20 years, still full crash on teleport

[Sid Nagy]

Just now, JeromFranzic said:

I didn't attend the opening ceremony, but I was in several crowded sims during that period. Not once did I crash, and in all cases, was able to maintain 25-30 fps overall. (Of course I had to turn off shadows and limit draw distance to 96m with the crowds, and no jelly dolls... not bad for a PC from 2018. )

And that is where I went wrong, didn't touch a slider and jumped right in. I should have known better after all these years.
But than I decided to watch Patch's speech on Youtube. The result was the same in the end.

[Paul Hexem]

4 hours ago, Love Zhaoying said:

It's definitely you.

And if you're going to tell me you've never seen the "view IMs or quit" dialog, I'm going to call you a liar to your face.

[Love Zhaoying]

2 minutes ago, Paul Hexem said:

And if you're going to tell me you've never seen the "view IMs or quit" dialog, I'm going to call you a liar to your face.

I see it extremely rarely. Maybe once every few months - and only if I log in daily or more. Sorry, your initial post made me think you still saw the error "often".

Edited August 8, 2023 by Love Zhaoying

[Paul Hexem]

7 minutes ago, Love Zhaoying said:

I see it extremely rarely. Maybe once every few months - and only if I log in daily or more. Sorry, your initial post made me think you still saw the error "often".

Often is subjective in this case. At the time of the post originally (note the date) I was seeing it somewhat often- that was due to teleports sometimes actually tripping my network load balancer, but the post is more complaining that it exists at all.

After two decades, not only do we still run the risk of being disconnected while performing a routine, required task but it's made worse by the fact that when it happens, Linden Lab who say they care about user retention, only offer the option to quit, not reconnect.

[Sid Nagy]

I think it all also depends on what other programs one is running at the same time with SL.
For texturing I often use PaintShop Pro. Not the best friends with SL. 
When I use both together for a longer time, I occasionally get the said screen from SL.
But who is to blame PaintShop, SL or both?

Edited August 8, 2023 by Sid Nagy
Oops, I thought it was Saturday already: my edit the text hobby day.

[Love Zhaoying]

17 minutes ago, Paul Hexem said:

After two decades, not only do we still run the risk of being disconnected while performing a routine, required task but it's made worse by the fact that when it happens, Linden Lab who say they care about user retention, only offer the option to quit, not reconnect.

I think the viewer developers were lazy, quite frankly. No matter how badly the viewers are coded, "restarting an application" certainly should be possible. Even with a hard crash.

[CandyCole]

I cannot remember the last time I crashed ..
.. which wasn't in or around a teleportation.

All these years and I still mentally cross my digits when teleporting.

[Henri Beauchamp]

2 hours ago, Love Zhaoying said:

I think the viewer developers were lazy, quite frankly. No matter how badly the viewers are coded, ”restarting an application” certainly should be possible. Even with a hard crash.

I'm afraid you are mixing things together...

A disconnection is not a crash. A crash happens when the viewer executes an illegal instruction (e.g. accessing some data out of the allocated address space, or jumping/branching/”returning” to a faulty address containing junk/random memory, causing faulty opcodes to be executed). In this case, the viewer will just ”vanish” or the OS will report a crash and close it. If this happens to you, you should report the crash to the viewer developers (with the necessary data, i.e. the stack trace or crash dump, and the logs), so that the crash can be fixed; I personally never crash any more on TP with my viewer (and the rare times I did in the past, I fixed the corresponding bugs).

On the other hand, a TP may fail because the viewer failed to connect to the arrival sim server; the said disconnection could be the result of a bad network (lost packets), or a race condition (messages arriving out of order or too late to allow a proper connection sequence), or a failed handover between servers (for the same kind of causes: network or race issues). In this case, the viewer does not have any server left to speak with (not even the departing sim, which has already disconnected), and it will report a failure and present you with a grey screen, indicating it is disconnected; at this point, the validity of the data it got cannot be guaranteed any more, and the viewer does not ”know” what was the reason for the disconnection. A reconnection is therefore impossible, and the best course of action is to ask the user to restart a fresh viewer session and reconnect. True, it would be theoretically possible for the viewer to clean up all its memory and propose a reconnection from the login screen (like if you had relaunched it), but it would be complex to implement for a rare occurrence, and the resulting second session would be handicapped by side effects of the previous one (such as virtual address space fragmentation); it is much cleaner and safer to quit and restart the viewer for good.

