292 replies to this topic

[Shadow Beast]

Karl may I just start by saying thank you for you post on the process that you had to go through to find the issue with the HUD elements. Having written programs in the past I appreciate the efforts that you and the PGI team have been putting in to resolve the issues that people have been reporting. Here's hoping that there are not more "features" like this in the CryEngine or Scaleform.

[MustrumRidcully]

Ramien, on 14 May 2013 - 11:22 AM, said:

I'm not sure if Karl's post made me reconsider my chosen field of study (Computer science and software development) or makes me want to do it even more. Either way, an excellent read!

Just recently we talked in our company about employee education and learning opportunities. I seriously consider trying to find an opportunity to learn more of the inner workings of C++.. I came from the Java world to the C# world and then to the C++ world, but in the C++ world, I rely on Qt - so much is abstracted away from me that I sometimes just am glad to know where to write the consts, &, * so that the compiler doesn't complain.

[syngyne]

Falcore, on 14 May 2013 - 11:50 AM, said:

First of all thanks for the ultra detailed response. I understood maybe half of that but the transparency was extremely helpful. I hope that you are this transparent with bugs and issues going forward. I think there would have been less complaints if you had been more transparent earlier in the process.

To be fair, I don't know how useful it would have been to post "Okay, we think we found the root cause... nope, trying something else" over and over until the resolution. If anything it would have probably just gotten people's hopes up and then frustrated them. A post mortem makes more sense.

[Butane9000]

A good read. Now if only PGI can make regular statements about issues plaguing the game in such fine form.

[MustrumRidcully]

Butane9000, on 14 May 2013 - 12:05 PM, said:

A good read. Now if only PGI can make regular statements about issues plaguing the game in such fine form.

A balance command chair post where they discuss discrepancies between gameplay realities and their balance model, or something like that.

[Jonathan Paine]

Reminds me of the early morning when I decided to give up on programming. Had finally figured out why my "bottle sorting program" (the excitement never ends!) failed to sort one specific type of bottle, fixed it, trimmed the code by a third... then realized that the last bus had left hours ago and I was stuck in the computer lab for another four hours until the busses started running again. Realized that I'd rather do other things at four in the morning

Kudos to you guys for sticking it out

[nuked4ever]

As an advanced processor design verification engineer, I understand and appreciate the effort that goes into debugging corner cases. Some bugs can escape, and it is sure as hell fun trying to reproduce the ones that do. Good job on this!

[Utilyan]

ahhh hah.......its quite clear the devs went INTO the matrix....and killed a librarian named Cryname.....

[Karl Berg]

Phaesphoros, on 14 May 2013 - 11:49 AM, said:

Would you mind replacing "STL" with "Standard Library"? Stop spreading the myth xD Great work finding that bug.

Heh, interesting read, thanks!

[Kitane]

Reading that debugging process just made me sick . Cases like this often end up on my table and its not doing anything good to my blood pressure.

Good job, really.

[Verbrand]

I'm sending this to a contact I have who works for Crytek, see what he thinks of all this...

Great job Karl and co!

[MavRCK]

*bow* what an analysis of the errors and an amazing feat of hardwork and dedication. Great read, too, for us non-programmers.

Edited by MavRCK, 14 May 2013 - 12:29 PM.

[RenegadeMaster]

syngyne, on 14 May 2013 - 12:04 PM, said:

To be fair, I don't know how useful it would have been to post "Okay, we think we found the root cause... nope, trying something else" over and over until the resolution. If anything it would have probably just gotten people's hopes up and then frustrated them. A post mortem makes more sense.

I wouldn't have expected a public post mortem that was this detailed, but as a QA Analyst for a software company, I too really appreciate the post mortem as well as the immense effort put into these insanely tough to track down bugs.
This makes the complexity of cloud computing platforms (like the ones I test) look like a cakewalk compared to the gaming platform/engine that MWO is being developed on.

[Stoicblitzer]

Interesting read. I commend the members of PGI involved in fixing this bug. This demonstrates to me that although they are a small firm and things take awhile to fix, some of them really do know what they are doing. This game might actually survive.

[stjobe]

Jonathan Paine, on 14 May 2013 - 12:11 PM, said:

Reminds me of the early morning when I decided to give up on programming. Had finally figured out why my "bottle sorting program" (the excitement never ends!) failed to sort one specific type of bottle, fixed it, trimmed the code by a third... then realized that the last bus had left hours ago and I was stuck in the computer lab for another four hours until the busses started running again.

Excellent, four more hours to really make the program perfect! Perhaps tidy up that ugly method, rename the variables to something less cryptic, and perhaps even comment the code!

I've had so many times when I've "just going to have a short look" at something, and re-surfacing five or eight hours later, in the middle of the night. It's beautiful when you're in the zone

[Coolant]

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?

Can we post the huge announcement of a pic of Canyon that we had already seen before, or the fact that we still don't have a lobby, still can't drop with the players we want, can't choose a map, and only have 6 maps total, and still can't paint our Founder's mechs that we dropped $30-$120 on?

[Zyllos]

That problem right there is why threaded applications are so difficult to mess with. One unrelated issue causing issues somewhere else which manifests itself as something unrelated.

Got to love programming...

Back to PLM/386!

Edited by Zyllos, 14 May 2013 - 12:40 PM.

[Adria Amoro]

Wintersdark, on 14 May 2013 - 11:25 AM, said:

I think my brain exploded.

Yea mine did too. I did not understand much out of that post except that it was bloody hard to find and fix. Good job to the guys that fixed it.

[zmeul]

Verbrand, on 14 May 2013 - 12:28 PM, said:

I'm sending this to a contact I have who works for Crytek, see what he thinks of all this...

Great job Karl and co!

I'm quite interested on what he has to say

[Hammerfinn]

Coolant, on 14 May 2013 - 12:32 PM, said:

Can we post the huge announcement of a pic of Canyon that we had already seen before, or the fact that we still don't have a lobby, still can't drop with the players we want, can't choose a map, and only have 6 maps total, and still can't paint our Founder's mechs that we dropped $30-$120 on?

Did you even read this? http://mwomercs.com/...23#entry2348823