These are chat archives for exceptionless/Discuss

20th
Jan 2016
Blake Niemyjski
@niemyjski
Jan 20 2016 01:53
so if (client.configuration.DefaultData.contains(“@user”) { submit session start..}
i like that it’s simple. I don’t like it for the same reason a plugin / lhamda could set the user
and we wouldn’t completely be killing those events from being sent
James Connor
@megakid
Jan 20 2016 11:40
Has anyone had any joy with running Exceptionless.Windows with .NET 4.5 ? I get a MissingMethodException on startup - looking into it now. Has anyone had this before?
Same binaries works fine on .NET 4.6
machine
Frank Ebersoll
@frankebersoll
Jan 20 2016 11:48
Good morning :)
Hey @megakid
Just stumbled upon the same thing right now
obviously it has been built with usage of "Array.Empty<T>" which only exists in .NET 4.6
we just cloned the .NET client and try to figure out if something is broken in regard to references
James Connor
@megakid
Jan 20 2016 12:00
@frankebersoll nice. Can you let me know your findings? I would have done the same, we can probably make it build targeting 4.5
Frank Ebersoll
@frankebersoll
Jan 20 2016 12:00
we didn't find it yet. the previous version (3.2) just works, so it seems to have been broken in between.
those are the changes
looking now :)
@frankebersoll It might just be the fact that it's been bumped from 4.5.1 to 4.5.2 ?
Frank Ebersoll
@frankebersoll
Jan 20 2016 12:10
thought about it, but i can't really tell.
aren't those changes in the config files only? if so, this shouldn't affect the runtime at all
let alone the build
James Connor
@megakid
Jan 20 2016 12:11
Hmm well it's the build is targeting v4.5.2 now
I just repro'd on a Win7 VM
Frank Ebersoll
@frankebersoll
Jan 20 2016 12:12
okay? i'm still on 4.5.1 over here
James Connor
@megakid
Jan 20 2016 12:12
yeh indeed
I think that's why, not sure where the MissingMethodException originates from but must be some internal .net method ?
can you change the target to 4.5.1 ? does it still compile?
Frank Ebersoll
@frankebersoll
Jan 20 2016 12:32
where do you see that the build is targeting 4.5.2?
James Connor
@megakid
Jan 20 2016 12:41
Well in the csproj files.
You're right though, it's only in the sample projects
James Connor
@megakid
Jan 20 2016 12:51
Hmm no clue
James Connor
@megakid
Jan 20 2016 13:41
I found the difference in dotPeek but I fear it's just a consequence of some build thing
ExceptionlessExtraConfigurationExtensions.ReadAllConfig(client.Configuration, (Assembly[]) Array.Empty<Assembly>());
vs
ExceptionlessExtraConfigurationExtensions.ReadAllConfig(client.Configuration);
it's a params method that hasn't changed between the versions
The strange thing is if I build v3.3.0 myself, it doesn't have that Array.Empty<Assembly>
Maybe the build script used to create the nuget binaries has overridden the targetted framework
James Connor
@megakid
Jan 20 2016 13:56
We've downgraded for now
Blake Niemyjski
@niemyjski
Jan 20 2016 14:12
Hmm
I'll take a look into this asap
Might be a resharper fix, I ran all the samples under 4.5.2.. Maybe this bug happens on machines that don't have .net 4.6 installed
thats everything thats changed
@megakid @frankebersoll do these machines you are running on have .net 4.6?
what’s the full stac trace that your getting
Blake Niemyjski
@niemyjski
Jan 20 2016 14:30
I just looked over the whole change log (did before we released as well
we used some c# 6 syntax (nameof and $”hi {worldVar}"
and changed the build os version to VS2015 but that’s just syntactic sugar
@megakid if I install Windows 7 SP1
Blake Niemyjski
@niemyjski
Jan 20 2016 14:35
can I install Vs2015 also? or do these machines not have anything installed on them?
please let me know how I can reproduce this
so I can work on getting it fixed asap
Frank Ebersoll
@frankebersoll
Jan 20 2016 14:45
hey blake! we have 4.5.1 without 4.6 here
just checking for the complete stack trace, give me a second.
