Would it be feasible for Hyperion's programmers to develop their own OS4 specific web browser? They, of course, know OS4 better than anyone. It should be faster and work better than ports of browsers designed for other systems. If they all worked together on it maybe it wouldn't be too huge of a project.
Of course I've never attempted to make a web browser so I don't know how much work it would be. I also don't know how busy the Hyperion programmers already are.
It's not a simple task to design a browser from scratch. It's easier to take a working browser engine or even a browser that uses it and port it. This is exactly what happens with OWB and NetSurf.
Jack
"the expression, 'atonal music,' is most unfortunate--it is on a par with calling flying 'the art of not falling,' or swimming 'the art of not drowning.'. A. Schoenberg
Would it be feasible for Hyperion's programmers to develop their own OS4 specific web browser?
No. Even multi-billion companies like Apple, Google, Nintendo, etc., don't try to do something like that but use existing engines (Apple and Google: WebKit, Nintendo: Presto) for their browsers (Safari and iPhone, Chrome/Chromium and Android, DS browser and Wii internet channel).
Quote:
They, of course, know OS4 better than anyone. It should be faster and work better than ports of browsers designed for other systems. If they all worked together on it maybe it wouldn't be too huge of a project.
If both full time OS4 developers would stop all other AmigaOS related work I guess they could complete an AmigaOS browser created from scratch in about 20-30 years. But only a browser comparable to the current 4 major engines, not something which will be usable when it's completed ...
If they'd port Mozilla/FireFox instead it would take at most 3-6 months.
As long as I'm working alone in my spare time on it I guess it will take about 5 years to create an AmigaOS 4.x browser comparable to FireFox, Opera, etc., from OWB. Unlike FireFox OWB isn't a complete browser but basically "just" a portable version of the WebKit engine and I have to implement a browser around it myself.
Today I just got that... you work alone on that project! Congratulations! In so short time you made something really amazing (looking for movies and pictures, since I could try it yet...).
Personally I was always against situation when we do not have any "fully" functional web browser (or any program) but we have/had Aweb, Ibrowse, Voyager, now OWB, NetSurf and people start thinking about Firefox...
Personally I think all this people would work together on one project we would have really amazing web browser... and yes, nothing for free :).
No it would not be feasible and it has not much sense. It is a huge task, Hyperion has other things to do and they are not web specialists. More, we already have a browser that need to be supported, OWB.
@Kreciu
Joerg does a amazing job on OWB and he works alone on it. OWB is often improved but we see that it take months with the engine almost done. Imagine what it could be if a whole browser would have to be created from scratch ?
Even if it takes 5 years, web technologies will change during this time ...
IF Hyperion was to do any browser for AmigaOS4 I really hope they would focus on Firefox. Firefox is well known on other platforms and is used by a lot of people and see frequent updates.
Making their own browser from scratch wouldn't make any sense. Perhaps if possible, buying IBrowse and using the webkit engine could make sense, but I say they should go for Firefox.
btw:I would really like to see someone take those bounty money.
Yes unfortunately there's too many people hoping for Mr Someone
And unfortunately a Mr. Someone is not enough for such a big project
Not being a developer myself (don't tell me to learn code, pls!) all I can do is hope and contribute with a little $$ for those who actually produce something in our community.
Yes unfortunately there's too many people hoping for Mr Someone
And unfortunately a Mr. Someone is not enough for such a big project
Not being a developer myself (don't tell me to learn code, pls!) all I can do is hope and contribute with a little $$ for those who actually produce something in our community.
You should learn to code.
I know that software development isn't for everyone and the donations are appreciated. I wish that I could work on my projects full time, but I also need money.
@people who can code
If you want to see something, then that Mr Someone is you. When people were moaning about no display-lists in MiniGL, I realized that no-one else was going to do it, so I did it myself. Now MiniGL 2.0 has advanced to the point where we could get something like Blender working (with others helping out too). Likewise, I decided that I couldn't wait until someone else wrote drivers for newer graphics cards, so I started working on it myself. The same is true for software developed by other developers such as Joerg, Afxgroup, Salass00, Rigo, SpotUP, and many others (who I can't think of right now).