The TP protocol in SL is however way too fragile (I already made a few suggestions in the Server User Group meetings to strengthen it). The lack of a proper handshaking protocol, and the fact the departing sim does not wait for the TP to complete (i.e. for the arrival sim to take over after a successful connection with the viewer) cause these cases when the viewer is ”left in the blue”, with no server to speak to !

Edited August 8, 2023 by Henri Beauchamp

[Ayashe Ninetails]

3 hours ago, Paul Hexem said:

And if you're going to tell me you've never seen the "view IMs or quit" dialog, I'm going to call you a liar to your face.

Omg. Like every time I'm logged in.

I teleport a lot. Store to store to store - especially if I'm grabbing demos/checking out weekend sales. It's pretty inevitable that one of those TPs will fail. I use Firestorm. I rarely teleport or go anywhere when I'm on Black Dragon since I've got all my graphics settings cranked on that one.

Whether or not I feel like logging back in is up to my mood at the time. Sometimes I don't bother, unless I was in the middle of a conversation (and I've since learned not to TP during convos!).

[Love Zhaoying]

28 minutes ago, Henri Beauchamp said:

I'm afraid you are mixing things together...

A disconnection is not a crash. A crash happens when the viewer executes an illegal instruction (e.g. accessing some data out of the allocated address space, or jumping/branching/”returning” to a faulty address containing junk/random memory, causing faulty opcodes to be executed). In this case, the viewer will just ”vanish” or the OS will report a crash and close it. If this happens to you, you should report the crash to the viewer developers (with the necessary data, i.e. the stack trace or crash dump, and the logs), so that the crash can be fixed; I personally never crash any more on TP with my viewer (and the rare times I did in the past, I fixed the corresponding bugs).

On the other hand, a TP may fail because the viewer failed to connect to the arrival sim server; the said disconnection could be the result of a bad network (lost packets), or a race condition (messages arriving out of order or too late to allow a proper connection sequence), or a failed handover between servers (for the same kind of causes: network or race issues). In this case, the viewer does not have any server left to speak with (not even the departing sim, which has already disconnected), and it will report a failure and present you with a grey screen, indicating it is disconnected; at this point, the validity of the data it got cannot be guaranteed any more, and the viewer does not ”know” what was the reason for the disconnection. A reconnection is therefore impossible, and the best course of action is to ask the user to restart a fresh viewer session and reconnect. True, it would be theoretically possible for the viewer to clean up all its memory and propose a reconnection from the login screen (like if you had relaunched it), but it would be complex to implement for a rare occurrence, and the resulting second session would be handicapped by side effects of the previous one (such as virtual address space fragmentation); it is much cleaner and safer to quit and restart the viewer for good.

The TP protocol in SL is however way too fragile (I already made a few suggestions in the Server User Group meetings to strengthen it). The lack of a proper handshaking protocol, and the fact the departing sim does not wait for the TP to complete (i.e. for the arrival sim to take over after a successful connection with the viewer) cause these cases when the viewer is ”left in the blue”, with no server to speak to !

No, I wasn't mixing things up - but I see how you may have thought I was. I did not mean to imply disconnects are "crashes", at all - just that a viewer could be written to restart / reconnect if there were a disconnect "or even a crash".

Please don't take this the wrong way: a lot of the reasons you give are not good excuses for not allowing a reconnect option - in my professional opinion.

My reason for mentioning a "crash" was: sometimes the disconnect message might be shown when it's "more than just a disconnect".

My additional assumption is: any application can be written so as to restart; the viewer developers (not YOU!) never added a "reconnect" feature because it's difficult to ensure everything is stable.

Reminder: I'm a C++ developer with more than 25 years experience.

 

Edited August 8, 2023 by Love Zhaoying

[Love Zhaoying]

10 minutes ago, Ayashe Ninetails said:

Omg. Like every time I'm logged in.

I teleport a lot.

