| [Files] AWeb 3.5 APL Lite Beta released | ANN.lu |
| Posted on 21-Jun-2004 16:51 GMT by Henrik Mikael Kristensen | 82 comments
View flat
View list | |
The first official AWeb 3.5 APL Lite Beta is now released. It can be found at http://aweb.sunsite.dk. Read more for changelog.
WHAT'S NEW
Here is an overview of the new features and major bugs fixed in this
release. For a complete list in "fine detail" see the Changelog
Included at the end of this document.
HTML
Support for <INS> and <DEL> tags
RENDERING
Double Buffered Rendering - Optional double buffering dramatically
reduces the flickering during rendering of long table based pages.
Background Images - no longer force a complete redraw off the
document, resulting in a dramatic speed up of some pages.
LAYOUT
Fixes and standardisation to the way tables are laid out, and to the
way background images in tables are align mean that many more pages
render "correctly".
INTERNAL IMAGEDECODER
Fix to the transparency bug in the internal gif decoder.
GOPHER SUPPORT
Improved gopher support.
CHARACTER SETS
Support for translating character sets enables pages using a
different character set to be displayed.
SPLASH SCREEN
New more attractive startup screen and about requester.
AWEBPLUGINS
Fixes to a major bug in the image decoder plugins where AWeb would
crash with bad memory trashing on repeated iconification.
AWEBLIBS
The aweblib version numbering has been changed, all version start
from 35 as of this release. This avoids some clashes that would other
wise occur with previous version of AWeb.
awebjs.aweblib has been renamed javascript.aweblib for the same
reason.
JAVASCRIPT
Fundamental support for Regular Expression objects has been
introduced in this version. However, the string methods using regular
expressions have not yet been implemented and as most real world
sites uses these methods few new sites will work, rather this is the
first step on the journey.
DEVELOPER COMPILE
A compile of the developer version of AWeb (AWeb.developer) is included in
the archive. It is not installed by the Install script. It allows developers
and beta testers to generate extended debug info and should only be used for
this purpose. It contains no extra browsing functionality.
Have Fun!
AWeb APL Development Team
|
|
| AWeb 3.5 APL Lite Beta released : Comment 1 of 82 | ANN.lu |
| Posted by brotheris on 21-Jun-2004 15:06 GMT | | This certainly is right move in right direction. Congrats. |
|
| AWeb 3.5 APL Lite Beta released : Comment 2 of 82 | ANN.lu |
| Posted by Joël EHRET on 21-Jun-2004 16:24 GMT | Yeahhhh
Bigfoooooooooooooooooooootttt !!
Please do the MorphOs Build ! :) |
|
| AWeb 3.5 APL Lite Beta released : Comment 3 of 82 | ANN.lu |
| Posted by Crumb // AAT on 21-Jun-2004 17:02 GMT | | keep up the good work guys! :-) |
|
| AWeb 3.5 APL Lite Beta released : Comment 4 of 82 | ANN.lu |
| Posted by Emeric SH on 21-Jun-2004 17:22 GMT | | Excellent job!!!! A MorphOS native version would be even more nice! :) |
|
| AWeb 3.5 APL Lite Beta released : Comment 5 of 82 | ANN.lu |
| Posted by Amon_Re on 21-Jun-2004 17:47 GMT | In reply to Comment 4 (Emeric SH):
Isn't the source available for download? It shouldn't take long to compile it i guess.
Anyway, it's nice to see my fave Amiga browser to make improvements :) |
|
| AWeb 3.5 APL Lite Beta released : Comment 6 of 82 | ANN.lu |
| Posted by Nate Downes on 21-Jun-2004 18:10 GMT | | I hate to be a cloud on this sunny day, but where is CSS support? Without that, you're still a crippled browser. |
|
| AWeb 3.5 APL Lite Beta released : Comment 7 of 82 | ANN.lu |
| Posted by Henrik Mikael Kristensen on 21-Jun-2004 18:23 GMT | In reply to Comment 5 (Amon_Re):
I would like to encourage developers to tell us, if they have done a specific version for MorphOS or OS4 so we can host it and learn of the changes that have been done, so others can do compiles for MorphOS or OS4 and keep everything under one roof.