So, if you see something that needs programming, and have the ability to it yourself, don't wait around for someone else to pick up the project and don't hope that Hyperion will do it. Get working yourself.
When you tell "Get working yourself", you are right but I have to add many developers are there to help, so "Get working yourself" does not mean "Get working alone". You did not say that but I just want to tell RacerX, Troels and others that they can feel confident if they want to start programming.
Once, I tried to port SQLite and it failed (almost done but still problems with partial locks on files). I didn't wait for someone else does the port. Even if that happens Still motivated I contributed to compile the last version on OS4 and I created a MUI example to explain how to use the sqlite lib. I didn't wait the last version was available. And today, both are available !
Sometimes programming is hard, disappoints, ... but that's so great !
There are many languages to start programming (AmiBlitz, C, C++, Ruby, Python, ...). And our website gurumed is back online one week ago after months/years with articles and tutorials (in french ...).
About the first post in this thread, we have to be realistic and use our resources in the most clever way. For internet browser, OWB was designed to be portable. So let's go, Joerg did the right choice, avoiding to spend months understanding internals of a browser engine (it may be the case though).
For internet browser, OWB was designed to be portable. So let's go, Joerg did the right choice, avoiding to spend months understanding internals of a browser engine (it may be the case though).
You don't have to understand the internals for a FireFox port either. OWB is portable (using SDL it runs everywhere without any changes) and you immediately get something usable and can implement the missing parts one after another, but unlike FireFox it's not a complete browser and you have to implement a lot yourself. An AmigaOS 4.x port of FireFox port is less work than creating a complete browser from OWB, the only problem is that for a FireFox port you have to work for some months porting all the libraries it needs to AmigaOS without getting anything usable. Working on improvements of an already usable browser like OWB is much more fun
When you tell "Get working yourself", you are right but I have to add many developers are there to help, so "Get working yourself" does not mean "Get working alone". You did not say that but I just want to tell RacerX, Troels and others that they can feel confident if they want to start programming.
Correct. But someone has to take the first big step in order to get a project started. After that you may find that others will want to help out.
@Joerg
Have the AWeb guys ever contacted you? IIRC, the plan was to create a new AWeb built around the KHTML toolkit (i.e., webkit). I have no idea how easy/hard it would be to rip out its display engine and replace it with OWB's core, but it is an idea.
IIRC, the plan was to create a new AWeb built around the KHTML toolkit (i.e., webkit). I have no idea how easy/hard it would be to rip out its display engine and replace it with OWB's core, but it is an idea.
If you'd do that the only parts of AWeb still usable would be some small parts of it's GUI (not the menus or anything else depending on it's ARexx port), the cookie handling and it's network functions (incl. memory and disk cache with cache browser), but that's something you definitely don't want to use, the only old AmigaOS browser with usable network support was Voyager (but the rest of it had way to much bugs), Voyager supports persistent HTTP 1.1 connections, AWeb and IBrowse don't but open a new connection for each request which is very slow.
There was Smithy which started a browser from scratch some years ago but stopped last year even if some parts were finished. The browser's name was paihia
in feb 2006 :
Quote:
Grande Paihia is a modern, small, fast, and portable web browser.? With complete support for XML, HTML 4, CSS 2.1, DOM level 3, and EcmaScript 3, Grande Paihia is more than capable of dealing with most of your internet needs.? Grande Paihia is written from the ground up and is not based on any existing web software.
