Torchlight III

Torchlight III

View Stats:
Jul 2 @ 11:56am
State of the Game: Week 3
Torchlight 3 Launch Weekend Postmortem - "We had to kill all of the zombies."

View full event information here:
https://steamcommunity.com/ogg/1030210/announcements/detail/2545045898938703643
< >
Showing 16-30 of 35 comments
Asenaz Jul 3 @ 3:43am 
I have a statue in my inventory that i already know and i cant sell it or destroy it, so i have one slot less in my inventory. Also micro-lags are a problem cause adds hit u and u cant even notice it and it is not problem of my connection. Pls fix this, ty for all the work u are doing !!
You know how you wouldn't have server issues? IF THE GAME WAS NOT DAMNED TO BEING TIED TO ALWAYS ONLINE
你們寧願打這一長串話,也不願快點修BUG。
Mephizo Jul 3 @ 7:12am 
I don't understand why people are being so hostile, it is obvious your team is working to squish the bugs. I am monitoring progress and patiently waiting for the right time to snag this game.:steamhappy:
Originally posted by Mephizo:
I don't understand why people are being so hostile, it is obvious your team is working to squish the bugs. I am monitoring progress and patiently waiting for the right time to snag this game.:steamhappy:

It's a combination of people being ignorant and people being wary because of past early access failures.

People are wary because there have been other early access games that have turned out a mess even after years of development, but I think a lot of people are just ignorant about what game development actually entails.

I also think a lot of people simply forget that there's human beings on the other end of these forums at Echtra (or whatever game company is making the game they're bashing) reading what they're writing, and we as a species have developed this sort of collective inability to stop and empathize before we say/write something.
Last edited by IndigoAK; Jul 3 @ 7:18am
sweater Jul 3 @ 10:36am 
Originally posted by fabulousfreep:
If this game is going to be an always online game like D3 i'm just going to go ahead and put you on ignore. Sad but never making that mistake again.

Jesus christ.

*facepalm goes though head*
Originally posted by IndigoAK:
Originally posted by Mephizo:
I don't understand why people are being so hostile, it is obvious your team is working to squish the bugs. I am monitoring progress and patiently waiting for the right time to snag this game.:steamhappy:

It's a combination of people being ignorant and people being wary because of past early access failures.

People are wary because there have been other early access games that have turned out a mess even after years of development, but I think a lot of people are just ignorant about what game development actually entails.

I also think a lot of people simply forget that there's human beings on the other end of these forums at Echtra (or whatever game company is making the game they're bashing) reading what they're writing, and we as a species have developed this sort of collective inability to stop and empathize before we say/write something.

You would think those people would just avoid Early Access titles but here they are, every time, buying them and then ♥♥♥♥♥♥♥♥ about Early Access problems while not actually reading anything about the plans for the game.
The14th Jul 3 @ 5:08pm 
Originally posted by Humanitarian:
Guy Somberg. The name sounded unique and familiar somehow. Then I remembered. He was an audio programmer in Flagship I think.

But more importantly, he is the guy that snitched on Flagship and the HG:L team. Then he was negated by David Brevik himself on public and fired.

GameIndustry: "He says that poor sales of the EA published game, and negative consumer feedback has left the studio with low morale and a dwindling talent pool."

Somberg: "The reason is that people are leaving. In droves, they’re leaving. We've had programmers, accountants, HR people, and artists leave," wrote Somberg on his personal blog."

Brevik: "Everyone has good days and bad days. Guy was having a bad day. ... He vented, exaggerated, and posted information that was inaccurate,"

Here's the source:
https://www.gamesindustry.biz/articles/flagship-programmer-says-staff-are-leaving-in-droves
https://www.gamesradar.com/flagship-were-not-sinking/

TL3 devs should be careful because things don't look good and they have a snitch inside.

lol :D

It looks this game was doomed from the start.
I don't trust much of what David Brevik says from back then, 'cus it was under his management that they made Hellgate: London. And yeah, of course Somberg is still working on this game. Runic Games was originally what was left of the Flagship Seattle studio, and many of the Runic team reformed to work on Torchlight 3.

