pouet - new platforms
category: general [glöplog]
i'm probably oldfashioned somehow but platforms in my mind are supposed to be stable and shouldn't somehow lose the ability to run prods that were specifically tailored for them. so browser as a platform sounds about as bad as tandy or trs-80 used to sound back in the day. smth with moving goalposts and whatnot, i guess... which is not to say that your arguments are wrong or invalid or anything like that, it's just my personal observation as someone not involved in this kind of decisions :)
or call it 'web' as it's supposed to run as a web application in your favorite browser and usually with a server (be it electron)
Very tricky one, Windows has obviously changed a lot in the last 20 years or so, no distinction is made for iteration on features so I think the same situation applies to browser based demos (just my opinion of course).
My thought would be just to consider the release date of the production for any iterative platform and make an educated guess from there.
My thought would be just to consider the release date of the production for any iterative platform and make an educated guess from there.
same argument can be made for any pc based platform, be it windows or linux or browser, with linux maybe being the worst in terms of backwards compatibility of demos.
still browser would be by now a much better platform than javascript, same as linux is a much better platform than "c++"
still browser would be by now a much better platform than javascript, same as linux is a much better platform than "c++"
The "moving platform" issue would be fixed / addressed by this suggestion.
Browsers are trying pretty hard to not break backwards compatibility with the web (I know, I'm working on one), but when it happens, it's usually in the name of security. Which is kind of hard to do anything about :-) (Of course, there's also insanity like enumerating all methods on some system object; intros are different from web pages in that regard.)
no saga it wouldn't actually solve anything, but mostly just serve to explain why stuff won't run on a given system. which is not totally useless, but far from a solution at the same time.
unless maybe someone cooks up some kind of sandboxing environment where the average user can run random outdated windows/directx/browser/etc versions. or emulate such things, idk, i'll gladly leave it to ppl who actually care about pc demos to decide about that
unless maybe someone cooks up some kind of sandboxing environment where the average user can run random outdated windows/directx/browser/etc versions. or emulate such things, idk, i'll gladly leave it to ppl who actually care about pc demos to decide about that
Using the tag list we'd then create an ebay shopping list for the hardware and a bootable self-contained image for the software, with a sleek one-click checkout so you get all the needed hardware and software delivered to your doorstep over the next couple months, as well as a custom generated assembly manual coming with a disk/cartridge/floppy of the prod in question, the manual will also contain the phone number of the main programmer of that prod, in case you need support.
Anyways we're straying far from the original request that sparked this conversation, which was to simply rename "javascript" to "browser", so its:
Adding tagging to platforms is cool and all that but is not a requirement for that change so lets not blow up the scope to "rewrite pouet from scratch" just yet.
I think the only thing holding us back from that is:
So since I had some support on the matter here we go:
(whole sheet with new icon replacing js icon, fetched 15 minutes ago)
Anyways we're straying far from the original request that sparked this conversation, which was to simply rename "javascript" to "browser", so its:
- More inclusive (shadertoy/glslsandbox/css prods)
- Prevents mislabeling ("windows" and "linux" prods being just "browser" prods)
- Is actually accurate (see my previous post)
Adding tagging to platforms is cool and all that but is not a requirement for that change so lets not blow up the scope to "rewrite pouet from scratch" just yet.
I think the only thing holding us back from that is:
Quote:
1. Don't bother suggesting unless you have an icon for it. Seriously. Pull your weight.
So since I had some support on the matter here we go:
(whole sheet with new icon replacing js icon, fetched 15 minutes ago)
"us"...? speak for yourself, and please do consider that this new category would need maintenance.
havoc, I'm speaking for the people who supported the renaming, that'd be NR4, gasman, v3nom and myself, isn't "javascript" being maintained already?
sure but that is problematic enough as-is and what you're suggesting is a clear case of moving goalposts so i wouldn't want to play that game, sorry
As far as I can tell the only way this moves goalposts is by allowing prods on pouet that are currently absent because they don't have a platform, it's either eventually introducing platforms for "glsl", "whlsl", "webassembly", "css" and "svg" (which are all languages, not platforms) or renaming "javascript" to what it should've been from the very start: "browser" because that's a platform not a language.
btw. "glsl" alone would have at least 20 prods that could be added, thanks to tokyo demofest (iirc there was also another party having a shadertoy compo at some point).
btw. "glsl" alone would have at least 20 prods that could be added, thanks to tokyo demofest (iirc there was also another party having a shadertoy compo at some point).
how would you envision shadertoy prod downloads?