Pretty please. :-)
Regards,
Henrik Mikael Kristensen
AWeb Development Team |
|
| AWeb 3.5 APL Lite Beta released : Comment 8 of 82 | ANN.lu |
| Posted by Henrik Mikael Kristensen on 21-Jun-2004 18:25 GMT | In reply to Comment 6 (Nate Downes):
CSS is pretty hard to implement in the current AWeb. Not impossible, just very hard. That's one of the reasons we are rooting for a KHTML port, which is maintained by an external source. |
|
| AWeb 3.5 APL Lite Beta released : Comment 9 of 82 | ANN.lu |
| Posted by brotheris on 21-Jun-2004 18:45 GMT | In reply to Comment 8 (Henrik Mikael Kristensen):
Lite means SAS/C ? |
|
| AWeb 3.5 APL Lite Beta released : Comment 10 of 82 | ANN.lu |
|
| AWeb 3.5 APL Lite Beta released : Comment 11 of 82 | ANN.lu |
| Posted by Moohoo on 21-Jun-2004 20:56 GMT | In reply to Comment 7 (Henrik Mikael Kristensen):
http://80.197.46.38/~bigfoot/files/AWebPPC-src.lha
Ported over 2 years ago. For some reason this port wasn't accepted as base for gcc AWeb. |
|
| AWeb 3.5 APL Lite Beta released : Comment 12 of 82 | ANN.lu |
| Posted by beauty on 21-Jun-2004 21:32 GMT | In reply to Comment 1 (brotheris):
nice.. but aweb is still ugly, and it's conception is crazy... why 4 settings menu items with no sense at all!
oh.. is it possible to make a flash plugin? there is some opensource projects on linux? |
|
| AWeb 3.5 APL Lite Beta released : Comment 13 of 82 | ANN.lu |
|
| AWeb 3.5 APL Lite Beta released : Comment 14 of 82 | ANN.lu |
| Posted by Nate Downes on 21-Jun-2004 23:22 GMT | In reply to Comment 8 (Henrik Mikael Kristensen):
hard or not, without it we're still discussing a 1995-grade web browser in 2004. That is 9 years obsolete. |
|
| AWeb 3.5 APL Lite Beta released : Comment 15 of 82 | ANN.lu |
| Posted by Amon_Re on 22-Jun-2004 05:53 GMT | In reply to Comment 7 (Henrik Mikael Kristensen):
Yea, i noticed that on your the site, earlier versions were already compiled for MOS & whatnot :)
Cheers |
|
| AWeb 3.5 APL Lite Beta released : Comment 16 of 82 | ANN.lu |
| Posted by Fabio Alemagna on 22-Jun-2004 06:00 GMT | In reply to Comment 11 (Moohoo):
> Ported over 2 years ago. For some reason this port wasn't accepted as base for
> gcc AWeb.
The reasons have been explained plenty of times: it's a quick and dirty port, full of #ifdef's and, basically, good only for MOS.
Anyway, the port to gcc was finished months ago, it's just that this is the first beta since then. |
|
| AWeb 3.5 APL Lite Beta released : Comment 17 of 82 | ANN.lu |
| Posted by Fabio Alemagna on 22-Jun-2004 06:02 GMT | In reply to Comment 14 (Nate Downes):
> hard or not, without it we're still discussing a 1995-grade web browser in 2004.
> That is 9 years obsolete.
Is there any amiga browser that supports CSS /in any acceptable form) at the moment?
This is a genuine question. |
|
| AWeb 3.5 APL Lite Beta released : Comment 18 of 82 | ANN.lu |
| Posted by Fabio Alemagna on 22-Jun-2004 06:02 GMT | In reply to Comment 14 (Nate Downes):
> hard or not, without it we're still discussing a 1995-grade web browser in 2004.
> That is 9 years obsolete.
Is there any amiga browser that supports CSS (in any acceptable form) at the moment?
This is a genuine question. |
|
| AWeb 3.5 APL Lite Beta released : Comment 19 of 82 | ANN.lu |
| Posted by Anonymous on 22-Jun-2004 06:28 GMT | In reply to Comment 16 (Fabio Alemagna):
>The reasons have been explained plenty of times: it's a quick and dirty port, full of #ifdef's and, basically, good only for MOS.
>Anyway, the port to gcc was finished months ago, it's just that this is the first beta since then.
Well, AWeb is my "daily" browse. So I'm very please to see new version.
You say that bigfoot's sources are dirty. Well, they are extremely clear compare to the mess on the official release.
Moreover, thisone is a bit more unstable than bigfoot's one
I'm not speaking about the sources themself, but about the build...
What about stoping using UNIX like build and make a proper makefile ?
