292 replies to this topic

[John Mechlane]

I had almost no friggin idea what i was reading, since i'm not a programmer. But this is the way you should communicate with players...I can see why fixing this bug took this long....it was a b*tch. Keep up the good work!

[Belorion]

Here is a more in depth explanation of red black trees if anyone is interested.

http://www.drdobbs.c...trees/184410531

[Varrin Coursca]

Thank you for showing us all just how complex the problem was. Hopefully folks will take this case as an example of why the developers can't just wave a magic wand sometimes (even though, I really wish they would... ).

[Brawler1986]

Just Wow... Amazing that you guys have located this fatal error. Just bizarre that this error (which is nested deeply into the CryEngine) makes it all the way to the top, causing UI errors. Kudos for finding this. This makes my money worth it! Keep posting/communicating like this.

Edited by Brawler1986, 15 May 2013 - 04:45 AM.

[pesco]

Thanks for this! Great read.

[Effectz]

Keep up the good work!Thank you for sharing why it took so long!

[Falcore]

It really sounds like they had 1 person working on the issue for a few weeks then they had to bring in some extra help from other areas because they couldn't figure it out. It happens, depending on how your give bugs priority low levels of being able to reproduce would most likely have only assigned it to one person. I assume they are using agile and sprints it would have been a sprint and bam no resolution well lets double the resources looking at it. Another sprint goes by still no resolution, you double the resources again. This is what we tend to do for non priority 1 bugs for my work. At some point someone hit a panic button and it was get this crap fixed. But I can see how it went 6+ weeks of no resolution, assuming 2 week sprints.

[Shumabot]

jakucha, on 14 May 2013 - 08:04 PM, said:

Except they just said Karl was working with them since the beginning of the game. Pretty clear you're just putting forth pessimistic speculation because you're upset over something.

The very beginning of the article states explicitly that Karl wasn't working on the HUD issues until late in the game. He was with the company, not working on HUD bugs, software development works in silos. Also, yes. I'm pessimistic and this game has done everything to deserve pessimism. The very first game I ever played had a missing hud and that was like 4 months ago, I'm also down to averaging ~20 fps, I was getting 40 2 months ago.

PGI accomplished something? Better make it into some sort of parade and embellish it to the point where the fact that they launched a public product with a major game beaking bug that had been pre existing and took half a year to fix no longer matters. I'm sure Karl is a very skilled programmer, I'm not trying to lesson his accomplishment. My issue is with the management and I wonder why they didn't dedicate manpower to this issue in the first ~8 months that it was a major problem.

Edited by Shumabot, 15 May 2013 - 05:45 AM.

[Zaptruder]

Congratulations to PGI on squashing this particular family of bugs. Must've been quite the moment once you guys were able to verify that there were no longer any occurrences of the issue.


But next time... next time... please give players a work around while you guys fix this.

i.e. allow us to use a console or a command to reinitialize the HUD, or reload the map - if we detect or identify problems with our games.

In this case, it's no longer required. But a function that would be supremely useful is in giving players the ability to reconnect to a game that they were just in - while you guys look for the issues that cause connection drop outs.

If it's not already a top priority - it should be made a top priority - even if it takes from the focus of actually fixing the connection issues; because those work arounds help to make frequent drop outs and other issues of a similar nature significantly more bearable, and because PGI, much less the playerbase can never be certain how long it will take or how complex the issues underlying these faults are.

It'd be senseless to wait on such a feature that can mitigate 75% of the harm of a disconnect if fixing those issues take 3-6 months. That's millions of hours of player times and player games wasted or degraded while waiting for a fix with no certain fix time.

[Tastian]

Great read. I'm a computer gaming programmer and I empathize with what you went through. Programming an integrated flash reader into our games, we also hit the case sensitivity issue with actionscript to c++. And, we dumped STL a long time ago because 1) its difficult to track and 2) its slow. And yes, we have bugs that takes us months to find - the worst offenders are generally lost pointers. I'm super excited you guys managed to find this. And I know you guys are as excited - hence the detailed write-up. Good job.

Zaptruder, on 15 May 2013 - 05:57 AM, said:

But next time... next time... please give players a work around while you guys fix this.

You aren't a programmer are you?

[StalaggtIKE]

First off, much more stable. Thanks.

I am still getting a few IFF issues. I "lose" some of my team mates on the minimap when not looking in their direction. And before anyone says it, this is not related to ECM. This happens even at the start of the match, where the enemy's ECM is too far away to distort my HUD. I hope this can be addressed as well.

Edited by StalaggtIKE, 15 May 2013 - 06:17 AM.

[Shumabot]

StalaggtIKE, on 15 May 2013 - 06:16 AM, said:

First off, much more stable. Thanks.

I am still getting a few IFF issues. I "lose" some of my team mates on the minimap when not looking in their direction. And before anyone says it, this is not related to ECM. This happens even at the start of the match, where the enemy's ECM is too far away to distort my HUD. I hope this can be addressed as well.

That actually doesn't sound like an interface/hud issue so much as something relating to how the game tracks player locations for the minimap. I'd wager that's more related to ECM changes than these if it's true that they disappear when you aren't looking at them (which is the behavior for enemy mechs).