Ironically, I had a rare disconnect yesterday just standing on my parcel. (I don't go places often, Home is where I stand and script!)

[CandyCole]

40 minutes ago, Henri Beauchamp said:

On the other hand

Your understanding is far superior to my own.

Semantics aside and without explaining the ins and outs of a duck's ar5e. One minute I'm playing SL. The next minute I'm not.

1GB fiber connection, cutting edge laptop (18 months ago). Teleport failure and all I can do is spin on the spot where I was.

Doesn't happen all the time, perhaps one in 20 or 25 TPs on average. I accept this seemingly random failure as part of SLife.

The world isn't perfect, neither is SL. It is what it is. Frustrating nevertheless.

At least it doesn't happen all the time, but I retain the mindset in which it could happen at any time.

p.s. Failed teleports occur enough for me to ALWAYS detach and re-add any attached item I may have edited or adjusted. Once detached/re-added the change just made will be preserved. If you adjust a fitting or texture or tint or something via a HUD etc. and subsequently become disconnected, your changes will not be saved. 

Edited August 8, 2023 by CandyCole
Just to add the p.s.

[Paul Hexem]

29 minutes ago, Henri Beauchamp said:

A disconnection is not a crash.

This is nothing against you personally, since you're spot on and providing accurate information; Unfortunately all of it is irrelevant in this particular case. I've parroted in the forums here a concept called "quit moments" in MMO's. These are things that make a player leave and not come back. SL has much in common with MMO's, including quit moments, and a popup that literally tells the user to quit... Is a pretty serious quit moment that absolutely hurts retention.

Linden Lab is objectively, inarguably dropping the ball on this one.

 

15 minutes ago, Ayashe Ninetails said:

Whether or not I feel like logging back in is up to my mood at the time. Sometimes I don't bother, unless I was in the middle of a conversation (and I've since learned not to TP during convos!).

And you're invested. Imagine a brand new user with nothing invested, in today's world of countless other distractions.

[Ayashe Ninetails]

5 minutes ago, Love Zhaoying said:

Ironically, I had a rare disconnect yesterday just standing on my parcel. (I don't go places often, Home is where I stand and script!)

I don't see that happen much, luckily. These are specifically teleports stalling out where the progress bar just slows to a crawl and eventually dies. If you cancel, the black screen goes away and I can see myself again at my starting point, but can't move or talk until the viewer notifies me it's over my shenanigans, the screen grays out, and I get the dreaded View IMs popup. 😂

 

Just now, Paul Hexem said:

And you're invested. Imagine a brand new user with nothing invested, in today's world of countless other distractions.

Exactly. It's so annoying that I'll put off my shopping trip or photography session and go do something else. I used to have a higher tolerance for that sort of thing (crashes when zoning between areas in an MMO was a constant thing back in the day), but not today, Satan!

[Henri Beauchamp]

58 minutes ago, Love Zhaoying said:

My reason for mentioning a ”crash” was: sometimes the disconnect message might be shown when it's ”more than just a disconnect”.

Sorry then, but your reasoning is flawed: if the viewer shows a message and then quits cleanly, it did not crash...

Again, a crash is due to illegal code being executed by the CPU, due to a bug; this is NOT the case here...

1 hour ago, Love Zhaoying said:

My additional assumption is: any application can be written so as to restart; the viewer developers (not YOU!) never added a ”reconnect” feature because it's difficult to ensure everything is stable.

I have been developing the Cool VL Viewer for the past 16 years. It never came to my mind that implementing a reconnect (from login screen, since there is no other way to reconnect cleanly to SL) would be ”nice to have” and even less worth it !

The solution to failed TPs is NOT to implement a ”reconnect without relaunch” feature, but to implement a reliable TP protocol with proper handshaking between the departing sim, the arrival sim and the viewer.

1 hour ago, Love Zhaoying said:

Reminder: I'm a C++ developer with more than 25 years experience.

Reminder: I'm an assembly, C and C++ developer with over 40 years of experience. 👴

[Love Zhaoying]

18 minutes ago, Henri Beauchamp said:

Sorry then, but your reasoning is flawed: if the viewer shows a message and then quits cleanly, it did not crash...