@Bigfoot: my hero, shoot again :-)
Bye |
|
| AWeb 3.5 APL Lite Beta released : Comment 20 of 82 | ANN.lu |
| Posted by Thomas Würgler/Pagan on 22-Jun-2004 06:36 GMT | In reply to Comment 19 (Anonymous):
It is extremely stable on my µA1 in OS4 - and has been for the last couple of internal betas. |
|
| AWeb 3.5 APL Lite Beta released : Comment 21 of 82 | ANN.lu |
| Posted by code on 22-Jun-2004 06:41 GMT | In reply to Comment 20 (Thomas Würgler/Pagan):
define extremely |
|
| AWeb 3.5 APL Lite Beta released : Comment 22 of 82 | ANN.lu |
| Posted by Fabio Alemagna on 22-Jun-2004 06:42 GMT | In reply to Comment 19 (Anonymous):
> You say that bigfoot's sources are dirty.
NOt exactly. I said it was a quick and dirty port... it couldn't have been done better in the very short time it was done, and it was done in an amazingly fast way. Mind you, this is not meant to be in any way a hash towards bigfoot.
> Moreover, thisone is a bit more unstable than bigfoot's one
If you find any bugs, please report them.
> I'm not speaking about the sources themself, but about the build...
> What about stoping using UNIX like build and make a proper makefile ?
I'm not sure what you mean by "proper makefile", but AWeb's build system has never been so clean :-) You just need make, and a couple of other utilities, which you have if you have gcc installed. Besides, some of us use crosscompilers to build AWeb. |
|
| AWeb 3.5 APL Lite Beta released : Comment 23 of 82 | ANN.lu |
| Posted by Fabio Alemagna on 22-Jun-2004 06:45 GMT | In reply to Comment 22 (Fabio Alemagna):
> [...] a hash [...]
Ehum... dunno how that came out, but it was meant to be "harsh" :-) |
|
| AWeb 3.5 APL Lite Beta released : Comment 24 of 82 | ANN.lu |
| Posted by Kjetil on 22-Jun-2004 07:09 GMT | In reply to Comment 22 (Fabio Alemagna):
If find it quite annoying that I need Automake, Imake, and some other tools that are not provided whit that AmigaOS40 SDK, every time I try to compile some thing.
Way not make a Makefile for MOS only one for OS40 only and one for AROS only and 68k Amigas only, and not depend on external utility’s every time,
I typed install binutils it looks like it broken inn UAE and AOS40, complains about wrong version numbers, when I check whit version command it’s the same..??? |
|
| AWeb 3.5 APL Lite Beta released : Comment 25 of 82 | ANN.lu |
| Posted by Fabio Alemagna on 22-Jun-2004 07:23 GMT | In reply to Comment 24 (Kjetil):
> I find it quite annoying that I need Automake, Imake, and some other tools that
> are not provided whit that AmigaOS40 SDK, every time I try to compile some
> thing.
I do too. In fact, you need nothing like that here.
> Way not make a Makefile for MOS only one for OS40 only and one for AROS only
> and 68k Amigas only, and not depend on external utility’s every time,
There's no need for different makefiles, AWeb doesn't use "standard" makefiles anyway, it uses some special features of GNU's make which allow for nifty things. The system we use is named "TUBS". Have a look at the various Makefile.tubs files you find in the source tree.
> I typed install binutils it looks like it broken inn UAE and AOS40, complains
> about wrong version numbers, when I check whit version command it’s the
> same..???
I'm not sure of what you're talking about there. |
|
| AWeb 3.5 APL Lite Beta released : Comment 26 of 82 | ANN.lu |
| Posted by Thomas Würgler/Pagan on 22-Jun-2004 07:48 GMT | In reply to Comment 21 (code):
extremely = "it has not crashed". |
|
| AWeb 3.5 APL Lite Beta released : Comment 27 of 82 | ANN.lu |
| Posted by Anonymous on 22-Jun-2004 08:19 GMT | Thanks! A good update -- keep `em coming.
Two sites that I know is handled correctly now,
compared to earlier versions of AWeb:
www.amigaworld.net
www.aros.org |
|
| AWeb 3.5 APL Lite Beta released : Comment 28 of 82 | ANN.lu |
| Posted by Anonymous on 22-Jun-2004 09:36 GMT | In reply to Comment 20 (Thomas Würgler/Pagan):
>It is extremely stable on my µA1 in OS4 - and has been for the last couple of internal betas.
