On Wed, 10 Aug 2016 22:45:41 +1000 onefang Rejected <onef...@gmail.com> said:
> On Thu, 4 Aug 2016 15:17:38 +0930 Simon Lees <sfl...@suse.de> wrote: > > > > > > > On 08/04/2016 02:57 PM, Carsten Haitzler (The Rasterman) wrote: > > > On Thu, 4 Aug 2016 15:16:03 +1000 David Seikel <onef...@gmail.com> > > > said: > > > > > >> On Thu, 4 Aug 2016 14:00:01 +0900 Carsten Haitzler (The Rasterman) > > >> <ras...@rasterman.com> wrote: > > >> > > >>> On Thu, 4 Aug 2016 13:23:33 +1000 David Seikel <onef...@gmail.com> > > >>> said: > > >>> > > >>>> On Thu, 4 Aug 2016 11:29:50 +0900 Carsten Haitzler (The > > >>>> Rasterman) <ras...@rasterman.com> wrote: > > >>>> > > >>>>> one thing with your evas 3d usage... it's not real/useful to > > >>>>> anyone else. if you made open source tools/apps that were easily > > >>>>> usable and downloadable (without needing special accounts you > > >>>>> have to pay for e.g second life) then you would be > > >>>>> interesting/relevant. > > >>>>> > > >>>>> let me give you a sample: > > >>>>> > > >>>>> if you made a really nice world clock app for seeing world > > >>>>> timezones, selecting yours, setting date/time etc. that even did > > >>>>> sexy stuff like used topological maps of the world so when you > > >>>>> zoom into an area you get some mountains/hills and so on... and > > >>>>> this was part of e's settings tools or clock tool with calendar > > >>>>> etc. ... well then that'd be really cool and useful to LOTs of > > >>>>> people and accessible to everyone. :) just saying. you're emails > > >>>>> are "i'm working on this thing.. here on my hard drive... it > > >>>>> does x/y/z and will do x/y/z". that's not REAL to anyone. it's > > >>>>> not accessible. it's not used daily thus important. :) you want > > >>>>> evas 3d to be useful to your BIGGER projects you do like this, > > >>>>> do some smaller ones out the front that people see and use > > >>>>> daily. :) > > >>>> > > >>>> Er, my project is for Second Life, OpenSim, AND an alternative to > > >>>> both, and is on Github. Your arguments are invalid. Hell, my > > >>>> general idea is to make server side so efficient that by default, > > >>>> anyone can run their own little server to invite their friends > > >>>> to, even on their phone. I'm trying to open up 3D virtual > > >>>> worlds to the masses. > > >>> > > >>> "that are easily downloadable and usable without special > > >>> accounts". someone has to set up a server, run it, need accounts, > > >>> etc. is it a tool readily usable out of the box that people > > >>> actually will need/use regularly. a secondlife or quivalent is > > >>> not. a terminal is. a wm is. a web browser is. ... if someone has > > >>> to set up a server themselves or register accounts somewhere else > > >>> at all you just failed the "useful to anyone" :) > > >> > > >> If you read SledjHamr.org, you'll see I have taken all of that into > > >> account. Now you are just making up excuses to ignore me. > > > > > > count the number of people here participating in any virtual world. > > > they use terminals. email clients. wm's. the number of people > > > involved in virtual worlds who would even look is small at best. > > > perhaps you don't like it, but i'm trying to tell you that the > > > reason no one is paying attention is you do not have something they > > > want. maybe others do, but no one here working on efl or e or core > > > users etc. is. or almost no one. if they were, then you wouldnt be > > > a periphery thing. i am not saying your work is bad or its stupid > > > or anything. i'm saying that expecting relevance is dependent on > > > things being relevant to others. if efl breaks terminology then > > > that break becomes relevant because LOTs of people use it here and > > > even elsewhere. but especially here. if evas_3d was being used in e > > > for eg 3d cube desktop switching effects... it'd be relevant to a > > > whole lot of people immediately. > > > > > > i'm not even going to look at those projects because the premise of > > > them as a virtual world is not interesting or relevant to me. it is > > > to you and that's great. not to me. and if you go around wondering > > > why they are not paying attention then this is the primary reason. > > > you have to have things in common that make the things you want be > > > relevant. :) i'm trying to help you here by pointing out that to > > > help the project you care about, you may have to do other side > > > projects that have relevance to others to bring that to the fore. :) > > > > > > > A gadget to display models of parts of virtual worlds inside E would > > be pretty interesting maybe. > > What, like the file requester showing 3D thumbnails that Raster keeps > arguing against? Yeah, I want that to. B-) it has nothing to do with you. i equally would reject someone saying "lets put a full pdf viewer in a thumbnail preview with form filling and page flipping etc.". you've paid 90% of the cost of loading it completely anyway. yes argue LOD all you like - the cost of creating a 3d scene, memory cost of loading osmesa (almost 100k on its own you can