Optimizations for lower-end?
Moderator: Moderators
- clericvash
- Posts: 1394
- Joined: 05 Oct 2004, 01:05
Optimizations for lower-end?
This is a question to any devs, i am wondering if you plan to include any options/optimizations for lower-end computers?
what if on sm2 maps each game tile was averaged out to a single colour then 1 single texture of each pixel representing a game square be used rather than lots of little tiles for single ground quads, it'd save memory and speed up ground drawing on slower PCs, especially those with integrated graphics.
- clericvash
- Posts: 1394
- Joined: 05 Oct 2004, 01:05
- FoeOfTheBee
- Posts: 557
- Joined: 12 May 2005, 18:26
There was this a while back:
http://spring.clan-sy.com/phpbb/viewtop ... g&start=23
Robed got it working I think, but it wasn't added. Don't know the details or the merits of his code, but the idea seemed promising.
http://spring.clan-sy.com/phpbb/viewtop ... g&start=23
Robed got it working I think, but it wasn't added. Don't know the details or the merits of his code, but the idea seemed promising.
- FoeOfTheBee
- Posts: 557
- Joined: 12 May 2005, 18:26
What I hope we could tell low enders is that the development community is committed to making/keeping Spring accessible to the vast majority of human beings in the world who can't afford a new computer.smoth wrote:lol, I feel sorry for the low enders.. spring is going to get more and more graphically intense. I do not know what to tell you.
Spring is already pretty good at running on older hardware. Making and keeping graphical/cosmetic features optional to include even older machines moves in the direction of Pareto optimality.
It is not a matter of things like particle systems. Those can be disabled sure.. but high detail textures, maps and models are going to be more and more prevalent in the future. Combined with mods and maps runing lua for gai player functions and the VAST amount of units and complicated things happening spring is going to become pretty hefty.
There has to be a limit to how much people are going to allow others to retard their progress. Things like reflective units/transparencies are already beyond the reach of some lower-end machines. I am not saying they have to buy a new machine but even my 6800 gets slow down in spring on ocassion. When you have a game that supports 8+ players... if someone is BARELY running spring then they are going to run into more troubles when the desync is fixed because more players will be in the games. Those extra players are going to be running all of that extra stuff that is without a way to disable. This such as very complex scripts and lua code.
There has to be a limit to how much people are going to allow others to retard their progress. Things like reflective units/transparencies are already beyond the reach of some lower-end machines. I am not saying they have to buy a new machine but even my 6800 gets slow down in spring on ocassion. When you have a game that supports 8+ players... if someone is BARELY running spring then they are going to run into more troubles when the desync is fixed because more players will be in the games. Those extra players are going to be running all of that extra stuff that is without a way to disable. This such as very complex scripts and lua code.
I'm going to keep saying it... if Spring does not end up with mods that are beautiful but require a decent (not insane, merely decent) computer, then Spring's audience will never grow. We've been static on numbers for months- the "roadmap" expected large growth, but for whatever reasons, we haven't really grown a larger audience yet.
I'm working on my last major project for this engine now... hopefully that will change that (we're planning a PR campaign, assuming the desync thing is conquered by release date) and bring in a lot more people.
I'm working on my last major project for this engine now... hopefully that will change that (we're planning a PR campaign, assuming the desync thing is conquered by release date) and bring in a lot more people.
I'm way ahead of you, Argh, so don't sweat it.Argh wrote:I'm going to keep saying it... if Spring does not end up with mods that are beautiful but require a decent (not insane, merely decent) computer, then Spring's audience will never grow. We've been static on numbers for months- the "roadmap" expected large growth, but for whatever reasons, we haven't really grown a larger audience yet.
I'm working on my last major project for this engine now... hopefully that will change that (we're planning a PR campaign, assuming the desync thing is conquered by release date) and bring in a lot more people.
A guy asked if I knew any free games so I pointed him to spring. He'd not really played RTS' much but tasclients GUI confused him and I had to give him step by step instructions click this that and then repeat them and re-explain them.
Once we got ingame he was faced with even more issues. First he couldnt see the text in the tooltips so he couldnt tell what he was going to build, that was after 5 mintues of him saying "there are no buildpics, just buttons saying things like fight or attack, how do I exit? =s shift+esc doesnt exit, wtf, wheres the exit button, I cant tell what I'm building?!?!?!"
Despite all the lua gadgetry nobody has yet to even attempt to sort out orders/build pics tabs, OTA style, or even copy another RTS, nobodies tried implementing proper tooltips even the failed 'New GUI' had tooltips, and nobodies even bothered setting the buildpics to appear before the commands by default.
And look at our GUI dvelopment process, its orientated around seperate widgets that on their own look 'cool' or do something 'cool', and even then HCI design doesnt seem to have a say, for example the tweak mode, "But you cant move things around, there's no buttons or controls to do it with, no you cant do it see no buttons nothing in the menus nothing"*in comes a dev with a magic key combo* "wtf made you hide it like that"
Or silly design choices, for example to get to spring settings have to start tasclient, open the options window, goto the settings tab, and press game settings. How is that user friendly? Why not just put a spring settings button at the top that runs settings.exe rather than hiding everything away? Nobodies making it all apparent and eays to use.