Versions for all the various Amiga systems are planned.? Progress is slow but steady.? As of November 2005, the CSS engine is all but done and being tested using the W3C test suite - biggest issue to arise from testing is performance.? The language parts of EcmaScript are done, the object library isn't.? And DOM level 3 is about 20% complete.? All of this and the source code for the Paihia web engine is less than 700k uncompressed.? Visual Studio compiles the engine to a binary around 300k in size - a tiny fraction the size of anything else in the field!
in March 2007, Smithy wrote : Quote:
After 3 years of Paihia development I think I've had enough. I am thinking about releasing the source code to interested people - but I'm not sure exactly yet. Perhaps the code might be worth something to me in years to come. There is 1.6Mb of code uncompressed. That's about 60,000 lines of C++. Lots of OO patterns used that have ensured everything is nicely decoupled. 40% of the code is unit tests.
For example. The CSS engine (not the layouter part, the bit that parses and calculates property values) could be quite easily integrated into another browser. All the other browser needs is a document tree that implements an interface. In turn the CSS engine maps quite straightforwardly into a display/layout engine.
Alternatively, another browser could replace their document tree stuff with Paihia's. This would give tree event support, including both bubble-up, top-down models. As well as the whole visual/CSS stuff.
Paihia's XML processor (a validating parser - it passes about 70% of W3 XML conformance test suite.) uses some fancy factories to ensure that there is no visual/display stuff in it. This means it could be used in people's applications that use XML but don't need all that visual extra stuff. And at 10,000 lines of code it's very small and fast.
The downsides:
The layout engine is most in need of work. It supports block/inline/float boxes, the box model (borders, padding, etc..), fonts, colours but no tables. Probably not much use to people.
There is a working, but messy Javascript interpreter that hasn't been compiled for 2 years - it doesn't compile against the current build. And none of the core Javascript library has been implemented either.
There is no HTML support. (it was going to be done via a CSS stylesheet) but that won't handle dodgy HTML.
No HTTP support.
No DOM interfaces have been written, although it should be quite easy doing it on top of the internal tree.
There is no documentation at all and code comments are few.
Erm, I've done all the development in Visual C++. The whole browser is in pure C++ though. The system-specific stuff is very small and there's a simple API to access it. The only implemented API is one for Java. The API is extremely simple anyway: opens windows, draws text and shapes. I've got no idea if it'll compile on GCC (it definately won't compile in SAS, Hisoft C++ or Storm because it uses lots of STL and new C++ stuff). But there are no MS-specific C++ stuff or libraries involved.
Basically, I'm fed up. And this isn't even thinking about the whole quagmire that the Amiga scene seems to be in. To get a complete browser will probably take me the rest of the year of working. I do enjoy programming but I feel this project has taken its toll on me in many ways. I suspect my future hobbyist coding projects will be somewhat smaller.
Anyway, is anyone interested or should I just consign it all to a CD and let it gather dust......?
Would be cool if Smithy can release its sourcecode, maybe directly on sourceforge or somethings like it, i'm quite sure that some code can be usefull for our community
Working on improvements of an already usable browser like OWB is much more fun
That's the main reason any developers code in their own time, without the fun element AmigaOS would be without a lot of programs. Please keep enjoying your Amiga!!!!
I guess the true answer is : - feasible -> yes (nothing is impossible), - however is it considered -> no (would be a huge waste of ressource and time, moreover if you consider that during this time the OS would not be updated so much)
'feasible' in my understanding means "likely" or "sensibly". I don't think it is feasible for Hyperion to go into browser writing. It *would* be a dreadful waste of resources.
Seriously, if you do want to contact me write me a mail. You're more likely to get a reply then.
News feed
OS4 Feedback forum
OS4Depot Feedback forum
Amigabounty forum
OpenAmiga forum
AmiCygnix forum
ABC shell forum
AmiKit forum
Cinnamon Writer forum
CodeBench forum
Digital Universe forum
Dopus 5 forum
E-UAE forum
Gnash forum
Ibrowse forum
JAmiga forum
Odyssey forum
OWB forum
Qt forum
SmartFileSystem forum
Timberwolf forum
TouchDevice forum
TuneNet forum
Unsatisfactory Software forum
Still motivated I contributed to compile the last version on OS4 and I created a MUI example to explain how to use the sqlite lib. I didn't wait the last version was available.