never get back once loaded), is the vast majority of the cost. load more or fewer triangles etc. wont be the issue. i also would reject that the file selector loaded full edje objects in a preview and had them run as this cost is non-trivial. i also reject if you said "hey use photocam to live load the whole image at scaled down and allow the user to pan/zoom in the preview". NO. no no no. generate a thumbnail or SERIES of thumbnails you can loop/ping-pong and then the cost paid to load this is a cost we already pay to load any icon or thumbnail and already paid - jpeg/eet etc. for the ui. have a background headless process generate thumbs and store on disk for re-use multiple times. no. i do not think having a full 3d model viewer in a file selector is a good idea. i do not think having a full pdf viewer or video player or music player or zooming/panning image viewer either is a good idea. yes. i know e's file selector plays full videos in the preview. it's a silly amount of overhead. it's only because e didnt thumbnail videos problem that it does this. i have a decentish stand-alone thumbnailer for rage that i've messed with. that's more like it but i'd use a cut down version that just collects 5 or 10 shots from the video and makes a still montage of it. > > Extra extra bonus points if you can merge the concept of a virtual > > world and a Window manager, I've always wanted to be able to walk > > between my windows its about time someone reinveted 3D window > > managers, if you started work along those lines id definitely be > > interested. > > Keep in mind that what I am working on is an entire 3D virtual world > system, which is a huge thing, including lots and lots and lots of > little bits. Any generic bit I can get others to write will be really > helpful. Taking time to write bits that wont be useful for the > virtual world work is anathema to me, it's all gonna take too much > time as it is. I tend to write things generically, so that I can reuse > them elsewhere. So extra extra bonus points if people actually looked > at my stuff, and picked out the things they might consider useful. > Don't leave it up to me, I can't market for shit. I don't know what > bribes EFL developers want, but I do know what virtual world users want. i can tell you that i am pretty much never going to go pick through a whole virtual world engine for 3d effects. i can more than manage them myself with less effort. in fact any 3d effects e might need are doable without evas_gl - evas_map can do all of it. simotek told you the same thing i did. if you took the time out to make what you want to see happen relevant as part of the day to day lives of the users of e, efl then what you care about would become important. you do say you don't have the time to do it. in fact you say it's an anathema to you. that will mean then your community is a virtual world/life sim community like secondlife and the people working on efl and e are not part of that and you're not magically going to make them part of it by wishing so. > The integration of EFL widgets displayed on Evas_3D models, and > Evas_3D models displayed on EFL widgets is very interesting to me, and > I do intend to make use of that in my virtual world project. > Ultimately the ability to show any arbitrary window in world, to other > people, is highly desirable. So that's part of my plan, and should > help with your 3D window manager idea. The lack of a useful internal > window manager means I have already started writing my own for this > project. By "internal window manager" I mean managing the internal EFL > windows inside the main window, even allowing internal windows to be > "ripped off" to become their own main window (a commonly requested > feature for Second Life and OpenSim users that no one has actually > implemented yet). > > Just off the top of my head http://SledjHamr.org/ already has - > > A working network transparent GUI system based on Lua and EFL. I even > intend to get it working in web browsers (Lua via WASM / asm.js and > maybe a HTML translation layer). This has lots of features, and > planned features, including the above mentioned internal window manager. > It's based on my earlier matrix-RAD system that was written in Java. > > Better versions of some internal EFL APIs that should be made public > APIs. > > A high speed Lua script runner that can handle thousands of scripts, > based on my edje_lua2 work and some academic work by others. > > So if EFL developers want little bribes so we can all have a decent 3D > system, then EFL developers should actually look at my stuff and pick > out suitable bribes. I ain't gonna take time out to build bribes that > wont be useful to ME, if that's the way this is to work. So go look at > http://SledjHamr.org/ and figure out what bribes con work for us all if > bribery is the thing we need to do. you may call it bribes. it's called shared interest. if you want me to pick over sledjhamr or whatever and build a fileselector preview for 3d models... not going to happen. 