And what about are widget picker? fine you've got the eyes on startup that tell the user to press f11, but where is that comment? The description bar? Why would the user think to look there? The same for descriptions of widgets, shouldnt they be by the widgets rather than in some arbitrary screen location? And why f11? Why cant there be a button to bring up the widget picker panel?
And why to host a battle do you have to enter the battlescreen, then click host battle? If your a new user its totally confusing as you might not have clicked battlescreen before..
Spring seems designed around emphasising the importance of things 'functioning' s much that what they made is out of reach of a large number of people who simply give up after a few days because its all too much hassle, hassle caused in a lot of cases by bad design decisions that are easily avoidable, easily fixable, and should never have been made to begin with..
Once we got ingame he was faced with even more issues. First he couldnt see the text in the tooltips so he couldnt tell what he was going to build, that was after 5 mintues of him saying "there are no buildpics, just buttons saying things like fight or attack, how do I exit? =s shift+esc doesnt exit, wtf, wheres the exit button, I cant tell what I'm building?!?!?!"
Despite all the lua gadgetry nobody has yet to even attempt to sort out orders/build pics tabs, OTA style, or even copy another RTS, nobodies tried implementing proper tooltips even the failed 'New GUI' had tooltips, and nobodies even bothered setting the buildpics to appear before the commands by default.
And look at our GUI dvelopment process, its orientated around seperate widgets that on their own look 'cool' or do something 'cool', and even then HCI design doesnt seem to have a say, for example the tweak mode, "But you cant move things around, there's no buttons or controls to do it with, no you cant do it see no buttons nothing in the menus nothing"*in comes a dev with a magic key combo* "wtf made you hide it like that"
Or silly design choices, for example to get to spring settings have to start tasclient, open the options window, goto the settings tab, and press game settings. How is that user friendly? Why not just put a spring settings button at the top that runs settings.exe rather than hiding everything away? Nobodies making it all apparent and eays to use.
And what about are widget picker? fine you've got the eyes on startup that tell the user to press f11, but where is that comment? The description bar? Why would the user think to look there? The same for descriptions of widgets, shouldnt they be by the widgets rather than in some arbitrary screen location? And why f11? Why cant there be a button to bring up the widget picker panel?
And why to host a battle do you have to enter the battlescreen, then click host battle? If your a new user its totally confusing as you might not have clicked battlescreen before..
Spring seems designed around emphasising the importance of things 'functioning' s much that what they made is out of reach of a large number of people who simply give up after a few days because its all too much hassle, hassle caused in a lot of cases by bad design decisions that are easily avoidable, easily fixable, and should never have been made to begin with..
its the way the world is, the easier and faster the gui is, the more incompetent people becomeMadRat wrote:C'mon, spring isn't that bad of a kludge. Some people are just gui clueless.
now, seriously, springs current guy is ok, not great, but you dont need to read a manual to use it (just some practice), its just the world that needs a start menu, a talking clip and a flash tutorial no know their way around
it wasnt so long ago when a game required you to read a 300 page manual, people forget too easily (those games usually were great, too)
About lower end clients, maybe it it not just about optimizations, but maybe it could be possible to automatically disable features to make it run faster. This in case this is the cause for slowness. If it is the simulation code then... other solution could be found.
About the gui, it all depends on what type of players do we want. There are many that manage the gui ... We want to have a greater community so that more people contribute and my opinion is that in general people who could have an important contribution would be able to handle the gui.
And it's not the case that there weren't big improvements in the gui (lua looks very nice...)
I wouldn't want the type of player that plays all day WoW to increase it's level by doing a set of repetitive actions...
About the gui, it all depends on what type of players do we want. There are many that manage the gui ... We want to have a greater community so that more people contribute and my opinion is that in general people who could have an important contribution would be able to handle the gui.
And it's not the case that there weren't big improvements in the gui (lua looks very nice...)
I wouldn't want the type of player that plays all day WoW to increase it's level by doing a set of repetitive actions...
Then quit bitching, and WRITE ONE... at least with Trepan's code we have that choice I plan to, for my project, but I am busy with a lot of other stuff that requires the rest of my talents and time, and I simply do not have the spare month free that will be required to make, skin, test, rebuild, reskin, etc., a completely custom GUI... until I have all the art assets in place and gameplay is mainly completed. I am hoping that, by then, people will have built some good frames that I can simply customize and add graphics to, frankly. And I see no reason to despair, given the very fast progress we're already seeing.Despite all the lua gadgetry nobody has yet to even attempt to sort out orders/build pics tabs, OTA style, or even copy another RTS, nobodies tried implementing proper tooltips even the failed 'New GUI' had tooltips, and nobodies even bothered setting the buildpics to appear before the commands by default.
I would just about kill for a LUA coder to do a decent GUI for us, so that I don't even have to bother with that, other than giving out design parameters and making graphics.
Unfortunately, such people don't even *exist* yet- the "widget guys" are the barebones, bootstrapping people getting things going, and they're making mistakes along the way, which is to be expected- while Trepan's work is brilliant, not all of his choices in example code are- meh, a lot of it's because he's a coder and should've grabbed a graphics artist before putting out the current GUI, so that he'd understand how they want things to feel like, but it's not that big of a deal. Eventually, they will start writing good GUIs from front to back, or helping those of us working on mods to do the same...