Hey, how about posting some screenshots of it? Seeing as AWeb uses Reaction,
I think it would blend in very nicely with the overall OS4 look. Add to
that anti-aliased fonts and I would imagine it looks very nice. Good for
simple web browsing/HTML documentation
Thanks |
|
| AWeb 3.5 APL Lite Beta released : Comment 29 of 82 | ANN.lu |
| Posted by Bill Hoggett on 22-Jun-2004 09:59 GMT | In reply to Comment 18 (Fabio Alemagna):
Fabio, the answer is "no". None of the currently available Amiga or MorphOS browsers support CSS in any usable way.
It's also true that this is a crippling factor in this day and age. I've been beta-testing an online multiplayer wargame based on the old 8-bit classic Lords of Midnight, which is played through a browser and will be going public in the next couple of months or so, but alas it needs CSS, so things like BeOS' NetPositive and QSSL's (QNX) Voyager and all the current Amiga browsers don't work. BeOS and QNX have Mozilla ports, but AmigaOS and MorphOS users are screwed. If anyone completes a port of KHTML then things might get interesting. I'm afraid I no longer believe the Amizilla project will ever bear fruit, but it would be cool if I was to be proved wrong. |
|
| AWeb 3.5 APL Lite Beta released : Comment 30 of 82 | ANN.lu |
| Posted by Don Cox on 22-Jun-2004 10:08 GMT | In reply to Comment 29 (Bill Hoggett):
"It's also true that this is a crippling factor in this day and age. I've been beta-testing an online multiplayer wargame based on the old 8-bit classic Lords of Midnight, which is played through a browser and will be going public in the next couple of months or so, but alas it needs CSS, "
Does it need anything else, such as Flash or Java ? Or is it just CSS and Javascript?
I agree about the crippling factor. All the efforts put into MorphOS and AOS4 are wasted until there is a complete browser, with CSS and up to date Flash. Streaming Real and WM are needed too. Very difficult. |
|
| AWeb 3.5 APL Lite Beta released : Comment 31 of 82 | ANN.lu |
| Posted by Bill Hoggett on 22-Jun-2004 10:16 GMT | In reply to Comment 30 (Don Cox):
Just CSS. That's the only limitation. The actual game is written in PHP with a MySQL back end, but all that's server based anyway. If it uses Javascript at all it would only be for the buttons, and that part works fine in IBrowse and AWeb (not tested in V). However the main display panel - the one that shows you what is happening - requires CSS. Without CSS the components of that panel are shown all over the page and do not make up the image you are supposed to see. The effect is exactly the same on all non-CSS browsers, including the non-Amiga ones. |
|
| AWeb 3.5 APL Lite Beta released : Comment 32 of 82 | ANN.lu |
| Posted by hooligan/dcs on 22-Jun-2004 10:17 GMT | | Amizilla booty is currently at $8748.30 + unlimited amount of res & glory, maybe even a place in heaven next to Elvis. |
|
| AWeb 3.5 APL Lite Beta released : Comment 33 of 82 | ANN.lu |
| Posted by Anonymous on 22-Jun-2004 10:27 GMT | OFF TOPIC
IBrowse 3.0 is supposed to be all this and more! Will handle CCS, Flash
and so on. But given that IBrowse 2.3 was released in January 2003 (1 1/2 years ago), and the small update 2.4 is still not released, I fear that v3.0 will be released in 2-3 years. That's just too late... If there was one thing Hyperion should pour funds into it is the web browser. But of course, there aren't enough competent coders around to tackle this, even with the money...
To summarize: Wait 3 more years before you will see a proper browser... |
|
| AWeb 3.5 APL Lite Beta released : Comment 34 of 82 | ANN.lu |
| Posted by Henrik Mikael Kristensen on 22-Jun-2004 11:38 GMT | In reply to Comment 33 (Anonymous):
Much of this waiting period could probably be eliminated, if people stopped reinventing the wheel all the time, because the "ultimate browser engine" will take a few years to invent no matter how hard you try. There are extremely good browser engines out there that can be ported. Why not take advantage of them? Everyone else does, even the one-man project SkyOS.
Reasons:
- You don't need to maintain anything, but the porting process for updated version of the engine. That is much, much less work than building a browser engine from the bottom up, especially after the initial porting process is done.
- There are hordes of developers maintaining browser engine technology today, some of those are in close contact with or working in major companies and institutions. They are skilled and have many years of experience in browser technologies.
- Levelling of technology, much like OpenGL is spread across many platforms, so should browser engine technology be. Most browsers today are no longer limited to one platform.
- Focus on extending the use of browser technology in other places than the browser, rethinking the use of the engine in the Amiga way where everyone can take advantage of the engine in their applications, rather than focusing on supporting all XHTML tags, where other platforms were years ago.