1. i have precisely ZERO such models these days to look at so i have no driver to do it (i used to have some for lightwave i created but hey... long gone by now). 2. in my priority list of things its not even on it - there is so much else cluttering it up, 3. it'd be STUPIDLY inefficient to stuff a 3d model viewer even as a lodable module in a file selector - ad above at the top. every piece of complex code like this is an opening for a bug that will crash the app with the file selector in it. the more complex the data being viewed, or the rarer its testing, the more chance for bugs that bring your whole app down. no! it should be a back-end that generates an imageset on disk in a file and the fileselector just loops/ping-pongs over a set of 10 or 20 images to give you multiple frames with an object rendered at multiple angles. same for videos. same for edje objects (more likely they will just be still single image thumbnails) etc. etc. - but it'll be the person who WANTS TO SEE 3d models in their file selector as icons/previews that will have to do this work. i'll happily do the work for things you might not care about but that i do. as for getting people interested in things like sledjhamr. look at the site. it's a wall of text that's links to more text files and then more and more of that as a diagram which makes me go "what on earth is this?" and then go back to my overflowing todo list instead. sit and complain " i can't do marketing" ... but unless you learn at leas SOME of it and not just learn but PRACTICE it, people are just not going to spend the time to dig. start with a page and atg least a screenshot that looks nice where people go "ooh i'd like to have that... so how?" i look at the site and no download link i can see, or get it or get started guide is obvious... oh wait ... buried as the 3rd last major bullet point "building..." not very obvious that's where i get it... oh wait a text file where links in it i can't click on - have to copy and paste.. but it discussed getting efl not your project.. oh a github page. finally.. click on that oh wait.. i see a repo with JUST a README?... so there is no code? what? oh wait i have to go to experimental? well gee finally something.. but this is scary. its telling me that this stuff likely doesn't work as its experimental and i have to clone and know how to switch branches so wait i have now a tree? what makefile? a cmake thing? is there autofoo? how does this thing build?... oh gee. and i don't even know what i will get after all this effort and now having to figure out more... not even a screenshot. enlightenemnt isn't some shining example of brilliance in marketing but i'm trying to help you here. your project just screams "go away". i don't even have the first carrot of "ooh look at this sexy stuff here... you could have this if you just did the following steps!". you need to do this man! i have nothing there that even close to wants me to dig further. but i did dig for the purposes of this mail and every level of digging just makes things hard, opaque and i wonder what was the reason i came this far? you claim to have some virtual world system. i don't even see screenshots. it likely doesn't even work. that's the impression people get. YES. you need to do better in marketing if you want people to have an interest and care. if you don't want to bother, then accept the fact you'll not have people take an interest. spend the time to put up carrots. spend some time clearly documenting how to get it, build it, run it, test it. do not expect someone to hunt page after page and copy & paste links and hunt through branches and then have to keep hunting. get them from zero to hero as fast as you can with some simple docs that work. your build system is entirely unknown to man and beast. you have to be very clear how it works. i have no clue how it does. again - i'm trying to help you here. i'm not claiming to be some gift-from-above in this, but a seasoned developer like me who knows various build trees when they see one is wondering... "so how? i run lua built.lua? is that it? i don't know! the README says nothing... oh wait there buried in the MIDDLE of a wall of text .. in a paragraph is "run build.lua" ... put it on a line by itself: ./build.lua so i can see that its obviously a build command! :) you need to slicken things up. you need to organize your website. you need to provide bait/carrots to get someone to bother following the rest of the chain. :) -- ------------- Codito, ergo sum - "I code, therefore I am" -------------- The Rasterman (Carsten Haitzler) ras...@rasterman.com ------------------------------------------------------------------------------ What NetFlow Analyzer can do for you? Monitors network bandwidth and traffic patterns at an interface-level. Reveals which users, apps, and protocols are consuming the most bandwidth. Provides multi-vendor support for NetFlow, J-Flow, sFlow and other flows. Make informed decisions using capacity planning reports. http://sdm.link/zohodev2dev _______________________________________________ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