Again, a crash is due to illegal code being executed by the CPU, due to a bug; this is NOT the case here...

I have been developing the Cool VL Viewer for the past 16 years. It never came to my mind that implementing a reconnect (from login screen, since there is no other way to reconnect cleanly to SL) would be ”nice to have” and even less worth it !

The solution to failed TPs is NOT to implement a ”reconnect without relaunch” feature, but to implement a reliable TP protocol with proper handshaking between the departing sim, the arrival sim and the viewer.

Reminder: I'm an assembly, C and C++ developer with over 40 years of experience. 👴

Damn, you're old! Yeah, I only worked 35 years now. I wrote 25 years earlier because of the current job. I started assembly language in 1983, and some BASIC in the 70's before microcomputers were available to the public, C in college, etc. I don't want to work on Viewers because that sounds like a "job", and you awesome TPV developers are doing a great job!

Its nice that we can agree to disagree. 😺

I never should have used the word "crash", as that seems to have caused confusion still on what I actually meant. You confirmed that it's often more than a "simple disconnect". (I do know how to debug hangs vs. crashes, etc. so I know what a "crash" is. Thanks for encourage me to be more precise in the future.)

No disrespect was intended  towards towards your massive experience and knowledge.

Edited August 8, 2023 by Love Zhaoying

[Paul Hexem]

17 hours ago, Henri Beauchamp said:

The solution to failed TPs is NOT to implement a ”reconnect without relaunch” feature, but to implement a reliable TP protocol with proper handshaking between the departing sim, the arrival sim and the viewer.

The real question is, which is quicker and easier?

Because that's the route LL will want to go, if they ever get off their asses at all.

[Paul Hexem]

17 hours ago, Love Zhaoying said:

I never should have used the word "crash", as that seems to have caused confusion still on what I actually meant.

To be fair, for the layman, even if it it includes an error message it's a "crash".

[Love Zhaoying]

3 minutes ago, Paul Hexem said:

To be fair, for the layman, even if it it includes an error message it's a "crash".

Yep! And despite what was said earlier, there are many types of crashes (not just "illegal instructions" or whatever Henri wrote). I was looking for the "CDB/WinDBG" list but it was easier to get the list from Wikipedia:

"Typical causes include accessing invalid memory addresses, incorrect address values in the program counter, buffer overflow, overwriting a portion of the affected program code due to an earlier bug, executing invalid machine instructions (an illegal or unauthorized opcode), or triggering an unhandled exception."

  https://en.wikipedia.org/wiki/Crash_(computing)

 

[Henri Beauchamp]

43 minutes ago, Love Zhaoying said:

not just ”illegal instructions” or whatever Henri wrote

I wrote:

Quote

A crash happens when the viewer executes an illegal instruction (e.g. accessing some data out of the allocated address space, or jumping/branching/”returning” to a faulty address containing junk/random memory, causing faulty opcodes to be executed).

And:

Quote

Again, a crash is due to illegal code being executed by the CPU, due to a bug

Accessing an invalid address (including buffer overflows) or corrupting the program counter were parts of what I wrote.

You should learn to read...

57 minutes ago, Paul Hexem said:

To be fair, for the layman, even if it it includes an error message it's a ”crash”.

Not when the error message is the result of a disconnection and the viewer just tells you that it cannot continue to render the 3D world, and instead lets you the choice to quit or continue reading your IMs, chat, then quits cleanly after you decide you had enough. There is no bug, no crash, just an impossibility to keep rendering after teh disconnection happened.

[Paul Hexem]

42 minutes ago, Henri Beauchamp said:

Not when the error message is the result of a disconnection and the viewer just tells you that it cannot continue to render the 3D world, and instead lets you the choice to quit or continue reading your IMs, chat, then quits cleanly after you decide you had enough. There is no bug, no crash, just an impossibility to keep rendering after teh disconnection happened.

I don't care what the error message is a result of. Adding a button to click after you "decide you had enough" of staring at a broken, unrecoverable viewer that as you say, can no longer function, that's just a crash with extra steps. 

It's the digital equivalent of saying "I meant to do that" after you trip and fall.