Browsers like Mozilla and Firefox are now shifting focus after a long period of engine development. There is time now to improve the interface and optimize things allround and you can feel by each version that the browser is getting faster, smaller and more user friendly.
We are actually doing the same for AWeb, not packing many more features in, but only optimizing existing ones. We are basically doing work, which is already done by others, where we could spend time porting a new engine.
The only three reasons the original AWeb is being maintained this way is so:
- That people with smaller Amigas still can use an optimized browser
- We don't have manpower to do more things than that
- Being an open source project, people can work at what they want to work with. We can only encourage, not force.
I also feel that instead of making this into a race of who-comes-first-wins-the-money, share knowledge, be a little altruistic about development (make it into something that everyone can take advantage of) than being concerned about milking money from this small platform.
It's better for the platform and better for everyone in the end, whether they are AmigaOS3/4, MorphOS or AROS users.
Regards,
Henrik Mikael Kristensen
AWeb Development Team |
|
| AWeb 3.5 APL Lite Beta released : Comment 35 of 82 | ANN.lu |
| Posted by Nate Downes on 22-Jun-2004 14:56 GMT | In reply to Comment 18 (Fabio Alemagna):
I hate to tell you, but the Amiga is irrelevent because of this kind of attitude. "Does any Amiga browser support it?" is a pointless question, the question remains "Does THIS browser support it? No? Then it is obsolete." Don't like this fact, then change it, but being an ostrich and bringing up really irrelevent issues just wastes my time, and pushes the whole platform further into the grave.
So my task for you is "what will you do to solve this glaring hole in the platform?" |
|
| AWeb 3.5 APL Lite Beta released : Comment 36 of 82 | ANN.lu |
| Posted by greenboy on 22-Jun-2004 15:29 GMT | In reply to Comment 29 (Bill Hoggett):
>I've been beta-testing an online multiplayer wargame based on the old 8-bit classic Lords of Midnight, which is played through a browser and will be going public in the next couple of months or so, but alas it needs CSS, so things like BeOS' NetPositive and QSSL's (QNX) Voyager and all the current Amiga browsers don't work.
I haven't followed the changes real closely but QNX has had CSS available for years now for its 6.x RTOS Voyager - Voyager simply was designed/modularized to become the UI for various engines starting with Opera's, then Mozilla and Access Netfront. There may be other engine options I'm not aware of that are also in use, or being ported/built.
The idea was to supply several footprints for various embedded profiles up to the developers' desktop machines.
>BeOS and QNX have Mozilla ports, but AmigaOS and MorphOS users are screwed. If anyone completes a port of KHTML then things might get interesting. I'm afraid I no longer believe the Amizilla project will ever bear fruit, but it would be cool if I was to be proved wrong.
Amen. I had hoped that we could get everybody on the same page, but it is within the realm of possibility we might end up with ... both? We'll see. |
|
| AWeb 3.5 APL Lite Beta released : Comment 37 of 82 | ANN.lu |
| Posted by Fabio Alemagna on 22-Jun-2004 15:35 GMT | In reply to Comment 35 (Nate Downes):
> So my task for you is "what will you do to solve this glaring hole in the
> platform?"
I'm not the Amiga saviour, and I only work for free if I can at least get fun as a reward. For money, though, I can work on anything. If you want, we can make a deal, a real one, and you can contract me for the work of, say, porting an existing engine to Amiga or Amiga-like platforms. |
|
| AWeb 3.5 APL Lite Beta released : Comment 38 of 82 | ANN.lu |
| Posted by greenboy on 22-Jun-2004 16:06 GMT | In reply to Comment 34 (Henrik Mikael Kristensen):
>Much of this waiting period could probably be eliminated, if people stopped reinventing the wheel all the time
- or telling us that spokes should be enough and that wheels are just needless eye-candy - horrible bandwidth wasters too! Great consolations, those... ; }
>because the "ultimate browser engine" will take a few years to invent no matter how hard you try. There are extremely good browser engines out there that can be ported. Why not take advantage of them? Everyone else does, even the one-man project SkyOS.
Yep. See QNX example, comment 36.
>Reasons: {...snipped}
I agree entirely. It is also a plus that the developers involved will not skew their efforts to exclude this "Amiga reimplementation", or that late-to-the-show "MorphOS clone" ; } ...They will simply not be drawn into such limited trench exchanges. Which practically guarantees that other political/cultural limitations will not be considered acceptable by them. They will be too busy trying to stay near the current norm for expectations of functionality, and maybe trying to keep things from being too cordoned by Microsoft fences.
