websockets client assembled.
It's already functional, but we just need to make it a bit more
robust when it comes to error handling, and apply some general
stylistic changes. And as you say, it is much easier to do minor
version updates than some new release for a different feature.
Thanks,
Nathan
On Thu, Apr 9, 2015 at 3:04 PM, Aaron Brown <
<mailto:>> wrote:
Hey Jordan,
Given the performance … eccentricities that have been encountered,
we’ll need to do another -rc with some changes to the warning
message (along with having it use the flash applet’s definition of
‘good environment’). Given that the WebSockets branch has been
stable, my thought is that we could merge that in for -rc4. Then
when we do work some more work on the websocket client, it’s just a
minor release to update the client instead of the whole websocket
environment too. What do others think?
Cheers, Aaron
-- Peter Boothe | ɯoɔ˙ǝlƃooƃ@ǝɥʇooqd <mailto:> |
ᴹ̶LAB | http://measurementlab.net
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
iQEcBAEBCgAGBQJVKFT0AAoJEL+9ounAjYBC7PsH/jp7vTD01C/6g+vM2iIT5AFh
PtTES6drgdjPcS/xqVgLJOOPrqo6QjPRIGHK5oo2rh9J9flxMGOK3hcw1QpJAapF
izL/+85PnpqO6pInkWnfiWYGUfkL/RkvGOfhI4o6r8dhuud49fppXxXLYCdseyr2
igzp5PhLU58ENwI9j864E0aFKjgC6hTFTusWid0AU+YoBE96HI8JfndtynMQsEjd
OU2Y0PMI5G3fvO+sFAnnLXQdup8xuX3YcGU8Ip42BMA4jfJjYDn3YIISE0O6yd8Y
SJrTMLXvEMzcZBE1PRmbcsjjzz3aqllZRMmch21uBsefBQUkJ2Opj8t5NqgkwXY=
=U96D
-----END PGP SIGNATURE-----
|
