Page 3 of 16

Re: The future of KGS

Posted: Mon Mar 04, 2013 6:13 am
by Charlie
Image

I know that apps are ideal for many purposes. They probably are for Go clients, too. (I cannot imagine browser based Go client being successful in any mobile browser. Try use dailyjoseki's browse feature on a phone and the problem becomes obvious.)

... BUT I don't think that the current technology cocktail (HTML, CSS, Javascript, jQuery...) is going to die in a hurry, not even if Apple actively persecutes it.

Kaya.gs has shown that you can build a very usable browser based Go board.

Re: The future of KGS

Posted: Mon Mar 04, 2013 6:50 am
by hyperpape
quantumf wrote:At the risk of sounding like a conspiracy nut, if HTML5 ever does take off, and actually starts to threaten Apple's app store revenue model, you can be pretty sure that HTML5 will become demonized. Any platform that threatens Apple's control will suffer the same fate as Flash and Java.
You do sound like one. Two small points:

1. For the first year of its existence, the iPhone only supported web apps. That was because Apple wasn't ready to release an SDK, but it was widely derided as a "[poo] sandwich".
2. Apple released the iPhone with the absolute best mobile browser at the time. My understanding is that with improvements to the builtin browser and the addition of Chrome, recent versions of Android are on par, but for a long time, iOS had better support for HTML5 and better performance. (My point in saying this is not to ignite some "APPLE IS BETTER/SUXXORZ" debate, but just to say: they don't dislike the web).

But the real point is this: you say "HTML5 will become demonized". What on Earth does that mean? Apple is not going to delete Safari from its phones. And it's not going to replace HTML5 with some other technology. Even if it liked it, it just doesn't have the ability. Unless you can explain what Apple could do that could possibly be comparable to not shipping Flash/Java, there's not a lot of reason to take this conspiracy theory seriously.

Re: The future of KGS

Posted: Mon Mar 04, 2013 7:45 am
by quantumf
amnal wrote:Flash is dying out because it's rubbish. Whilst apple were among the first to make a big statement to that effect, everyone else is also doing it because far better alternatives are maturing, not because apple has conspiratorial power over them.


Flash is not rubbish. That is an absurd statement. Flash is phenomenal. Of course, it can be improved (what can't?), but Adobe didn't seem to have what it took to win this fight and seem to have given up.

Flash (and eventually, perhaps, HTML5) offer a model for rich applications to be supplied directly to users. This is what users want, but it certainly isn't what Apple (or Google, or anyone with an app store) want. There is a tension there, until recently Microsoft and Android shipped Flash because they saw it as a way to attract market share away from Apple. But once you have the market share, Flash (and eventually, HTML5) is a terrible thing, as it takes Apple/Google/MS out of the revenue stream.

Re: The future of KGS

Posted: Mon Mar 04, 2013 7:58 am
by amnal
quantumf wrote:Flash is not rubbish. That is an absurd statement. Flash is phenomenal. Of course, it can be improved (what can't?), but Adobe didn't seem to have what it took to win this fight and seem to have given up.

Flash (and eventually, perhaps, HTML5) offer a model for rich applications to be supplied directly to users. This is what users want, but it certainly isn't what Apple (or Google, or anyone with an app store) want. There is a tension there, until recently Microsoft and Android shipped Flash because they saw it as a way to attract market share away from Apple. But once you have the market share, Flash (and eventually, HTML5) is a terrible thing, as it takes Apple/Google/MS out of the revenue stream.


Sure, its stated goal is great, it's the way the actual program has always peformed poorly in many ways that I don't like. I've mostly experienced it to be a buggy, poorly supported mess, and I'm rather glad to see the back of it.

Part of this might be using linux, which adobe never supported very well, but that's part of the problem - it's a closed standard and when the plugin is poor there's no way to get a good experience. Even if flash's kind of technology is the best way to do something, that technology could be implemented very significantly better for everyone.

Modern html and javascript websites have superseded flash for many of the interactivity elements it used to be necessary for. And for whatever reason, though no doubt helped by being open official standards, they are extremely well supported with multiple competitive browser engines which avoid most or all of the problems flash has always given me. My experience is far better, and I don't really see why anyone would laud flash compared to this.

The exception could perhaps be things that flash can still do that modern open standards can't, in which case sure it's great, but as far as I can see it's so flawed in reality that it retains its usefulness only until something better comes along. Problems like inefficiency and lack of support aren't inherent to the exercise, and as long as there's so much clear room for improvement I can't bring myself to call flash phenomenal.