>I also feel that instead of making this into a race of who-comes-first-wins-the-money, share knowledge, be a little altruistic about development (make it into something that everyone can take advantage of) than being concerned about milking money from this small platform.
I don't think it such a bad thing that some financial "thank you!" be available for all who work toward the worthy goal of an all-can-use browser solution. But it has made me wonder if treating it like PRIZE money and thus encouraging splinter cells to go off to win a race is the best way of utilizing very limited programmer resources, and then if some code should make its way to the finish line, what will it be like when code maintenance and updates are needed? And they will.
It doesn't seem like our little neighborhood has enough people around who grok what is good {arguably at least some parts are ;} about Open Source mindsets. I'm hoping that some obstructionist tendencies which I think are totally against the spirit of various licenses about sharing the sources and the knowledge - well, I hope some will get reappraised - and that those who have a clue about this will through their example and patience kind of infect all the others who could be making it easier for our teensy little niche platforms.
...It also seems like Bounty programs are springing up left and right. After everybodys' wallets get chipped away at for a few months or years and a Handout Mentality becomes even more established, where will all those attempting other ambitious voluntary or commercial projects be left? Will expectations become even less ideal? There should also be some incentive for people to form small companies and to fill some OTHER software category holes - or we should just forget about our Amiga-derived paths. Because they are just too stunted to be either fun OR reasonably complete tools.
Just some other things to think about. |
|
| AWeb 3.5 APL Lite Beta released : Comment 39 of 82 | ANN.lu |
| Posted by dantheman on 22-Jun-2004 16:45 GMT | >Much of this waiting period could probably be eliminated, if people stopped
>reinventing the wheel all the time, because the "ultimate browser engine" will
>take a few years to invent no matter how hard you try. There are extremely
>good browser engines out there that can be ported. Why not take advantage of
>them? Everyone else does, even the one-man project SkyOS.
This seems like a bit of a dig at Paihia. We looked into porting a browser
engine in some depth, but decided it'd be a vast amount of more work to
maintain. I'll repeat our reasoning here:
>Reasons:
>- You don't need to maintain anything, but the porting process for updated
>version of the engine. That is much, much less work than building a browser
>engine from the bottom up, especially after the initial porting process is
>done.
>- There are hordes of developers maintaining browser engine technology today, >some of those are in close contact with or working in major companies and
>institutions. They are skilled and have many years of experience in browser
>technologies.
This is the most common myth when it comes to the argument between new browser vs ported browser. The myth is that these hoardes of developers from other platforms will implement all these super new web standards and everything will be wonderful. But of course, there is no such thing as a free lunch, and when you look further into what porting actually involves you'll discover a very expensive lunch...
This myth is based on the failure of Amiga-like browser authors to implement modern-day web standards. It assumes the real reason for their failure is
because the modern-day web standards are large/complex/difficult, but in fact all these standards specify is how to draw text/lines/boxes on a
screen and various different languages to write these things. Blah blah. No big deal. What this myth ignores is the fact that new web standards are
released few and far between, whereas these magical browser engines and the technologies they're dependent on are released very often. And these new versions aren't implementing new web standards, they're just fixing bugs and adding user-agent features. I'll repeat this point because it's important: for one release of a web standard, there are hundreds, if not
thousands, of releases of new browser engines and the technologies they use.
By porting a web browser engine you are multiplying your workload by hundreds! Instead of just having to support new web standards - of which a major one might
be released once every 5 years, you have to keep up with every single new release of your browser engine. And don't give me any guff about it being just a simple re-compile. The enabling technologies of the popular browsing engines Qt, and Gtk (KHTML depends on Qt and Gtk depends on Gtk ) are updated all of the
time too, many of their new features will of course be used by new versions of the browser engines. And this isn't even taking into an account of Mozilla (you can choose to either re-implement X or win32's enormous APIs).
So. To conclude my meandering points, either write a browser engine yourself and update it in 5 years when a new standard comes out. Or port a browser engine and spend every day of the next 5 years porting new parts of X's, win32's, Qt's, or GtK's APIs, all of which are absolutely enormous and have had thousands of people spending decades working on them, and that doesn't even include time to work on your user agent features. Of course, that's assuming you manage to port one of these APIs in the first place, and correctly duplicate all of the side-effects and caveats.
Once again, in a soundbite:
Initial development time (new browser): Long time
Initial development time (ported browser): Very long time (if it's even possible to correctly re-implement a huge and old API)
Maintenance work (new browser): A lot
Maintenance work (ported browser): Less
Woop! Doing! Pah! Hopefully I've torpeodoed this issue once and for all now!