Blake Niemyjski
@niemyjski
Jan 20 2016 14:46
ok
Frank Ebersoll
@frankebersoll
Jan 20 2016 14:47
MissingMethodException
MissingMethodException: Method not found: !!0[] System.Array.Empty()
Blake Niemyjski
@niemyjski
Jan 20 2016 14:48
anything else
Frank Ebersoll
@frankebersoll
Jan 20 2016 14:48
bei Exceptionless.ExceptionlessClientExtensions.Startup(ExceptionlessClient client, AppDomain appDomain)
bei Exceptionless.ExceptionlessExtensions.Register(ExceptionlessClient client)
bei ConsoleApplication1.Program.Main(String[] args) in c:\Users\febersol\Documents\Visual Studio 2013\Projects\AccsCoreTest\ConsoleApplication1\Program.cs:Zeile 14.
bei System.AppDomain._nExecuteAssembly(RuntimeAssembly assembly, String[] args)
bei System.AppDomain.ExecuteAssembly(String assemblyFile, Evidence assemblySecurity, String[] args)
bei Microsoft.VisualStudio.HostingProcess.HostProc.RunUsersAssembly()
bei System.Threading.ThreadHelper.ThreadStart_Context(Object state)
bei System.Threading.ExecutionContext.RunInternal(ExecutionContext executionContext, ContextCallback callback, Object state, Boolean preserveSyncCtx)
bei System.Threading.ExecutionContext.Run(ExecutionContext executionContext, ContextCallback callback, Object state, Boolean preserveSyncCtx)
bei System.Threading.ExecutionContext.Run(ExecutionContext executionContext, ContextCallback callback, Object state)
bei System.Threading.ThreadHelper.ThreadStart()
Blake Niemyjski
@niemyjski
Jan 20 2016 14:49
going to be fun to debug this if I can’t install VS2015
dang just googled MissingMethodException: Method not found: !!0[] System.Array.Empty()
Frank Ebersoll
@frankebersoll
Jan 20 2016 14:49
yes, somehow it seems to be a compiler thing
Blake Niemyjski
@niemyjski
Jan 20 2016 14:50
dotnet/roslyn#4889
Microsoft/msbuild#173
those dicks
Frank Ebersoll
@frankebersoll
Jan 20 2016 14:54
:-D
Another question. We're just incorporating the .NET client in our core.
What about inner exceptions? Should that automatically work?
Blake Niemyjski
@niemyjski
Jan 20 2016 14:54
yeah?
yeah, we break everything down that we can
with the exception of aggregate exceptions (we might miss some inners)
Frank Ebersoll
@frankebersoll
Jan 20 2016 14:55
Working in 3.2 already?
Blake Niemyjski
@niemyjski
Jan 20 2016 14:55
yes
Frank Ebersoll
@frankebersoll
Jan 20 2016 14:55
okay, i test again.
Blake Niemyjski
@niemyjski
Jan 20 2016 14:55
that should be working since 1.0
@all so what’s the verdict roll back to old build server and code that uses c# 6 syntax
just sucks, how are we going to support .NET core if this is still an issue
dotnet/roslyn#4889 read my comment
was reported over 6 months ago, they should of been on that so fast our heads were spinning
Blake Niemyjski
@niemyjski
Jan 20 2016 15:10
public static void ReadAllConfig(this ExceptionlessConfiguration config, params Assembly[] configAttributesAssemblies)
{
if (configAttributesAssemblies == null || configAttributesAssemblies.Length == 0)
ExceptionlessConfigurationExtensions.ReadFromAttributes(config, Assembly.GetEntryAssembly(), Assembly.GetCallingAssembly());
else
ExceptionlessConfigurationExtensions.ReadFromAttributes(config, configAttributesAssemblies);
ExceptionlessExtraConfigurationExtensions.ReadFromConfigSection(config);
ExceptionlessExtraConfigurationExtensions.ReadFromAppSettings(config);
ExceptionlessExtraConfigurationExtensions.ReadFromEnvironmentalVariables(config);
ExceptionlessConfigurationExtensions.ApplySavedServerSettings(config);
}
@frankebersoll what client is being referenced by your project
.NET 4.0 or 4.5?
