|
|
Post by colin22cans on Aug 11, 2015 16:01:14 GMT
|
|
Lord Ba'al
Supreme Deity
Posts: 6,260
Pledge level: Half a Partner
I like: Cats; single malt Scotch; Stargate; Amiga; fried potatoes; retro gaming; cheese; snickers; sticky tape.
I don't like: Dimples in the bottom of scotch bottles; Facebook games masquerading as godgames.
Steam: stonelesscutter
GOG: stonelesscutter
|
Post by Lord Ba'al on Aug 11, 2015 17:34:16 GMT
Yay. An update. And it isn't even Friday. Strange things are afoot.
So what topics will Konrad be adressing in the next few weeks?
|
|
|
|
Post by mindless on Aug 12, 2015 15:22:22 GMT
Someone over on the steam forums pondered a couple of questions which i'll re-iterate here.
a) Would having a larger team working on the game reduce the amount of time it takes to identify and fix these bugs?
b) also is the fact that these bugs have been growing in magnitude to the point that the game is now unplayable a result of taking experience coders away from the game?
|
|
|
|
Post by 13thGeneral on Aug 12, 2015 15:31:17 GMT
It's really unfortunate that they're working through stability issues that should/could have been addressed back in the Aplha and Beta phases of development - when the core base of the game was unstable, and they didn't seem to care. Here we are, two years into Early Access, and it's still a hulking mess of a game.
Glad to see that they are trying to figure it out though.
|
|
|
|
Post by colin22cans on Aug 12, 2015 15:40:06 GMT
Someone over on the steam forums pondered a couple of questions which i'll re-iterate here. a) Would having a larger team working on the game reduce the amount of time it takes to identify and fix these bugs? b) also is the fact that these bugs have been growing in magnitude to the point that the game is now unplayable a result of taking experience coders away from the game? Hi Mindless a) Not really, faster processors and faster RAM would help more than larger team. Dimitri & Gary have been helping out with the issue as well due to their knowledge and experience with the codebase etc. b) It's more that quick fixes and patches we're been put in place without going back and spending significant time fixing everything. Experienced coders are working on Godus  |
|
|
|
Post by colin22cans on Aug 12, 2015 15:41:51 GMT
Yay. An update. And it isn't even Friday. Strange things are afoot. Wasn't the last update on a Thursday? |
|
|
|
Post by morsealworth on Aug 12, 2015 17:05:41 GMT
Is it just me or Konrad's way of speech is suspiciously unlike him?
Or was that someone else rearranging what Konrad said?
|
|
Lord Ba'al
Supreme Deity
Posts: 6,260
Pledge level: Half a Partner
I like: Cats; single malt Scotch; Stargate; Amiga; fried potatoes; retro gaming; cheese; snickers; sticky tape.
I don't like: Dimples in the bottom of scotch bottles; Facebook games masquerading as godgames.
Steam: stonelesscutter
GOG: stonelesscutter
|
Post by Lord Ba'al on Aug 12, 2015 17:15:37 GMT
Yay. An update. And it isn't even Friday. Strange things are afoot. Wasn't the last update on a Thursday? Could be. I just remember they have often been on a Friday. Not that it matters. |
|
|
|
Post by muumipeikko on Aug 12, 2015 23:29:04 GMT
Wow, this speaks volumes to me as a C++ developer and manager. As a developer I know the temptation is to do the profiling later as lets face it it's not exciting work Vs greenfield dev. As a manager I know the importance of doing it now as these errors tend to multiply and if you let is go on the only decision you have left to make is whether to rewrite the code or scrap the product... I'm guessing poor Konrad and Richard are not getting much oversight from their management...
|
|
|
|
Post by Gmr Leon on Aug 12, 2015 23:32:34 GMT
Management never was 22cans' forte. I'm pretty sure even with Simon on board, there remains this fast and loose atmosphere of work on...something...related to this...greater concept. We'll hammer out the details along the way.
|
|
|
|
Post by colin22cans on Aug 13, 2015 9:16:34 GMT
I'm guessing poor Konrad and Richard are not getting much oversight from their management... They are - Which is why they're having to fix it. |
|
|
|
Post by Crumpy Six on Aug 13, 2015 13:25:14 GMT
Well... it seems like they are having to fix it now that it has literally got to the point where they can't even launch the game without it breaking. I'm not sure that qualifies as having much oversight.
|
|
|
|
Post by muumipeikko on Aug 13, 2015 17:44:53 GMT
I'm guessing poor Konrad and Richard are not getting much oversight from their management... They are - Which is why they're having to fix it. Sounds like the horse has well and truly bolted. I've been in situation where we suddenly see a massive drop in performance or memory ramping up, the difference is though, a well run team spots it early and know is must be something done in the last few days and rolling back then incrementing forward is an option to find the issue quick... More than a few week or two and you're going to have to find it the hard way... Good luck!
|
|
|
|
Post by colin22cans on Aug 14, 2015 11:04:19 GMT
I spoke to Richard about this a bit more seeing that everyone seems to have jumped to their own ideas about what's happening.
RW - "One of the aspects which makes tracking down these sorts of memory issues difficult is that (perhaps unintuitively) the issues aren't necessarily in new code. It's unfortunately not as simple as saying "Okay, we didn't crash on yesterday, but we are crashing today. So it must be an issue with something we wrote in the last 24 hours!". For example, the current iteration of the AI sculpts quite differently to players - it makes lots of small, frequent sculpts and represents using the sculpting code in a different way to how it's been used before. This can mean that a memory issue in the core sculpting code which took 36 hours of continuous play to show itself previously could become problematic in as little as 10 minutes. Even worse, if the memory's getting corrupted, it could be happening from an entirely different area of the codebase!"
|
|
|
|
Post by muumipeikko on Aug 16, 2015 9:18:05 GMT
For the past 20 years I've been writing C++ systems so big in terms of lines of code and complexity they make godus look like a little hello world app. What you saying is correct in that the problem may not lie in the new code but if you profile regularly you have a base line which you can easily roll back to where the issue doesn't occur. Then you roll your changes forward until you hit the issue then you look at the changes and go "Oh, I see Konrad is calling PullOverEyes() with the args wool and community and that's causing the issue lets look at that in detail".
Every night we run a suite of tests on the code base to check memory, speed and CPU utilization and every weekend we run extensive tests and have never got into the situation where we have had to say we need to shutdown the trading for a few weeks while we try to resuscitate our code base..
Like I said good luck, I know how painful these times will be...
|
|
Lord Ba'al
Supreme Deity
Posts: 6,260
Pledge level: Half a Partner
I like: Cats; single malt Scotch; Stargate; Amiga; fried potatoes; retro gaming; cheese; snickers; sticky tape.
I don't like: Dimples in the bottom of scotch bottles; Facebook games masquerading as godgames.
Steam: stonelesscutter
GOG: stonelesscutter
|
Post by Lord Ba'al on Aug 16, 2015 10:20:12 GMT
Heheheheheh!
|
|