Really though, why are you digging for a villain here?
Well, that was a FANTASTIC description... of your incompetence on release.

And the game itself... well its almost like this is not even the same team that made TL1 and 2, because the map generation - you know that ACTUAL LEVELS we play on, is terrible.

What is going on with that? Its all shiney, but there is nothing organic at all to it, the repeating parts and graphics just make it actually inferior to TL2, D2, D3, and even the terrible generated maps from POE feel more interesting.
Pid Jul 4 @ 2:55am 
Another dev here, struggling daily with refactoring monoliths to microservices on a PaaS with Hor/Ver scaling with containers and pods. This write-up was very entertaining and... I have a rubber duck sitting right under my monitor :) So... thanks for opening up that way. Rest assured there are kids not understanding the struggle is real, just wanting to play. But there are also other folks understanding very well what is going on.

So now I envite everyone to re-evaluate their first impression and giving this game a second look. Albeit, it's still in EA, so... let's be constructive and make it great!
Valics Jul 4 @ 3:40am 
@me when the game is released
sad that people who gives a negative review on STEAM EA games will not read this article to understand how complex and time consuming is develop a game like this... and why EA is meant to be a mess buggy and fixes time... :stoneheart:
B166ER Jul 4 @ 9:15am 
You never actually tested the Zone Controller during a large-scale stress-testing phase, or else those 410 HTTP errors would've presented themselves on a silver platter, right?

> This was where we started to realize that if the upstream services like Zone Lookup and the content service were lagging, then they were causing blocks in the ZC.

Sounds like a pretty obvious test case you never bothered to test.
"What happens if ZC as the single-point-of-failure doesn't get data fast enough due to other servers or processes being slow or lagging".

> So we started asking questions about how the ZC works and how it spins up and monitors zones. As a part of this, one of our engineers noticed that we are sending events on a Go channel while holding a lock on the map.

You're asking yourselves on a live environment what your own code actually does?
Impressive, in the most negative sense imaginable.

> However, the buffer length is finite and eventually the buffer gets filled up as the other routines try to hit up the Zone Lookup.

Yet another obvious test case. "What happens when the buffer is full considering the amount of data we do buffer"

> The lesson here for us is that the first impression of what a problem’s root cause is is unlikely to be the actual problem, so we need to constantly be re-evaluating our fundamental assumptions.

The first lesson is "Extensively test your software based on priority/severity".
Nkov Jul 4 @ 12:17pm 
Originally posted by B166ER:
You never actually tested the Zone Controller during a large-scale stress-testing phase, or else those 410 HTTP errors would've presented themselves on a silver platter, right?

> This was where we started to realize that if the upstream services like Zone Lookup and the content service were lagging, then they were causing blocks in the ZC.

Sounds like a pretty obvious test case you never bothered to test.
"What happens if ZC as the single-point-of-failure doesn't get data fast enough due to other servers or processes being slow or lagging".

> So we started asking questions about how the ZC works and how it spins up and monitors zones. As a part of this, one of our engineers noticed that we are sending events on a Go channel while holding a lock on the map.

You're asking yourselves on a live environment what your own code actually does?
Impressive, in the most negative sense imaginable.

> However, the buffer length is finite and eventually the buffer gets filled up as the other routines try to hit up the Zone Lookup.

Yet another obvious test case. "What happens when the buffer is full considering the amount of data we do buffer"

> The lesson here for us is that the first impression of what a problem’s root cause is is unlikely to be the actual problem, so we need to constantly be re-evaluating our fundamental assumptions.

The first lesson is "Extensively test your software based on priority/severity".

♥♥♥♥♥♥♥♥ RIGHT!
rofl
Last edited by nowhere man; Jul 5 @ 11:51am
< >
Showing 16-30 of 35 comments
Per page: 15 30 50