can you send me the full path in the properties window when clicking on the assembly ref
Frank Ebersoll
@frankebersoll
Jan 20 2016 15:13
\packages\Exceptionless.3.3.1470\lib\net45\Exceptionless.dll
\packages\Exceptionless.3.3.1470\lib\net45\Exceptionless.Extras.dll
Blake Niemyjski
@niemyjski
Jan 20 2016 15:13
this a console or service app
ok
I’m downloading a vm right now
Frank Ebersoll
@frankebersoll
Jan 20 2016 15:13
console
this is two nested TargetInvocationExceptions with an inner ArgumentException
the innermost stack frame is nowhere to be seen
Blake Niemyjski
@niemyjski
Jan 20 2016 15:15
Can you create a unit test that reproduces and I’ll take a look.
Be easiest to debug and confirm it’s fixed
It’s hard to see it without actually seeing the real stack trace and inners
Frank Ebersoll
@frankebersoll
Jan 20 2016 15:16
            ExceptionlessClient.Default.Configuration.ApiKey = "q10yVifWUsduPXQrfF0ujwiq2SVcO1ose0XaTC94";
            ExceptionlessClient.Default.Register();

            try
            {
                try
                {
                    try
                    {
                        throw new ArgumentException("This is the inner argument exception", "wrongArg");
                    }
                    catch (Exception e1)
                    {
                        throw new TargetInvocationException("Target invocation exception. Blah blah blah blah.", e1);
                    }
                }
                catch (Exception e2)
                {
                    throw new TargetInvocationException("Outer Exception. This is some text of the outer exception.", e2);
                }
            }
            catch (Exception exception)
            {
                exception.ToExceptionless().Submit();
            }

            ExceptionlessClient.Default.ProcessQueue();
;-)
Blake Niemyjski
@niemyjski
Jan 20 2016 15:17
can you break on that submit ex handler
and paste the stack trace
Frank Ebersoll
@frankebersoll
Jan 20 2016 15:18
i did
Blake Niemyjski
@niemyjski
Jan 20 2016 15:18
just wanted to compare that stack trace with the one I see
Frank Ebersoll
@frankebersoll
Jan 20 2016 15:18
i even subscribed the Submitted event and looked into the event model and it contained the complete stack, with nested exceptions and frames
Blake Niemyjski
@niemyjski
Jan 20 2016 15:18
it should be an error module with an inner with an inner
Frank Ebersoll
@frankebersoll
Jan 20 2016 15:19
okay
System.Reflection.TargetInvocationException: Outer Exception. This is some text of the outer exception. ---> System.Reflection.TargetInvocationException: Target invocation exception. Blah blah blah blah. ---> System.ArgumentException: This is the inner argument exception
Parametername: wrongArg
   bei ExceptionlessTest.Program.Main(String[] args) in c:\Users\febersol\Documents\Visual Studio 2013\Projects\ExceptionlessTest\ExceptionlessTest\Program.cs:Zeile 22.
   --- Ende der internen Ausnahmestapelüberwachung ---
   bei ExceptionlessTest.Program.Main(String[] args) in c:\Users\febersol\Documents\Visual Studio 2013\Projects\ExceptionlessTest\ExceptionlessTest\Program.cs:Zeile 26.
   --- Ende der internen Ausnahmestapelüberwachung ---
   bei ExceptionlessTest.Program.Main(String[] args) in c:\Users\febersol\Documents\Visual Studio 2013\Projects\ExceptionlessTest\ExceptionlessTest\Program.cs:Zeile 31.
