776 réponses à ce sujet

[Bryan Johnson]

Blair is actually a producer

[Allan Schumacher]

Generic epic trailer music #542, I hoped for something more.....unique and original. 

 

Oh well.   

 

You can watch the other trailer for something unique, if you'd like.

[Barrett Rodych]

38frames per second? Pre order canceled!

Sometimes when you play a debug-mode build, you might get 1 frame per second!

[Barrett Rodych]

EP, title change to "That mage came from the Moon"?

Right?

[Barrett Rodych]

One can almost see the blood, sweat and tears of JEpler as he twitters each day.

I wonder if he tweeted back when he was in QA...

[Allan Schumacher]

Are the writers not at the BBQ? Also about the (newish) demo video that was loaded, I like how everyone in the crowd basically shut up when the dragon fight started.

 

The BBQ was the online team getting together.  Smaller team get together (of which I raided an extra hamburger, while talking about some Keep issues I noticed with Brian Mills )

[Allan Schumacher]

Just enough time to get attached to people before they die a terrible death!

 

Mua hah hah hah hah

[David Gaider]

 

You have been around a while, eh? It was that a kitten was punted through a plate glass window, actually. Think of the kittens!!

 

Nope. That was used for "every time a fan says something would be easy to do".

[Allan Schumacher]

To make myself feel better about Allan being whatever the canadian version of a brat is, I like to think that a good portion of the Marketing team and/or other Devs are keeping an eye on his twitter sweating it out over him accidentally revealing something. 

 

Good chance that I am sweating more than they are

[Allan Schumacher]

What would happen if you did reveal something?

 

They would probably ask me to be more careful and use more discretion with my interactions.  The horror!!! D:

[Allan Schumacher]

A stern lecture and a blood sacrifice to the writer's pit, probably. 

 

 

 *edit* my answer was better anyway.

 

My technological and biological distinctiveness would be merged into the collective.

[Allan Schumacher]

So, theoretically speaking, how much can you reveal without getting into trouble? 

 

No one's ever told me, which I take as a sign of good faith that they trust me to know the limits before I cross them.

[Allan Schumacher]

There was no need to warn him at all he was super cryptic about everything

 

I'm not too worried about it.  Aaryn favourited it some of the tweets himself!

[Barrett Rodych]

Twitter thread is where the real party's at.

[Barrett Rodych]

It's also a holiday today for Bioware Edmonton so I'm not expecting much...

Slackers. We're hard at work in Mtl! Though by hard at work I might mean giving DAI a whirl and maybe tweeting about it... 

[Allan Schumacher]

I know this, pretty sure she's working on ME4. It was just odd to start doing it now is all. 

 

Jessie can't tell me what she is working on, and I'm 99.99999999% sure she's not working on the new Mass Effect (otherwise she'd be able to tell me...).

 

 

Jessie is awesomesauce, but I wanted to make sure that people didn't take tweets she was making as reflective of stuff that was going on in DAI (or even BioWare).  By all means follower her, however, as she likes to share insights on game development on her Twitter.

[Allan Schumacher]

Bugger.....  I won't say I was wrong (Even though I am)..... I was just misinformed.......

 

No worries.  She's a dev and talks with us a lot and is pretty talented and shares art.  People often thought the same of _psdo as well

[Barrett Rodych]

Been there...and an even MORE awesome feeling is the knowledge that your work is done once you open a ticket.  SOMEONE ELSE gets to fix it!

QA work is never done, there is always another bug. Always. ALWAYS. Like tribbles, they are.

[JamesLeung]

Been there...and an even MORE awesome feeling is the knowledge that your work is done once you open a ticket.  SOMEONE ELSE gets to fix it!

It's also QA's responsibility to confirm the fix once the programmer or designer returns the bug to us as Resolved. If it isn't, we send it back.

[Allan Schumacher]

I think a programmer actually tweeted that one time too haha

[Allan Schumacher]

I confess, a marsh is not a feature I was looking for. Can't the just demo Allan hunting Fennecs?

 

This is giving me fun ideas for prerelease.... XD (assuming I'd be allowed lol)

[Allan Schumacher]

I hope that doesn't mean people will look at my archer-spec'd rogue and call her a dirty thief. That would irritate me more than a little bit seeing as you don't have much choice in the matter if you want to shoot pointy things and aren't necessarily pickpocketing everyone you come across.

 

The context of mine was referring to a choice *I* could make in response

[Barrett Rodych]

People who find reliable repro steps before filing bugs are good people. Nay, great people. The engineer's job is much, much easier when repro steps exist.

1. The engineer can quickly determine if the bug still exists by the time they work on the ticket (bugs always come in faster than they can be fixed, so for lower priority bugs there's a time gap between "bug found" and "engineer begins looking at bug").

2. The engineer knows the code path that causes the bug, or can quickly find it, by working backwards from the actions in the repro steps. Even small production codebases contain more lines of text than even the longest novel. Being able to narrow down the location of the problem is vital.

3. Having narrowed down the problematic code, the engineer can try different things until the bug goes away. They'll know exactly when the bug goes away because they'll try the repro steps and the bug won't appear.

My favourites are rare frequency race condition/timing-specific crashes with garbage release mode callstacks which can't be reproduced in a debug mode environment.

[Barrett Rodych]

Better yet when you print trace data to the console so you can track the code state in release mode while hunting down that bug, and that causes the bug to go away, only to return when you remove those trace calls. /FML

Quantum bugs are so next gen.

[Allan Schumacher]

My favourites are rare frequency race condition/timing-specific crashes with garbage release mode callstacks which can't be reproduced in a debug mode environment.