Being a mostly C++ developer my excuses is,
1) I don't get home until 8.00pm and I am totally fuc*ed.
2) I just about know HTML, XML, CSS, Javascript let alone having to parse them.
3) I haven't programmed a Amiga seriously since 1995.
4) I really don't like trundling through others peoples code as I have to do that all day.
5) I have more interesting things to do.

But I would like to contribute financially. Is there a link, preferably paypal?

It's interesting you say there is a SkyOS port. The embarrasment! This new kid on the block with a port way before us!

Out of interest, are there any decent Java browsers out there and if so would not the time be better spent doing a fully Sun compliant JVM?

_________________

mmm..

I'm just started with classic C, not C++. I'd like to help you but my experience is very low. Hope many good developers here helps you in this mission. It is strange that are many os with a KHTML port while AOS no.

just two words.. Good Luck!

_________________
http://www.amigasoft.net

I'm going to be honest here.

I didn't like the feel of AWeb since the start. I liked Voyager, apart from the constant (And it was) crashing in it's 68k variants. IBrowse didn't get a look in until v2.2 but when I got my mits on that, I was hooked.

IB2.3 is my current browser of choice.

Fair point, it doesn't support things like CSS &, as you say, HTML4 but I rarely hit a website where any of these actually cause a problem. IB still remains very functional. There is also a lot of work going into IB2.4 & beyond so I think it's fair to give those guys a chance to prove themselves with the 'big boys'.

I think at the end of the day whether it's surfing, purchasing goods, banking, chatting on forums etc... The browsers on the Amiga get the job done. They aren't in THAT bad a state. There are more pressing issues in the short term.

That & I can't code

FuZion.

Last edited by FuZion on 20-Oct-2004 at 10:59 PM.

I had the impression that IBrowse does support some of the HTML4 tags...
Kind of partial support for HTML4....
Not all though, but which browser does that?
Not even IE seems to support HTML 4 100%.

About the KHTML, I'm sorry, can't code anything more advanced than Basic or Arexx...
This should be something that could be a joint-operation between OS4- and MOS-camp...?
I wish

_________________
Site admins are people too..pooff!

SkyOS' webbrowser is called SkyKruzer. It's more than a year old. Details can be seen here:

http://www.osnews.com/story.php?news_id=2584

About our plan: It's the intent that the engine is ported alone, so we can use it anywhere, in a program just like you would use, say a MUI custom class, not encapsulating it in a webbrowser. It should be documented and have a proper API. Much of this information can be copied from how Apple did it with Webcore, because it works that way.

KHTML is both being continually developed by the original team and Apple themselves are contributing with bugfixes, and access to really talented people.

I even forgot to mention that AtheOS even has a KHTML browser, called ABrowse. It can be seen here:

http://www.campusprogram.com/reference/en/wikipedia/a/ab/abrowse.html

It seems to be done the same way as Apple Webcore...

Regards,
Henrik Mikael Kristensen
AWeb Development Team

Last edited by HMK on 20-Oct-2004 at 11:07 PM.

From the Abrowse-link:

"Because of difficulties with the amount of work required to maintain current ports of Khtml on the Syllable operating system, ABrowse's use of the Khtml layout engine shall be dropped by the Syllable Internet Tools Team in favour of the more-portable Gecko layout engine."

So, opinions differ i suppose.

Quote:

I had the impression that IBrowse does support some of the HTML4 tags... Kind of partial support for HTML4.... Not all though, but which browser does that? Not even IE seems to support HTML 4 100%

It doesn't. When I developed the amiga.dk site, I discovered this as vertical centering of text in table cells doesn't work in IBrowse. That feature is part of HTML4. Thus the text looks a bit funny in IBrowse but looks perfectly in all other browsers.

Investigating this further, I found more things that weren't supported and finally found that IBrowse officially only supports HTML3.2.

I don't want to put IBrowse off in the least. It's a great browser, just with an engine that is a little behind, compared to Voyager and AWeb in terms of features.

Comparing them to modern browsers however shows that HTML is today, if you take advantage of the technology, only a very small part of how a webpage can be built, using the Document Object Model, where you can do some really nifty interactivity stuff, which would totally kill things like AmigaGuide. CSS is not even enough and more things are coming. You can build almost full GUIs with the Document Object Model and when people realize that on a larger scale, they'll start using it, just like they started using CSS.

Amigabrowsers will then not be able to display such pages at all.

Quote:

This should be something that could be a joint-operation between OS4- and MOS-camp...?