This is what happens on ToString
Blake Niemyjski
@niemyjski
Jan 20 2016 15:22
ok
just verified that we are breaking this down right
"inner": {
"message": "Target invocation exception. Blah blah blah blah.",
"type": "System.Reflection.TargetInvocationException",
"inner": {
"message": "This is the inner argument exception Parametername: wrongArg",
"type": "System.ArgumentException"
},
guess the ui is doing a poor job representing that it’s an inner exception eh
I’ll take a look into this today
the build breaking people is my first priority
installing a vm now
James Connor
@megakid
Jan 20 2016 15:26
@niemyjski - didn't read everything yet but it's caused by ExceptionlessExtraConfigurationExtensions.ReadAllConfig(client.Configuration, (Assembly[]) Array.Empty<Assembly>());
if you use decompiler on the Extra binary you will see that
when I compile the same commit myself in Release, I decompile and Array.Empty is gone
Blake Niemyjski
@niemyjski
Jan 20 2016 15:27
I used dot peeek
James Connor
@megakid
Jan 20 2016 15:27
it's just a symptom of wrong .NET targeting. Not sure how.
Blake Niemyjski
@niemyjski
Jan 20 2016 15:27
and the code is how it should be
I’m installing a windows 7 vm right now I’ll take a look
James Connor
@megakid
Jan 20 2016 15:27
yes but the same code is compiling to something different vs the nuget package
i agree the code is correct
Blake Niemyjski
@niemyjski
Jan 20 2016 15:29
take a look at the twitter link above
Blake Niemyjski
@niemyjski
Jan 20 2016 15:29
there might be a compiler optmization doing all of this when the build machine has .NET 4.6 installed
James Connor
@megakid
Jan 20 2016 15:29
and tested with 4.5 (failed) - then installed 4.6 (worked)
Frank Ebersoll
@frankebersoll
Jan 20 2016 15:29
no idea. seems like it
James Connor
@megakid
Jan 20 2016 15:30
I have .NET 4.6 installed on my build machine and I can build v3.3.0 without issues (works on .NET 4.5)
with the same source
Blake Niemyjski
@niemyjski
Jan 20 2016 15:30
wonder what the issue is then?
James Connor
@megakid
Jan 20 2016 15:31
I can send you the Exceptionless.Extra.dll v3.3.0 if you want to check the dotPeek to see the difference, but not sure the issue
That twitter is interesting
Think it must be that
Blake Niemyjski
@niemyjski
Jan 20 2016 15:33
sure
I’m using the one from nuget
just creating a console app and calling Register should reproduce right?
can I target 4.5.2 or should I target just for 4.5?
Well just v4.5 like your lib
the console app should repro it, yes
I have uploaded my dll here, you can see the different in that method
one has Array.Empty, one does not.
sorry - to be clearer. The official nuget binary (v3.3.1470.0) has Array.Empty. My self-built one (on 4.6.1 machine) does not (v3.3.0.0)
Blake Niemyjski
@niemyjski
Jan 20 2016 15:37
that assembly looks good too
what methods should I look at
James Connor
@megakid
Jan 20 2016 15:37
ExceptionlessClientExtensions.Startup
ah maybe it has PDB - 1 sec
Blake Niemyjski
@niemyjski
Jan 20 2016 15:40
I canceled out of downloading symbols and just looked at the decompiled assembly
gotta install windows updates and choose .net 4.5
not installed on windows 7 by default
James Connor
@megakid
Jan 20 2016 15:41
I just downloaded dotNetFx45_Full_setup.exe
no windows updates
Blake Niemyjski
@niemyjski
Jan 20 2016 15:42
from msdn/
James Connor
@megakid
Jan 20 2016 15:43
Ok so I removed my local PDB
Blake Niemyjski
@niemyjski
Jan 20 2016 15:43
?
James Connor
@megakid
Jan 20 2016 15:43
not sure if its the same but my link reproduces
so I removed local PDB and decompiled both your binary and my own - I get different ExceptionlessClientExtensions.Startup methods
Blake Niemyjski
@niemyjski
Jan 20 2016 15:46
yeah because pdb contains the real code and the other one is disassembled
James Connor
@megakid
Jan 20 2016 15:46
I removed the pdb
this is native disassembled
Blake Niemyjski
@niemyjski
Jan 20 2016 15:47
hmm
James Connor
@megakid
Jan 20 2016 15:47
maybe it gets from symbol server?