-dantheman, the Paihia Team |
|
| AWeb 3.5 APL Lite Beta released : Comment 40 of 82 | ANN.lu |
| Posted by Bill Hoggett on 22-Jun-2004 16:51 GMT | In reply to Comment 36 (greenboy):
"I haven't followed the changes real closely but QNX has had CSS available for years now for its 6.x RTOS Voyager - Voyager simply was designed/modularized to become the UI for various engines starting with Opera's, then Mozilla and Access Netfront."
Except that this does not explain why some sites work in Mozilla dn Opera, but not with Voyager. So what's the explanation?
I can only tell you categorically that Voyager on QNX 6.2.1 (the last official NC) does not handle the site in question, reacting exactly like all the other browsers which don't support CSS. Opera does handle the site (I didn't test Moz), although it has glitches that don't appear in any other port.
The problem with providing a single rendering module for everything is that if that module is less than perfect you're stuck with the deficiencies across the board.
Maybe there's a default setting in QNX I need to change to enable CSS? If so, I haven't seen it. |
|
| AWeb 3.5 APL Lite Beta released : Comment 41 of 82 | ANN.lu |
| Posted by greenboy on 22-Jun-2004 16:56 GMT | In reply to Comment 39 (dantheman):
>Woop! Doing! Pah! Hopefully I've torpeodoed this issue once and for all now!
Really, dantheman: if your team can do it, I'm sure most everybody will be pleased to see the results! : }
But in the meantime, no actual torpedos are to be had at any price in waters hereabouts - only promises, rumors, and preliminary blueprints for torpedos have been sighted ; }
Land ho? |
|
| AWeb 3.5 APL Lite Beta released : Comment 42 of 82 | ANN.lu |
| Posted by greenboy on 22-Jun-2004 17:25 GMT | In reply to Comment 40 (Bill Hoggett):
Bill, I don't know for sure, but I think the current QNX 6.3 eval version downloads with the ACCESS Netfront engine. I just recall that was the plan last year. Time to do some catching up, I guess. |
|
| AWeb 3.5 APL Lite Beta released : Comment 43 of 82 | ANN.lu |
|
| AWeb 3.5 APL Lite Beta released : Comment 44 of 82 | ANN.lu |
|
| AWeb 3.5 APL Lite Beta released : Comment 45 of 82 | ANN.lu |
| Posted by Bill Hoggett on 22-Jun-2004 17:50 GMT | In reply to Comment 43 (greenboy):
Precisely. The back end - the one doing the rendering - is different. With Opera what you get is the Voyager front end and the Opera back end, which is why the site works in Opera and not in Voyager. (This is Voyager from QNX 6.2.1, as I'm not going to bother with the 6.3 eval)
Since the part I'm talking about is the rendering engine, you can assume that I'm just talking about the "server" or back end as it were.
Voyager 6.3 may well have the problem sorted, but I really can't be bothered installing it. Last time ONX was updated the only way to upgrade was to wipe it clean and re-install, and I have absolutely no interest in doing that again. The next time QNX gets wiped it stays wiped.
Back on-topic though, CSS support is crucial and anything that does not support it is quite obsolete these days. Sadly, the Amiga mentality leads me to believe people would rather keep developing their own obsolete solutions rather than co-operate to produce a KHTML "server" component that all the various browsers will be able to plug into. |
|
| AWeb 3.5 APL Lite Beta released : Comment 46 of 82 | ANN.lu |
| Posted by greenboy on 22-Jun-2004 18:02 GMT | In reply to Comment 45 (Bill Hoggett):
>Precisely. The back end - the one doing the rendering - is different.
I thought I pretty much said that. If not, I certainly pointed at that, in my first post in the thread ; }
>Back on-topic though,
All that QNX stuff was actually on-topic - if one assumes we here could learn from other communities/ideologies. QNX has. AROS has. People here and there in this 'hood get it, and actually understand the beauty and progressive potential of making Open Source code easily accessible as well.
>CSS support is crucial and anything that does not support it is quite obsolete these days. Sadly, the Amiga mentality leads me to believe people would rather keep developing their own obsolete solutions rather than co-operate to produce a KHTML "server" component that all the various browsers will be able to plug into.
Perhaps you'd like to comment on dantheman's post. |
|
| AWeb 3.5 APL Lite Beta released : Comment 47 of 82 | ANN.lu |
| Posted by greenboy on 22-Jun-2004 18:04 GMT | In reply to Comment 45 (Bill Hoggett):
>Precisely. The back end - the one doing the rendering - is different.
