by

Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
  • 03:22
    a14907 commented #629
  • 03:19
    ejsmith closed #629
  • 03:19
    ejsmith commented #629
  • 03:11
    a14907 commented #629
  • 03:09
  • 00:51

    niemyjski on discard-join

    Added some more test data (compare)

  • Jun 02 21:29

    ejsmith on master

    Update k8s resources (compare)

  • Jun 02 13:18
    niemyjski commented #620
  • Jun 02 06:28
    jasonwang1109 starred exceptionless/Exceptionless
  • Jun 02 05:13
    chengfeng19 commented #620
  • Jun 02 02:15
    JustinJane starred exceptionless/Exceptionless
  • Jun 02 01:47
  • Jun 01 12:13

    dependabot-preview[bot] on nuget

    (compare)

  • Jun 01 12:13
    dependabot-preview[bot] closed #633
  • Jun 01 12:13
    dependabot-preview[bot] commented #633
  • Jun 01 12:13
    CLAassistant commented #634
  • Jun 01 12:13
    dependabot-preview[bot] labeled #634
  • Jun 01 12:13

    dependabot-preview[bot] on nuget

    Bump Stripe.net from 35.9.0 to … (compare)

  • Jun 01 12:13
    dependabot-preview[bot] opened #634
  • May 31 03:01
    houlintao starred exceptionless/Exceptionless
Blake Niemyjski
@niemyjski
yeah
Eric J. Smith
@ejsmith
exactly
Blake Niemyjski
@niemyjski
I’ll add a test to the master branch
and then pull it into this one
I’ll make that a toodo
Sander Rijken
@srijken
ex collegue used to do that.. optimize for optimal reuse (and then it doesn’t get reused), and for optimal speed, when speed isn’t the issue there
Blake Niemyjski
@niemyjski
yeah
I know
I’ll write a test and add some good test cases
Sander Rijken
@srijken
especially the first thing leads to code more complex than necessary
livecodign has delay :) I see the message come in, and then I see you typing it
Eric J. Smith
@ejsmith
drives me insane when people make architectural design decisions based on microbenchmarks.
build for simplicity and maintenance… profile the entire thing and fix the real hotspots.
Blake Niemyjski
@niemyjski
you really think I need to do smart timers
lots of extra code for just this that runs every 60 seconds
Eric J. Smith
@ejsmith
probably not
Blake Niemyjski
@niemyjski
kk
sounds good
Eric J. Smith
@ejsmith
its kind of nice to only fire when you need to, but probably not worth it.
Blake Niemyjski
@niemyjski
I’ll commit these changes and we’ll have to test this
yeah
I don’t think it’s worth it
Eric J. Smith
@ejsmith
sounds good.
Blake Niemyjski
@niemyjski
we gotta make sure we don’t use any c# 6 features in this client or it breaks the build
Sander Rijken
@srijken
ah
oh right that thing
Blake Niemyjski
@niemyjski
Any chance you could update all occurrences of ?.
Sander Rijken
@srijken
that sucks.. wish I knew about this
Blake Niemyjski
@niemyjski
yeah :(
Sander Rijken
@srijken
I didn’t see the build turn red either when I committed stuff
brb
Blake Niemyjski
@niemyjski
k
Sander Rijken
@srijken
@niemyjski there’s this R# key you can put in to have that warn about C#6. I’m gonna use that to find all the problems, but maybe you want to have that in there as well?
Blake Niemyjski
@niemyjski
yeah
where can I put that in
should be in the resharper settings for that project
Sander Rijken
@srijken
would be nice if that worked
tried, it came up with them, and then they were gone
gotit
properties window of the .Portable project
UGH
the first fix it suggests is “turn on C#6 support"
and there’s no other fix
Frank Ebersoll
@frankebersoll
Had a discussion at work today about performance of string comparisons. Colleagues wanted to do GetHashCode to compare two long strings, but it turned out to be a bad idea: GetHashCode would need to be
...called every time and it takes about 4 times longer than comparing the strings themselves
Sander Rijken
@srijken
makes sense
Frank Ebersoll
@frankebersoll
that's because string comparison and GetHashCode both need to look at every single character, but GetHashCode needs to multiply and xor things.
So, actually it is more costly and doesnt add any performance when used on few instances.
So, maybe we need more than one hashcode? Could we have a fail-fast-hash that only uses few properties?