Tastian, on 15 May 2013 - 06:15 AM, said:

Great read. I'm a computer gaming programmer and I empathize with what you went through. Programming an integrated flash reader into our games, we also hit the case sensitivity issue with actionscript to c++. And, we dumped STL a long time ago because 1) its difficult to track and 2) its slow. And yes, we have bugs that takes us months to find - the worst offenders are generally lost pointers. I'm super excited you guys managed to find this. And I know you guys are as excited - hence the detailed write-up. Good job.




You aren't a programmer are you?

It's not an unreasonable request for a bug that is as old as the game itself. The ability to reinitialize the in game interface or reconnect to the game at all are both features the game still needs.

Edited by Shumabot, 15 May 2013 - 06:22 AM.

[Sprouticus]

Shumabot, on 14 May 2013 - 01:01 PM, said:

Yeah, it says that Karl was hired late into the process and that they seemingly didn't want to get serious about fixing the issues until fairly recently. These bugs are as old as this game, this didn't get overnighted by a crew of astronaut gold medalists, it looks like PGI couldn't get its **** together and hired some additional talent when they realized they were out of their depth.

No, it says he was BROUGHT ON to process towards the middle, not hired. HE was working on other stuff and brought in when they realized the low level code review that would be needed to accurately access the issue. (see other Dev posts pointing out that he has been on the team a long time)


It is a shame that it took this long to iron out the bug. And I am NOT saying I agree with some of the design and systemic decisions PGI has made. I have been quite vocal when I think they are doing something wrong.

But your posts make you sound like a petulant child who wants his toys NOWNOWNOW. The whole Veruca Salt act is getting old to many of us. Polite, reasonable criticism is a good thing, it ensures our voices are heard and the game has features and play styles the community likes. Acting like a spoiled brat and not actually reading what is written is useless and should get the scorn it deserves.

[Belorion]

Shumabot, on 15 May 2013 - 06:21 AM, said:

It's not an unreasonable request for a bug that is as old as the game itself. The ability to reinitialize the in game interface or reconnect to the game at all are both features the game still needs.

And how do you propose they would have worked around an intermittent bug that effected the UI as a whole?

[Shumabot]

Belorion, on 15 May 2013 - 06:35 AM, said:

And how do you propose they would have worked around an intermittent bug that effected the UI as a whole?

Because, as has been stated repeatedly, the bug applied itself at random. Players could simply initialize their hud until it didn't appear. The ability to restart the process of loading into the game environment would also help with a lot of other uncommon bugs.

Sprouticus, on 15 May 2013 - 06:33 AM, said:

No, it says he was BROUGHT ON to process towards the middle, not hired. HE was working on other stuff and brought in when they realized the low level code review that would be needed to accurately access the issue. (see other Dev posts pointing out that he has been on the team a long time)


It is a shame that it took this long to iron out the bug. And I am NOT saying I agree with some of the design and systemic decisions PGI has made. I have been quite vocal when I think they are doing something wrong.

But your posts make you sound like a petulant child who wants his toys NOWNOWNOW. The whole Veruca Salt act is getting old to many of us. Polite, reasonable criticism is a good thing, it ensures our voices are heard and the game has features and play styles the community likes. Acting like a spoiled brat and not actually reading what is written is useless and should get the scorn it deserves.

It's a bug that is literally as old as the game itself. How is it "now now now" to not be celebratory that it took into late stage open beta with publicly advertised sales (that don't mention its a beta) to fix a massive game breaking bug? I'm sorry, but as much of a victory as this is, this is a victory that came a half a year late at least. You sound like a sycophantic tool if I sound like a petulant child. Having a legitimate criticism for the extreme lag in fixing this bug doesn't make me wrong. I read the post. I'm not the one making up the fact that this bug is "new" and was "handled with expediency". It was clearly a low priority.

Edited by Shumabot, 15 May 2013 - 06:47 AM.

[ElLocoMarko]

Feedback... I played 20ish games last night. The map and enemy markers worked perfectly... no crashes, didn't see many disconnects of other players. Overall stability is very good.

One issue still out there (for me). Getting some texture flicker on the buildings in frozen city snow as described in this thread.
http://mwomercs.com/...s-bug-persists/
Looks just like this screenshot.
It is pretty consistent for me on that one map.

[Shumabot]

Thontor, on 15 May 2013 - 06:45 AM, said:

Karl was probably moved to this bug hunt when the frequency of it went through the roof a few patches ago. Yeah it's been around since closed beta, but something happened that made it happen a lot more frequently. I'm guessing that in PGI's efforts to optimize the HUD, they made it more susceptible to this bug.

In my experience it's no worse now than it was during the stalker patch. I suspect that after gold fractals they decided to actually prioritize bugs and bug testing. It seemed like a wakeup call to them when they accidentally made the game unplayable for a quarter of their population, since then they've seemed much more bug focused.

[NRP]

This was a great command chair post! It demonstrates why game development is hard. And it also demonstrates why most devs would prefer to write their own game engines, if only they had the resources to do so.

[Yankee77]

Hammerfinn, on 14 May 2013 - 11:19 AM, said:

Can we post this in every thread where people complain about laziness or incompetence at PGI?

And this is why I say an angel dies whenever someone claims game developers are lazy or something would be easy to implement/fix.

Many angels died due to these bugs, but it is good that their loss has been vindicated.

[scJazz]

Wow... excellent read! Notably all the false starts sections. To everyone who posted something along the lines of...

The HUD should be a painfully simple 2 second fix and PGI sucks because they can't fix it I want to say...

You should be forced to read that wall of text 1000 times before ever posting a "It should be easy to fix" post ever again!