Re: The future of KGS

Posted: Mon Mar 04, 2013 8:05 am
by Kirby
quantumf wrote:
Kirby wrote:...

Say it's the 90s, oh about the time that KGS was initially being developed. Making a go server and client functionally capable and stable would have been a pretty epic task. And for what, exactly?


To create something that wasn't there.


Same situation now. If a web client that doesn't need Java is made, it's something that wasn't there.

quantumf wrote:...

You don't need 4 different versions of java - java is backwards compatible, so you only need the latest one.


If installing the latest version of java were like most normal software, it would update or uninstall the currently installed version on your machine.

This isn't what happens. If I install a new JRE, I have two versions on my machine that are in different directories. This is annoying to me as an end-user.

Re: The future of KGS

Posted: Mon Mar 04, 2013 8:32 am
by Phelan
amnal wrote:Modern html and javascript websites have superseded flash for many of the interactivity elements it used to be necessary for. And for whatever reason, though no doubt helped by being open official standards, they are extremely well supported with multiple competitive browser engines which avoid most or all of the problems flash has always given me. My experience is far better, and I don't really see why anyone would laud flash compared to this.

While I'm glad you've had a better experience, this is completely disingenuous. There's a lot of variance in javascript and css, if not html, between browsers. A lot of web developers complain about testing sites in IE to make them cross compatible with all browsers. I haven't delved much into HTML5 yet, but I believe the reason it exists, is to avoid a lot of the problems that exist right now.

Re: The future of KGS

Posted: Mon Mar 04, 2013 8:46 am
by amnal
Phelan wrote:
amnal wrote:Modern html and javascript websites have superseded flash for many of the interactivity elements it used to be necessary for. And for whatever reason, though no doubt helped by being open official standards, they are extremely well supported with multiple competitive browser engines which avoid most or all of the problems flash has always given me. My experience is far better, and I don't really see why anyone would laud flash compared to this.

While I'm glad you've had a better experience, this is completely disingenuous. There's a lot of variance in javascript and css, if not html, between browsers. A lot of web developers complain about testing sites in IE to make them cross compatible with all browsers.


You're right of course, there are still problems, but I perceive them to be of a fundamentally different nature to those of flash, as well as much reduced in severity. This is part of why I consider the situation far superior now.

With modern web standards, we have just that...standards. Whilst there are differences between browsers, they're centred around working to fulfil these standards such that, in principle, they will ultimately all behave the same way. In the case of IE as you mention, this is a relatively new thing - the great victory of firefox was to kickstart the modern trend towards standards being important - but even then the most recent version meets many more html5 standards than was ever the case before. The traditional complaining about IE has always been because it didn't follow those standards, thus making things harder for everyone in an (initially very successful) attempt to prevent browser competition.

Now, although we get some jostling for position in exactly how to do things we also get competition. The lack of that is a big thing that led to IE6s web stagnation, it basically didn't change at all for several years. With competition, you can't get away with doing things really badly, because there are easily accessible and well known alternatives. In fact, it's more recently led to doing somethings incredibly well, with (for instance) google chrome existing partly just to run javascript really fast and reliably so that everyone else would want to catch up...which benefits google with their increasingly popular web applications.

In contrast, flash does not have competition. When it's slow, or buggy, or simply not available for your architecture you can do...nothing at all. So both in principle and in practical terms it's a terrible thing to rely on, it isn't accessible to everyone and cannot be openly improved beyond the closed development path that adobe choose to put it on. In contrast, open web standards can be and are discussed and improved in a way that makes the web better without locking anyone out.

Edit: Actually, perhaps to be clear I should say that flash does not have competition in the field of implementing flash's language, since it's not open and is very difficult to reverse engineer. It does have competition in the form of modern javascript and html capabilities, and has proven severely unequal to that competition.

So no, things still aren't perfect, but I'm not sure there is a perfect solution and I much prefer this one because I can't think of a better way to do things. Flash, in particular, has many flaws both in theory and in practice.

I haven't delved much into HTML5 yet, but I believe the reason it exists, is to avoid a lot of the problems that exist right now.


One of the big things about html5 is that it makes adds a lot of new stuff like html video to the standard. This wasn't originally there simply because the web hadn't got to that point yet (hence flash happening to become popular as the best thing available at the time), but the standard is open and ultimately responsive to desired changes.