Blake Niemyjski
@niemyjski
Jan 20 2016 15:47
maybe
James Connor
@megakid
Jan 20 2016 15:47
not sure dotPeek behaviour
Blake Niemyjski
@niemyjski
Jan 20 2016 15:47
yeah it will grab from symbol server first
I clicked cancel
James Connor
@megakid
Jan 20 2016 15:47
me too
notice the Array.Empty<Assembly>()
Blake Niemyjski
@niemyjski
Jan 20 2016 15:48
well the great news is I can repo
yeah
James Connor
@megakid
Jan 20 2016 15:48
yey :)
Blake Niemyjski
@niemyjski
Jan 20 2016 15:48
dam
this stinks
James Connor
@megakid
Jan 20 2016 15:48
Yeh, sucks, surely frameworkTarget is 4.5 then it should actually be able to run on 4.5 lol
what a shitty bug
Blake Niemyjski
@niemyjski
Jan 20 2016 15:49
yeah
James Connor
@megakid
Jan 20 2016 15:50
from the twitter bug linked
seems as if you may have had 4.5.1 on your appveyor server
then straight to 4.6
Blake Niemyjski
@niemyjski
Jan 20 2016 15:50
yeah
James Connor
@megakid
Jan 20 2016 15:50
if you had gone 4.5.2 then 4.6 it wouldn't be an issue
Blake Niemyjski
@niemyjski
Jan 20 2016 15:51
so you think my build server doesn’t have 4.5.2
it just has 4.6? is 4.6 an inplace upgrade?
James Connor
@megakid
Jan 20 2016 15:51
that's what the bug report 6 months ago seems to suggest
not sure if you can install 4.5.2 over 4.6 then install 4.6 again
The bug suggests:Project target: 4.5.2
Build server: 4.6 (without 4.5.2 reference assemblies properly available)
Runtime server: 4.5.2
this is just stupid
James Connor
@megakid
Jan 20 2016 15:55
Agreed. Massive bug in the .NET compiler
How anyone can argue against that I'm not sure.
Blake Niemyjski
@niemyjski
Jan 20 2016 15:55
so until this is fixed we can’t use any c# 6 syntax features and we have to go back to 4.5 looks like
James Connor
@megakid
Jan 20 2016 15:56
hmm why?
hmm I think you can use c# 6 syntax, just targeting 4.5 - we do that here
Blake Niemyjski
@niemyjski
Jan 20 2016 15:56
we also can’t support .NET CORE unless we do a different repo
nope
requires the new compiler
James Connor
@megakid
Jan 20 2016 15:56
Ah, yes but you just need to make sure 4.5.2 assemblies are available before you install the new compiler
then all is good
so just uninstall 4.6 (if you can) then install 4.5.2
then install 4.6
Blake Niemyjski
@niemyjski
Jan 20 2016 15:57
ah
then I’ll be fine
James Connor
@megakid
Jan 20 2016 15:57
yes
It's the only way by the sounds of it
Blake Niemyjski
@niemyjski
Jan 20 2016 15:59
sent our ci an email
in the mean time I guess I should revert to use older syntax
and redeploy
cause they won’t update for a bit
James Connor
@megakid
Jan 20 2016 16:03
                c.Configuration.DefaultData[Event.KnownDataKeys.UserInfo] = new UserInfo
                {
                    Name = Environment.UserName,
                    Identity = Environment.UserName,
                };
I used that as a workaround for v3.2
Blake Niemyjski
@niemyjski
Jan 20 2016 16:05
don’t worry I’ll have a fix out in about an hour
James Connor
@megakid
Jan 20 2016 16:05
I see you can disable the heartbeats - would it be possible to disable the session stuff altogether?
in 3.3
Eric J. Smith
@ejsmith
Jan 20 2016 16:07
Sessions get automatically created when you send events with a user identity, but the auto created session events are not counted against your plan limit.