I thought I pretty much said that. If not, I certainly pointed at that, in my first post in the thread ; }
>Back on-topic though,
All that QNX stuff was actually on-topic - if one assumes we here could learn from other communities/ideologies. QNX has. AROS has. People here and there in this 'hood get it, and actually understand the beauty and progressive potential of making Open Source code easily accessible as well.
>CSS support is crucial and anything that does not support it is quite obsolete these days. Sadly, the Amiga mentality leads me to believe people would rather keep developing their own obsolete solutions rather than co-operate to produce a KHTML "server" component that all the various browsers will be able to plug into.
Perhaps you'd like to comment on dantheman's post. |
|
| AWeb 3.5 APL Lite Beta released : Comment 48 of 82 | ANN.lu |
| Posted by greenboy on 22-Jun-2004 18:21 GMT | By the way, Henrik and the Aweb coders - good work : } Hopefully people around here begin to understand the OpenAweb interest in what is essentially two efforts. One I think is to supply classic Amigas with some improved browsing. And the other, which I think is much more important, is to get less platform locked when that can be advantageous. KHTML is a good choice.
I think some of the MOS/Amizilla people are pretty cool too, about doing something that maybe all can benefit from. They've expressed less interest in the "bounty" than they have that the bounty might have given some visibility and that has arguably attracted a few like-minded people to work together - "just because". Much in keeping with Henrik's mention of being altruistic, though MOS/Amizilla has to thank Bill Pangouleas for putting some fuel to spur this grouping initially. |
|
| AWeb 3.5 APL Lite Beta released : Comment 49 of 82 | ANN.lu |
| Posted by Bill Hoggett on 22-Jun-2004 18:40 GMT | In reply to Comment 47 (greenboy):
"I thought I pretty much said that. If not, I certainly pointed at that, in my first post in the thread ; }"
I think the confusion arose from what we referred to when we said QNX Voyager. I was talking about the server and you were talking about the client.
The QNX Voyager 6.2.1 server does not support CSS, or at least not properly.
"Perhaps you'd like to comment on dantheman's post."
So far it's all talk, and will remain so until we see actual results. I have to admit it sounds like a pile of excuses and BS to me, but I stand ready to be proved wrong by actual working available software. There have been a lot of people who have been big on "authoritative" talk about Amiga developments, but only a very small fraction of them have delivered anything at all, and of those most have delivered less than they said they would.
I'll pay attention to the Paihia team when their browser is available to the public, and not a day before. We've had vapour Amiga browsers before (no pun intended). From all that talk you'd expect a Firefox beater, and in all honesty I'm not convinced. So far all the proprietary custom built Amiga browsers have sucked compared to every other platform, so history is not on the Paihia team's side. |
|
| AWeb 3.5 APL Lite Beta released : Comment 50 of 82 | ANN.lu |
| Posted by greenboy on 22-Jun-2004 19:05 GMT | In reply to Comment 49 (Bill Hoggett):
Bill,
Pretty much my take as well. And I'd gladly be proven wrong, too.
The other problem I see is that commercially supporting and updating a totally proprietary solution historically (as you say) has not led to catching up or keeping up. It hasn't been any great financial incentive for the authors of the [non]current choices; has NOT kept them working on those browsers and with frequent update, to keep their users fully enabled to deal with the web as it actually exists today.
Why would another commercial effort in this 'hood have a better chance or more resolve to always actually stay in the same decade as the rest of the world as regards the web? I'd place no more faith in THAT than I would in the Open Source mindset of doing it "just because". In fact the OS motivations have proven to be very useful when the money can NOT be a practical answer.
I'm noticing that some of the work being done for post-BeOS browsing solutions has some common code and coding with QNX-oriented solutions, BTW. Seems hopeful. |
|
| Anonymous, there are 82 items in your selection [1 - 50] [51 - 82] |
|
- User Menu
-
- About ANN archives
- The ANN archives is powered by #AmigaZeux. It was updated daily (news last: 22-Oct-2004; comments last: 18-May-2005).
ANN.lu was created, previously owned and maintained by Christian Kemp, www.ckemp.com.
- Contribute
- Not possible at this time!
- Search ANN archives
- Advanced search
- Hosting
- ANN.lu was hosted by Dreamhost. Sign up through this link, mention "ckemp" as referrer and he will get a 10% commission on any account you purchase.
Please show your appreciation for any past, present and future work on ANN.lu by making a contribution via PayPal.
|