Quote:
The question here is, what icon to use, I have trouble coming up with a vendor agnostic icon for a browser, maybe a globe? Or go and use the old, now defunct internet explorer icon?
ye ol' Netscape logo!
…just scaled down a bit.
iirc tokyo demofest uses glsl-sandbox so the shaders are just plain glsl, it sure would be nice to provide a simple stand-alone generic player in the zip of the glsl-sandbox and shadertoy prods for the sake of preservation and ease of use, but it's not like amiga entries come with the hardware or an emulator bundled, glsl-sandbox is open source and has quite a few forks and there are plenty of shadertoy player projects for a plethora of platforms.
Sesse: not sure that logo will be legible in 10x10 pixels ;)
gasman: same for me. though oddly enough, that file doesn't seem to run on the platform LJ was proposing. so your answer appears to be incomplete, or out of context.
LJ: thank you, yes this adresses the core of the problem i've been trying to hint at. some kind of generic/foolproof/userfriendly solution to run those files is necessary if the aim is to keep the site interesting for non-experts. imho that solution doesn't have to be included in every release archive but it should be easily accessible, preferably free, and (most importantly) operable by relatively dumb/uninitiated userlevel persons. like a fantasy console... i can download it, and use it to run an entire category of prods. maybe a proper shadertoy player could perform this function too, i'm not sure though because i've tried a few of those and with each try managed to find stuff that didn't work so properly within an hour or so. maybe the situation with glsl-sandbox is less confusing for the average user, i'm not sure, i never explored glsl-sandbox far enough to know.
all of these are questions that could be but as of yet have not been answered. aka, the potential for a platform is there, it just needs a little more work. but also the basic idea/proposal that you pitched needs a bit of work i think. "browser" as a category looks to be an oversimplification, like "windows", "linux" and "ms-dos" are too. the user would need to master quite advanced knowledge about browsers to get most of the prods in the proposed category to actually run, it's nothing like the ease of use we know from fantasy consoles for example. also, this all combined poses the question, is it even a good idea to add a generic platform category when we already can predict that for each category within the proposed platform we'll need a separate player program? from a maintenance POV it then certainly seems easier to define "shadertoy" (etc) as a category than the ultra generic "browser".
LJ: thank you, yes this adresses the core of the problem i've been trying to hint at. some kind of generic/foolproof/userfriendly solution to run those files is necessary if the aim is to keep the site interesting for non-experts. imho that solution doesn't have to be included in every release archive but it should be easily accessible, preferably free, and (most importantly) operable by relatively dumb/uninitiated userlevel persons. like a fantasy console... i can download it, and use it to run an entire category of prods. maybe a proper shadertoy player could perform this function too, i'm not sure though because i've tried a few of those and with each try managed to find stuff that didn't work so properly within an hour or so. maybe the situation with glsl-sandbox is less confusing for the average user, i'm not sure, i never explored glsl-sandbox far enough to know.
all of these are questions that could be but as of yet have not been answered. aka, the potential for a platform is there, it just needs a little more work. but also the basic idea/proposal that you pitched needs a bit of work i think. "browser" as a category looks to be an oversimplification, like "windows", "linux" and "ms-dos" are too. the user would need to master quite advanced knowledge about browsers to get most of the prods in the proposed category to actually run, it's nothing like the ease of use we know from fantasy consoles for example. also, this all combined poses the question, is it even a good idea to add a generic platform category when we already can predict that for each category within the proposed platform we'll need a separate player program? from a maintenance POV it then certainly seems easier to define "shadertoy" (etc) as a category than the ultra generic "browser".
Quote:
generic/foolproof/userfriendly solution to run those files is necessary if the aim is to keep the site interesting for non-experts.
since when is pouet caring about foolproof/userfriendly solutions and about keeping the site interesting for non-experts. non-experts are typically already scared away already by the layout and navigation - how to even find something worthwhile as a newbie?