James Connor
@megakid
Jan 20 2016 16:09
Ok, I just didn't think the session entry and exit are very nice in the website, e.g. http://snag.gy/T8z7f.jpg
hard to tell which session is what
Blake Niemyjski
@niemyjski
Jan 20 2016 16:10
are you self hosting
James Connor
@megakid
Jan 20 2016 16:10
yes
Blake Niemyjski
@niemyjski
Jan 20 2016 16:10
yeah
that’s the problem, we haven’t updated the server or api builds yet
James Connor
@megakid
Jan 20 2016 16:11
ah cool
Blake Niemyjski
@niemyjski
Jan 20 2016 16:11
when we do it will look much much nicer
hopefully will do that this week
James Connor
@megakid
Jan 20 2016 16:11
ok cool, happy to leave it in then :)
Eric J. Smith
@ejsmith
Jan 20 2016 16:12
@niemyjski Maybe it should be a project level setting though.
for people that only send errors it seems silly.
hard for us to know what the right thing to do it.
is
unless we were to keep track of how many events were submitted for a single user identity and wait until that number was 2 within a window of time before we auto created the session event.
Blake Niemyjski
@niemyjski
Jan 20 2016 16:15
then you cant get number of user reports and other things
James Connor
@megakid
Jan 20 2016 16:15
Just an application level setting would be good
Eric J. Smith
@ejsmith
Jan 20 2016 16:15
what do you mean? user reports?
James Connor
@megakid
Jan 20 2016 16:15
then we can set our clients to create sessions but not our server-side services
Eric J. Smith
@ejsmith
Jan 20 2016 16:16
we can’t get any useful session / user level information right now unless people send session start events for all events.
otherwise our stats will say that you have had 100 sessions with 100% errors and 100% unhappy users
Sander Rijken
@srijken
Jan 20 2016 16:18
when changing a config value on the website, it's automatically pushed down to the clients right?
Eric J. Smith
@ejsmith
Jan 20 2016 16:18
kind of makes me wonder if we shouldn’t be automatically creating sessions. Maybe the sessions page should tell the user that they need to call SubmitSessionStart with the user’s identity.
@srijken yes
Blake Niemyjski
@niemyjski
Jan 20 2016 16:19
that sucks
Eric J. Smith
@ejsmith
Jan 20 2016 16:19
right now most people just send errors… and if that is all they do then creating sessions feels very useless.
but if they send anything else… then its nice that sessions get created automatically.
Sander Rijken
@srijken
Jan 20 2016 16:20
is that sent along with request/responses the client makes, or is it also pushed? ie if nothing happens.. does that still happen?
Blake Niemyjski
@niemyjski
Jan 20 2016 16:20
I need to be back in a bit
I need to focus on getting this issue fixed
Eric J. Smith
@ejsmith
Jan 20 2016 16:20
if they send log messages or feature usages or page views… then it would be nice if they didn’t have to specifically send a session event.
@srijken it happens during event submission… when the client submits an event a header comes back with the config version.
if the version is wrong then it will fetch the new config.
so you do have to wait for the next event to be submitted.
we did that because we didn’t want to have tons of requests polling to see if the config has changed.
Sander Rijken
@srijken
Jan 20 2016 16:22
right, the thing is, I want to configure wether to log or not..
yeah I understand
Eric J. Smith
@ejsmith
Jan 20 2016 16:23
@srijken the setting will get picked up as soon as you send another event. If you wanted to, you could have an endpoint on your site that sends a dummy event to trigger it to update.
most apps send events enough that it is pretty much real time.
Blake Niemyjski
@niemyjski
Jan 20 2016 16:23
no
you can just call SettingManager.UpdateSettings() and that will go and update them
Eric J. Smith
@ejsmith
Jan 20 2016 16:24
no what?
Sander Rijken
@srijken
Jan 20 2016 16:24
yeah, I'll do that, good to know
oh right, even better
Eric J. Smith
@ejsmith
Jan 20 2016 16:24
yeah, forgot about that.
do we have a method for that off of the client?
like client.Config.UpdateSettings() ?