Although to be honest, that's all irrelevant anyway, like I said. There's really no defending this- The fact that this still happens after 20 years and the option is "quit" and not "reconnect" makes LL blatant liars when they say they care about user retention.

[EnCore Mayne]

14 hours ago, Paul Hexem said:

I don't care what the error message is a result of. Adding a button to click after you "decide you had enough" of staring at a broken, unrecoverable viewer that as you say, can no longer function, that's just a crash with extra steps. 

It's the digital equivalent of saying "I meant to do that" after you trip and fall.

Although to be honest, that's all irrelevant anyway, like I said. There's really no defending this- The fact that this still happens after 20 years and the option is "quit" and not "reconnect" makes LL blatant liars when they say they care about user retention.

i share your incredulity that the Lab's pool of experienced software developers by which they can draw from can't seem to fix a simple error.

however, and i'm not defending the Lab's inability to manage their spagetti code, have you ever looked at the source files?

like i said before, to great accolades of positive responses, the code is a steaming pile of discombobulated patchwork c***. finding, let alone fixing ANYTHING has seemingly disassociated and catastrophically deleterious effects. it's no wonder they don't want to do anything with it. we is on thin ice. best rest on our laurels.

[Zalificent Corvinus]

On 8/8/2023 at 10:52 PM, Ayashe Ninetails said:

I don't see that happen much, luckily. These are specifically teleports stalling out where the progress bar just slows to a crawl and eventually dies. If you cancel, the black screen goes away and I can see myself again at my starting point, but can't move or talk until the viewer notifies me it's over my shenanigans, the screen grays out, and I get the dreaded View IMs popup. 😂

I've had this a lot... At the time that the Great Oz  was claiming crash to desktop TP fails were "only 2 %" and refusing to comment on "view im/chat or quit" TP failure, I was typically seeing the disconnect fails about 1 TP in 7.

But here's a weird thing.

If you alt-tab out of SL, click your internet connection thing in the task trayy, DISCONNECT from the internet, instantly RECONNECT, and alt-tab back to SL, you have maybe a 50/50 chance that the "spinning in place for 60 seconds before grey screen of view im/chat or quit" thing will be fixed and you can move normally again.

 

Now, if you tried something like that back in the days of Real Arizona Data Centre, instant disconnect logout.

But the assorted kludges the devs have had to make to compensate for the awful quality of Cloudcrap, now sometimes allow you to get away with this, literally toggling your internet off and back on before the busted ass server code notices..

Theoretically, you should NOT be able to do this, but you can. At least from the other side of the Atlantic, with 170ms plus ping times, might be less ssuccessful if closer with lower ping. Somebody should do some SCIENCE here.

Go figure.

[Gabriele Graves]

24 minutes ago, Zalificent Corvinus said:

But here's a weird thing.

If you alt-tab out of SL, click your internet connection thing in the task trayy, DISCONNECT from the internet, instantly RECONNECT, and alt-tab back to SL, you have maybe a 50/50 chance that the "spinning in place for 60 seconds before grey screen of view im/chat or quit" thing will be fixed and you can move normally again.

I've even had my cat accidentally hit the sleep button on my keyboard whilst I'm inworld which hibernates the computer and shuts almost everything off (lights off, no fan noise, everything) except it waiting for keyboard/mouse to hit/moved before waking up again and still been able to continue with my logged in SL session without having to restart the viewer.

For me this works just fine 100% of the time as long as I wake it up as soon as I can - round trip is maybe a minute, minute and a half.

[Gabriele Graves]

The viewer never crashes during TP for me.  However, it does fail to teleport maybe 1 in a dozen times and then disconnects me.  This is especially noticeable during weekend sales and it's very annoying.

I have a world-class quality gigabit fibre connection to the internet (for at least 5 years) which is incredibly reliable and I don't suffer from any significant packet loss according to the viewer stats ~0.1% even if logged in for hours and hours.

Yet these disconnections still happen.  Here's the kicker.  It has been a lot better in the past.  I remember a time a few years ago when LL were working on region crossings that it seemed from my experience that TP disconnects were a thing of the past and were no longer happening.  Gradually over time they've become more prevalent for me again.

I think it could be better.  It has been better.