The plan really is to make it completely AmigaOS-clone agnostic, so it can be used on AmigaOS3.x, MOS, OS4 and AROS. This is something so valuable for the general OS experience that we shouldn't try to compete with eachother at all. Much better results will come from working together.

Regards,
Henrik Mikael Kristensen
AWeb Development Team

Quote:

So, opinions differ i suppose.

Of course something that should be taken into consideration. Looking closer, it seems either development has stopped or moved, so it's hard to compare Gecko and KHTML on the developer level.

We chose KHTML because it's a much smaller engine than Gecko. There's about 2 MB of sourcecode in Apple Webcore, which includes the KWQ layer which separates KHTML from the underlying parts of the OS. In short it's very Amiga like.

Gecko is probably even more portable, but then I think we would have seen a Gecko browser running on some AmigaOS clone by now...

KHTML renders about 90% of what Gecko does, but it does it much faster and with a smaller memory foot print.

The key here is that KHTML has already been ported to other places, we just need to duplicate those methods.

Here's an interesting (but aged) link to read which compares KHTML and Gecko, which explains basically why Apple chose KHTML:

http://developers.slashdot.org/comments.pl?sid=50694&threshold=4&commentsort=0&tid=154&mode=nested&cid=5082502

Quote:

Out of interest, are there any decent Java browsers out there and if so would not the time be better spent doing a fully Sun compliant JVM?

I don't think that's a good idea, based in if we want to see a browser that should be able to run on lowend PPC Amigas, such as CyberstormPPC and Blizzard PPC. KHTML would probably even run fairly well on a 68060 with 32 MB RAM. Can't do that with Java. This would be going in the totally opposite direction of the original Amiga philosophy of keeping things simple and small. The browser with Java would take up as much diskspace alone as the rest of the OS...

We would also loose the ability to use the HTML engine as part of other applications. We want an intelligent implementation of a good engine.

Regards,
Henrik Mikael Kristensen
AWeb Development Team

I totally agree with this. Not necessarily the route we should take, but we HAVE TO HAVE an updated browser. This is a must. I like Ibrowse, my browser of choice, and I have confidence in the Ibrowse team. Porting an html engine from open source could not hurt the Ibrowse team any! I understand why our browsers couldn't do all those things in the past, we were limited by our hardware! Not anymore! If we have any prayer of becoming an attractive alternative OS, we have to have a good/up to date, browser.

I also think this is something KMOS should be paying for someone to port. What do they think they are going to do with OS 4.0 with no up to date browser?


Yogi27

@HMK

Quote:

The plan really is to make it completely AmigaOS-clone agnostic, so it can be used on AmigaOS3.x, MOS, OS4 and AROS. This is something so valuable for the general OS experience that we shouldn't try to compete with eachother at all. Much better results will come from working together.

That sounds like a great plan. Would it be wrong to think you are looking at a 3.X conversion, and then recompile, make changes, etc for the AROS, MOS, and OS4 versions? I wish something could happen to bring everyone back together developing again.

Quote:

About our plan: It's the intent that the engine is ported alone, so we can use it anywhere, in a program just like you would use, say a MUI custom class, not encapsulating it in a webbrowser. It should be documented and have a proper API. Much of this information can be copied from how Apple did it with Webcore, because it works that way.

So what you are saying is that once the KHTML port is done, it could be used as a new engine for IBrowse, AWeb, Voyager or whatever, bringing them up to modern standards? Am I understanding that correctly?

Last edited by Herewegoagain on 21-Oct-2004 at 12:02 AM.

Could be a very good incentivate to know that who port KHTML and does a browser withj it (i think that can be usable and not need to be complex at first)�@will can get the Amizilla $8000+ USD booty!!!!!!!! So if you can do it, port KHTML and do a small but usable (not bugs all time) browser, and you will win some money and help to your platform(s)...

Personally I think that aweb guys can help to Bill Panagouleas with AmiZilla...

Quote:

But I would like to contribute financially. Is there a link, preferably paypal?

Sorry, no. We don't operate AWeb development with finances because of a lot of problems that would give, such as who should have the money for how much development. It can't be done fairly when there are many people involved. Continous contributions would also be required to keep development going.

@FuZion
"They aren't in THAT bad a state."

Without wanting to sound too negative, I think you're trying to kid yourself. They are in a horrible state.