Blake Niemyjski
@niemyjski
Jan 20 2016 16:25
no
we didn’t want users doing it
Sander Rijken
@srijken
Jan 20 2016 16:25
lol
Blake Niemyjski
@niemyjski
Jan 20 2016 16:25
we had a discussion about how we do a good job at managing it
Eric J. Smith
@ejsmith
Jan 20 2016 16:25
so what instance of config is that calling for?
Blake Niemyjski
@niemyjski
Jan 20 2016 16:25
and for those that want to do something hard core they could just call the settings manager
you pass it in
so the argument could be that, maybe a helper / extension method would be nice there
Blake Niemyjski
@niemyjski
Jan 20 2016 17:14
exceptionless/Exceptionless.Net@a4af6ca
@frankebersoll @megakid I have a build on myget
can you try it
it works on my windows 7 vm now
Blake Niemyjski
@niemyjski
Jan 20 2016 17:19
feel free to look over https://github.com/exceptionless/Exceptionless.Net/compare/v3.3.0...master to make sure I didn’t miss anything
can’t deploy due to nuget sigh
Eric J. Smith
@ejsmith
Jan 20 2016 18:26
nuget down?
Blake Niemyjski
@niemyjski
Jan 20 2016 18:38
yeah
myget or symbolsource
Blake Niemyjski
@niemyjski
Jan 20 2016 20:07
@frankebersoll @megakid pushing out new builds as we speak
Frank Ebersoll
@frankebersoll
Jan 20 2016 20:12
Great stuff. Will test tomorrorw morning in the office
Blake Niemyjski
@niemyjski
Jan 20 2016 20:12
:)
thanks guys'
Frank Ebersoll
@frankebersoll
Jan 20 2016 20:13
Thank you!
Blake Niemyjski
@niemyjski
Jan 20 2016 21:08
@frankebersoll so our ui only shows one level of indention on nested stack traces
which might makethis harder to view
wonder if we should indent everything x spaces
Blake Niemyjski
@niemyjski
Jan 20 2016 21:16
exceptionless/Exceptionless.UI#64
Do you agree with that?
Sander Rijken
@srijken
Jan 20 2016 21:55
Not sure who you’re asking, but it sounds better
Blake Niemyjski
@niemyjski
Jan 20 2016 21:56
like right now everything is indented 2 spaces
so all inner exceptions are just 2 spaces in from the root, what I suggested is each inner is +2 spaces indented'
Eric J. Smith
@ejsmith
Jan 20 2016 21:57
the current output mimics .net output.
Sander Rijken
@srijken
Jan 20 2016 21:58
might also help with this? exceptionless/Exceptionless.Net#15
Blake Niemyjski
@niemyjski
Jan 20 2016 22:01
I get that but @frankebersoll or @srijken contacted me earlier (don’t remember which one) to verify that the inners were captured correctly. And it’s kind of hard to see what was an inner
Eric J. Smith
@ejsmith
Jan 20 2016 22:02
doesn’t it show — inner —— ?
Blake Niemyjski
@niemyjski
Jan 20 2016 22:02
@srijken yeah it might help with that, but still need to address it
@ejsmith yeah, we show it but it gets confusing when your more than one level deep..
if we just indented each inner child by 2 spaces it would be much clearer

@all I closed a bunch of issues relating to notifications and I created this exceptionless/Exceptionless#177 If you could give your feedback on this it would be greatly appreciated.
Frank Ebersoll
@frankebersoll
Jan 20 2016 22:22
@niemyjski IIRC parts of the innermost stack were not shown
in particular, the source location
I will check again tomorrow
Blake Niemyjski
@niemyjski
Jan 20 2016 22:39
I see everything
what part
can you paste the raw stack trace string here exceptionless/Exceptionless.UI#64
I don’t see anything that is missing, I’ll have to run that sample tomorrow to double check
.. I need to run to a meeting shortly