I fail to see how this can be the deciding factor for making an already broken category/platform less broken, by renaming it to sth. that makes more sense and is more consistent.
ever since the site exists v3nom. take for example disk images, offering those is done to make things easier for the average user. even if it's not done consistently, each instance where it is done, it makes someone's life easier. or so we hope...
your gripes about layout (etc), those may be valid but let's dilute the discussion here with such topics, feel free to add them to a more suitable thread (where i won't answer because i don't design the site, i just maintain the data shown on it)
your gripes about layout (etc), those may be valid but let's dilute the discussion here with such topics, feel free to add them to a more suitable thread (where i won't answer because i don't design the site, i just maintain the data shown on it)
+1 for having a browser platform over javascript.
wouldn't mind seeing other specific sub-categories either, like shadertoy, glsl, etc. but doubt it would ever get approved on pouet.
regarding the concerns of archiving shadertoy/glsl entries, i see it as a non issue for the renaming/merging of the category that is currently being discussed since the problem already exists on current database regardless if the entries are on javascript or browser or wild.
but if you seriously would want to address it, it wouldn't be that hard to pull off either, with some basic userscripts to store the assets, text and metadata on a basic import/export standard. there already are some tools for doing that (shaderboi for example lol). and it wouldn't be hard to talk with the few parties that actually hold those compos to start being more careful in the future using them to make sure things are archived properly instead of the current statusquo of keeping a .txt with a link or not archiving at all. so i think this derailment is not about it being doable at all but about who should be doing it and why they haven't been doing it and the knowledge that it probably wouldn't be done by pouet moderators. which, as mentioned, is beyond the scope of what categories the prods fall under discussion imho.
wouldn't mind seeing other specific sub-categories either, like shadertoy, glsl, etc. but doubt it would ever get approved on pouet.
regarding the concerns of archiving shadertoy/glsl entries, i see it as a non issue for the renaming/merging of the category that is currently being discussed since the problem already exists on current database regardless if the entries are on javascript or browser or wild.
but if you seriously would want to address it, it wouldn't be that hard to pull off either, with some basic userscripts to store the assets, text and metadata on a basic import/export standard. there already are some tools for doing that (shaderboi for example lol). and it wouldn't be hard to talk with the few parties that actually hold those compos to start being more careful in the future using them to make sure things are archived properly instead of the current statusquo of keeping a .txt with a link or not archiving at all. so i think this derailment is not about it being doable at all but about who should be doing it and why they haven't been doing it and the knowledge that it probably wouldn't be done by pouet moderators. which, as mentioned, is beyond the scope of what categories the prods fall under discussion imho.
How many early Windows demos run without issues on "modern" Windows 10/11?
How many DOS demos run without issues on a stock DOSBox configuration?
Moving platforms have always been an issue, and I don't quite understand why we should draw a line at the Web Browser platform now, especially considering that it has a comparatively good track record w.r.t. compatibility.
How many DOS demos run without issues on a stock DOSBox configuration?
Moving platforms have always been an issue, and I don't quite understand why we should draw a line at the Web Browser platform now, especially considering that it has a comparatively good track record w.r.t. compatibility.
keyj: my hope is to (perhaps, where possible) avoid such issues as we've witnessed with older platforms on the new "browser" (group of) categor(y/ies). edge cases are a given, in the future as they were in the past, single platform categories spanning multiple decades with wild inconsistencies in terms of technological possibilities/backgrounds as we all know and suffer from, such as dos and windows, is what i'm hoping to avoid.
Quote:
some kind of generic/foolproof/userfriendly solution to run those files is necessary if the aim is to keep the site interesting for non-experts. imho that solution doesn't have to be included in every release archive but it should be easily accessible, preferably free, and (most importantly) operable by relatively dumb/uninitiated userlevel persons.
For browser demos, the usable-by-non-experts solution is the "view online" link - which will of course be a link to shadertoy.com in the case of shadertoy prods. I know that doesn't cut it for archival purposes, but the download link fills that gap. Unless I'm missing something, I don't think there's any reason why the "archival quality" and "suitable for casual consumption" links have to be the same thing. (Sure, if shadertoy.com disappears from the internet overnight then it'll need some knowledgeable people to bring those shaders back to readily-viewable form, but I think we can live with that.)