You might think that because most of the pages you visit mostly work that's ok. But as a web developer, there's no way I'll design anything to work with HTML 3.2 browsers. If the pages work in a current Amiga browser that's more luck than judgement, and I know that much of the styling on the pages I manage will be broken on IBrowse, AWeb, etc.

A browser is engine is #1 priority for the platform right now.

Chris

@Yogi27: "Porting an html engine from open source could not hurt the Ibrowse team any!"

You seem to forget that IBrowse is a commercial product - they expect people to pay money for it. Because of that, they MUST NOT use any code from a GPL product, and most Open Source products are GPL.

The problem with Open Source code is that it can only be used in other GPL projects. That effectively kills any prospect of commercial sales and means that all the programmers have to work for nothing.

Professional people have better things to do.

tony

_________________
cheers
tony

Hyperion Support Forum: http://forum.hyperion-entertainment.biz/index.php

Quote:

Would it be wrong to think you are looking at a 3.X conversion, and then recompile, make changes, etc for the AROS, MOS, and OS4 versions? I wish something could happen to bring everyone back together developing again.

Right now AWeb 3.5 beta is being ported to OS4 native and soon there will be a release of the 3.5 beta port. The process that went underway for this port, should be similar to that of porting it to MorphOS and finally to AROS (if it ever gets ReAction support).
It's taken a few months to get there, but Fabio Alemagna and Andy Broad are doing an excellent job at this. A lot of knowledge has been gained.

The key here is getting GCC support working across all of the platforms. AWeb can currently be compiled on AmigaOS3.x, Linux, MacOSX and probably soon also OS4. It would be nice to write code and crosscompile it on GCC on a fast and cheap x86 Linux cluster with MOS, AROS, OS3.9 and OS4 settings without too much work to get native versions for each OS, right?

It would be similar to writing an application for Linux and then port it to NetBSD, FreeBSD, OpenBSD and MacOSX. A little work for each platform, but not too much.

Quote:

So what you are saying is that once the KHTML port is done, it could be used as a new engine for IBrowse, AWeb, Voyager or whatever, bringing them up to modern standards? Am I understanding that correctly?

Yes, something like that. You probably can't just stuff it in there in 5 minutes, but it could be possible. Or build a new webbrowser around it. That's why it'll be important that a finished port needs to be developer friendly with a proper API and documentation for it.

This way we could get other browsers updated very fast. Their authors could focus a lot more on interface and browsing experience rather than engine debugging and updating.

SimpleMail and YAM could benefit from it too. Possibly real HTML editors will be possible too.

Regards,
Henrik Mikael Kristensen
AWeb Development Team

@tonyw
"You seem to forget that IBrowse is a commercial product - they expect people to pay money for it. Because of that, they MUST NOT use any code from a GPL product, and most Open Source products are GPL"

That's not really true and commercial browsers based on KHTML already exist. Here's one:

http://www.omnigroup.com/applications/omniweb/

You pay for the interface and featureset, not the engine. Changes to KHTML get fed back and everyone benefits. Likewise, Apple don't open-source all the code for Safari, but feed back heavily into KHTML.

So yes, IBrowse could take advantage in the same way.

Chris

So are you saying something in effect that you'd want to build components similar to Apple's WebCore and JavaScriptCore to make them more modular and make them more friendly for multiple browser development efforts?

What would be easier to port; Gecko or KHTML?

_________________

Quote:

Yes, something like that. You probably can't just stuff it in there in 5 minutes, but it could be possible. Or build a new webbrowser around it. That's why it'll be important that a finished port needs to be developer friendly with a proper API and documentation for it. This way we could get other browsers updated very fast. Their authors could focus a lot more on interface and browsing experience rather than engine debugging and updating. SimpleMail and YAM could benefit from it too. Possibly real HTML editors will be possible too.

@HMK

Thanks for the info. This is really sounding like something very positive now. I too see a very nice potential for this project. I've always liked IBrowse best of all of the Amiga browsers, but I can see where it would benefit from such efforts.

Quote:

But as a web developer, there's no way I'll design anything to work with HTML 3.2 browsers.

Indeed. But I do test designs in Links and Lynx. These (the latest versions) are available for AmigaOS aren't they?

But I agree fully that AmigaOS needs a decent browser at the level of Safari at least. I'm sure that there are lots of developers who could create a Reaction interface for the browser, it just needs someone to port KHTML into a Reaction widget with a defined API.

Apple and the KHTML team have done a tonne of work already on a lot of these things. It just needs someone to change the OS dependent code (preferably so that the changes go back into the main KHTML codebase too) to add AmigaOS support.