I think it is time, to finally consider stability

it is a complex issue, but keep in mind fixing technical issues in 4 Acts, is harder than in 3.

by now, GGG should have both a nice financial cushion, and a stable enough base of PR and followers, to allow taking on something, which in my opinion is a lot more important than more content.
they will obviously have a deadline for this, thus can't take a year off content and MTX to do only tweaking - but my point is it's unlikely they even tried starting a high-priority code quality task, as a team, until now.
the highest-impact quality task I remember in 2 years, is just hiring a Guru and throwing most of the work at him.
that's hardly a viable way to achieve results.

P.S: I like the new supporter pack idea. although it implies being at a low point, and needing to blackmail funds, to actually do what most developers including myself, consider an absolute must.

"

Boem wrote:

GGG could in fact release a new support pack that aims to get $$$ from supporters just for this purpose alone "optimization supporter pack" sort of speak, from which the funds go to a pool of resources dedicated to just this issue.

Yes, I've suggested something like that too. The 'problem' is that in this case they would have an obligation to actually do something, instead of sweeping it under the rug :) (just look at 'rain' and how GGG is ignoring this issue)

Imo the situation is pretty simple: do they really plan the longevity of this game for another 5 - 7 years? If yes, they will have to fix the engine at some point.

"

morbo wrote:

Imo the situation is pretty simple: do they really plan the longevity of this game for another 5 - 7 years? If yes, they will have to fix the engine at some point.

which is where I come in, and say that if they plan to keep the level of optimization pretty much as is, I doubt even the greatest programmer in mankind's coding history himself, may be able to fix their code at that point, ~5 years in the future.
and that's if the game's still functioning then. a huge "if".

thing about serious code problems is, they need to be taken care of as soon as possible, as efficiently as possible in terms of impact on the end user, and preferably with as much team work as possible, instead of sitting on your damned hands waiting for some genius to barge in and single-handedly sort out the mess you made.

"

johnKeys wrote:

"

morbo wrote:

Imo the situation is pretty simple: do they really plan the longevity of this game for another 5 - 7 years? If yes, they will have to fix the engine at some point.

which is where I come in, and say that if they plan to keep the level of optimization pretty much as is, I doubt even the greatest programmer in mankind's coding history himself, may be able to fix their code at that point, ~5 years in the future.
and that's if the game's still functioning then. a huge "if".

thing about serious code problems is, they need to be taken care of as soon as possible, as efficiently as possible in terms of impact on the end user, and preferably with as much team work as possible, instead of sitting on your damned hands waiting for some genius to barge in and single-handedly sort out the mess you made.

GGG should totally hire you, you've got it all figured out.

Trolling as usual, but you understood exactly what I meant.
Small problems become bigger and tougher over time.
And here are the really bad news - optimization in PoE never really was a small problem.

I believe you are the one trolling, as usual. The world does not begin or end with johnkeys. That potato? You could upgrade it? A thought. Or wait for outwards things outside of your control to change, it's on you ol' boy.

"

geradon wrote:

please, win7x64 was released 7 years ago, hardware supporting 64 bit was available centuries ago.

computers running a 32 bit os which only supports 4 GByte ram maybe sufficient for writing letters with word or playing tetris but on most non optimized machines win7 itself with it's caching mechanisms and the shitload of virus scanners, snakeoil personal firewalls and whatnot started alongside the os is taking up the max available ram already.
and that's not the only problem, missing ram can be virtualized but on 32 bit operating systems this is restricted to ... 32 bit as well.

That's all nice, but you aren't saying anything of value.

The game's min spec allows for 32-bit systems. There are a lot of 32-bit machines out there and GGG would like to make money off their owners. To do that, they need to make sure their game is stable enough on such systems that users don't have to go digging around in threads for workarounds like shutting off the sound or using a non-Steam executable for a game available on Steam just to be able to reliably cross zones in the end of act 3.

That's all that's relevant. That 32-bit systems are on the way out (even though companies are for some reason still releasing 32-bit variants of new OSes) doesn't matter. There's still enough out there that GGG supports them.

Nor is it a case where the game needs to have 64-bit. Besides it not being a requirement, the game is not some cutting-edge piece of tech, pushing the envelope. It looks and plays like something done on a 32-bit systems. And prior to 1.3, while it had some memory issues, it did just that. And with 1.3, if you use certain work-arounds like shutting off the sound or not using the Steam client, it still runs fine. But in 1.3, they changed something that made the game stability much, much worse.

It is a bug they introduced (or rather made a much larger issue) and they need to fix it. Furthermore, if they don't fix it now, as they continue to tack on things, the bloat will only continue. And next time, there won't be sound to turn off to solve the problem, because the sound is already off.

Currently, the state of the game for 32-bit users is embarrassingly bad. If GGG were required to use a 32-bit system with the Steam client (things they claim their game supports) for all demonstrations by and for their producers, the constant crashing would quickly find optimizing things placed at the top of the priority list.

Now, if they start making PoE 2 and decide to go 64-bit only, that's fine. But PoE is a game that supports 32-bit systems. Therefore, it should run on them reliably. Period.

"

TheWretch wrote:

I believe you are the one trolling, as usual. The world does not begin or end with johnkeys. That potato? You could upgrade it? A thought. Or wait for outwards things outside of your control to change, it's on you ol' boy.

you either are completely not getting this, or deliberately trying to get on my nerves.
if it's the latter - it's working.

I'm going to try one more time.
one last time before you're going to get ignored.
so read carefully.

will I buy a new PC just to handle current-state Path Of Exile? maybe.
will that solve the problem I'm presenting? no. it will only make the game more playable for me, temporarily. until the next big "milestone" in stacking poor optimizations on top one another.

and this isn't about me.

so, will someone who isn't as addicted to this game as I am, doesn't care about it as I do, and/or lacks the financial capability I do - be willing to participate in a "hardware arms race" with GGG's bloated code, which spirals out of control with virtually every content patch?
constantly upgrade their PC to play one game?

hell no.

so the only remaining question is, will the rate of players leaving/stopping to buy MTX because the game becomes unplayable for them, equal the rate of players doing so, because GGG promised Act 4 in - say - January, but will have to move that to - say - March, in order to do a series of major patches competely dedicated to stability and code fixing?
I think it can. I even think the former is greater.
so as someone who wants to be here playing in 1-2 years time, and wants GGG to still be here maintaining this game - I think it is about fucking time they do something with their code.

you on the other hand - as Axterix13 cleverly put it with the help of Monty Python - are just here being a Black Knight. one who doesn't give a fuck about the game as long as none passes your damned post.

I hope now I made myself clear.
now feel free to ask serious questions.

The solutions with the most impact on performance I see so far is in the way the game handles lights, particle effects, and the rain in docks. After that would be more predictable AI for the Desync monsters(snakes). Next would be a collision forgiveness program like Allstate. Set an area around a player location so that when a movement skill is used if the ending position that the client says is achieved is reachable within that error zone then have the server accept that position as valid. A check like that which only happens on movement skills isn't too bad for server load. I mean if you have to rubberband the same person 4 times in a second and this happens to .01% of the population of a server per second you are going to save server load by implementing that last one.

TL:DR and more importantly dont care. Do whatever you want John, ive told ya that multiple times. Your semi-weekly threads that moan over performance issues really helps..They got old awhile ago.