So...yeah, html5 exists to make things better, including solving problems and adding many new features. These new abilities, as above, cover a lot of what flash was originally popular for. Their being fully integrated into the page rather than as a plugin, and their implementations being subject to standards allowing implementation competition between browsers, has led to this way of doing things rapidly replacing flash in many places. It can be faster, easier, better integrated and less resource intensive or generally susceptible to flash's own problems.

Re: The future of KGS

Posted: Mon Mar 04, 2013 11:30 am
by Pippen
Charlie wrote:
I don't want to raise the issue of "duty" but I do think that WMS is being irresponsible. He should make up his mind: either support KGS properly or let people like Pasky make clients like Cgoban-h without fear or retribution, allowing the community to fix the things that are important to them.



You forget the third option of any owner: Let it drown and die out. F. de Groot did it with his "MoyoGo" recently. I also wonder why WMS doesn't invite some people to know his source code, giving them the opportunity to write certain improvements and test them and then it's WMS who decides if or not these programs will be implemented. It sounds risk-free though it might be wishful thinking, because I am not a programmer. But again: Maybe WMS plans to make money with KGS at one point and then he has good reason to keep it tight and that's just one reason that comes to my mind right now.

Re: The future of KGS

Posted: Mon Mar 04, 2013 11:35 am
by Javaness2
How off topic can you get.

I'd like to see the following features on KGS

[1] Different coloured boards being available again
[2] Able to navigate through the move history on a demo board
[3] Tibetan or Sunjang Baduk would be fun (yeah, it'll never happen :))
[4] The abililty for admins to have a teaching account
[5] A block challenge feature (in addition to the existing block user interaction one)

The rest seems fine to me :)

Re: The future of KGS

Posted: Mon Mar 04, 2013 12:38 pm
by Pippen
Javaness2 wrote:How off topic can you get.

I'd like to see the following features on KGS

[1] Different coloured boards being available again
[2] Able to navigate through the move history on a demo board
[3] Tibetan or Sunjang Baduk would be fun (yeah, it'll never happen :))
[4] The abililty for admins to have a teaching account
[5] A block challenge feature (in addition to the existing block user interaction one)

The rest seems fine to me :)


Well, an official stats tool where u can search with different parameters and download whole games from accounts wouldn't harm either. :) But the most important thing is: Do maintain KGS constantly, because right now it's the best server out there, so keep it winning^^.

Re: The future of KGS

Posted: Mon Mar 04, 2013 1:06 pm
by quantumf
Kirby wrote:
quantumf wrote:
Kirby wrote:...

Say it's the 90s, oh about the time that KGS was initially being developed. Making a go server and client functionally capable and stable would have been a pretty epic task. And for what, exactly?


To create something that wasn't there.


Same situation now. If a web client that doesn't need Java is made, it's something that wasn't there.



There is a pretty substantial difference between "no KGS"/"KGS" and "HTML5 KGS"/"Java KGS"

Re: The future of KGS

Posted: Mon Mar 04, 2013 1:24 pm
by PaperTiger
Pippen wrote:Maybe WMS plans to make money with KGS at one point and then he has good reason to keep it tight and that's just one reason that comes to my mind right now.


KGS is already commercialized. There are the ads, there is KGS+, and there's the Android client. However, obviously it's not enough income that he can be bothered to fix something as basic as stone clicking sounds.

And I hate to say it, but KGS is obviously "good enough" that people don't abandon it for one of the three, yes three, new servers that are already functional and actively being worked on that work in HTML5 without a client: kaya.gs, nova.gs, and funnode.com/go.

Re: The future of KGS

Posted: Mon Mar 04, 2013 1:59 pm
by Javaness2
go9dan.com is four, and there was Alex Selby's HTML server 5 years before any of these came out.

I heard the first rumour about a KGS HTML client in Alpha being ready for testing soon last month :) I doubt that it will have any new features. It will probably just be the same, but without any java GUI.

Re: The future of KGS

Posted: Mon Mar 04, 2013 3:29 pm
by Mef
Javaness2 wrote:there was Alex Selby's HTML server 5 years before any of these came out.



Is this GoShrine? or is that another to add to the list?

Re: The future of KGS

Posted: Mon Mar 04, 2013 4:56 pm
by xed_over
Kirby wrote: If I install a new JRE, I have two versions on my machine that are in different directories. This is annoying to me as an end-user.

as an developer and an end-user, I find this its best feature :)