...

View Full Version : what browser for testing



jeorg
10-28-2002, 07:49 AM
for testing my css2 I use mozilla , abd you ? what is for you the best browser for xhtml and css2 ?

and what do you think of Opera ? :-(((

thank you

jkd
10-28-2002, 11:09 AM
Originally posted by jeorg
for testing my css2 I use mozilla , abd you ? what is for you the best browser for xhtml and css2 ?

and what do you think of Opera ? :-(((

thank you

1. The same. Typically a recent nightly off the 1.2 branch currently.

2. Mozilla of course.

3. Opera 6 supports a surprising amount of CSS2 (more so than IE), but still lacks. (Can't style form elements, doesn't support XSLT, etc).

jeorg
10-28-2002, 12:21 PM
yes I was really surprised of Mozilla ! wondefull !

I use xhtml 1 transitional

and I have a lot of diffilculties to get what I want with the last Opera... it works for IE 6, netscape 6, and Mozilla


thank you

cg9com
10-28-2002, 03:16 PM
Originally posted by jkd
Can't style form elements

i hate that, opera annoys me sometimes
i like IE

brothercake
10-28-2002, 03:28 PM
For the record - it's not accurate to say Opera can't style form elements; it's more accurate to say Opera won't style form elements - this was a deliberate decision to prevent them from being able to be made to look like anything other than form elements.

jkd
10-28-2002, 03:51 PM
Originally posted by brothercake
For the record - it's not accurate to say Opera can't style form elements; it's more accurate to say Opera won't style form elements - this was a deliberate decision to prevent them from being able to be made to look like anything other than form elements.

Thanks for clearing that up. I almost always forget that it was an accessibility decision...

cg9com
10-28-2002, 05:44 PM
Originally posted by brothercake
this was a deliberate decision to prevent them from being able to be made to look like anything other than form elements.

why would they do that?

jeorg
10-28-2002, 06:45 PM
yes I cannot undestand such a decision too ! when everybody has allready wait too long for common standarts

at least CSS2 nearlly works every where, and with a good xhtml we have nearlly the feeling we knows what we are doing...

jeorg
10-28-2002, 06:49 PM
yes my english is a bit buggy too :-))

jkd
10-28-2002, 06:54 PM
Originally posted by cg9com


why would they do that?

One word: Accessibility

Though a user can always define their own stylesheet to keep form elements styled to the OS default...

But I use Mozilla, so I really don't care too much about why Opera decided to remove this or not. I have mouse gestures, tabbed browsing, and the best standards support anywhere - that's all I need. :)

jkd
10-28-2002, 06:55 PM
Originally posted by jeorg
at least CSS2 nearlly works every where

Ha! I wish. CSS1 works just about everywhere. Internet Explorer for Windows however is severely lacking in the CSS2 department, esp. with regards to selectors.

Bosko
10-28-2002, 06:57 PM
I use Mozilla,Konqueror and Opera for testing XHTML/CSS2 stuff.Theres is almost no diference though,unless some DOM stuff is involved which doesn't work in Opera 6 (Opera 7 will have DOM support).

jeorg
10-28-2002, 08:26 PM
but how do you get Konqueror ?? I have tried to download it !! and couldn't find the way

jkd
10-28-2002, 09:07 PM
It is a part of KDE3. You need to run XFree86, which is typically a *nix thing.

dauvm
10-28-2002, 09:37 PM
jkd, you have tabbed browsing and mouse gestures thanks to opera, who, as far as I know, introduced them as standard fare as far as popular browsers go.

I use opera for all my testing... stuff. Chances are really good that I would use mozilla, but it takes 10 times longer to start up than opera. Yea you can make mozilla cache itself at startup or something... but that's kinda cheating.

Opera may not support everrrything in standards, but they pretty much ONLY support standards, i.e. they don't put or support propriety code... which makes them good for testing. other features like turning off style sheets, tables, and images w/ a shortut key plus validating page code from the right click menu are all feature put there mostly for web developers use. It's also one of the only browsers I know that can display WML.

still... you have to pay to get the ads off it... which sucks. What I always do is test for the strictest browser, and THEN add things like styling form elements.
and I have a lot of diffilculties to get what I want with the last Opera... it works for IE 6, netscape 6, and Mozilla This is because those three browsers are bloated with support for poorly written code, whereas opera is not. :)

-Doug

zoobie
10-28-2002, 09:54 PM
Google's toolbar only uses IE...nuf said :D

realisis
10-28-2002, 10:02 PM
Can't comment on xhtml and css2, but regarding Opera, I've done a fair amount of testing with it.

While I like many of its features and generally how it renders web-pages, I think they really dropped the ball in a couple of important areas. These points really irk me:

1) Opera positively refuses to enumerate its built-in objects. Considering how their website mentions that only a small percentage of webpages take their browser into account, you figure they'd make it easy by allowing enumeration via for...in loops - what better way to sniff out the browser's DOM? Even Microsoft finally implemented enumeration with IE 5.0...

2) Opera response time to javascript in general, and document.write in particular is extremely sluggish: it takes about 8 seconds to doc.write a page in Opera which other browsers can deliver in about one second... Plus if you happen to mouse over a link while Opera is actively writing, the browser hangs for an additional 10 seconds or so, while it monitors the mouse-target. "Fastest Browser on Earth!" says Opera.

3) Opera still doesn't do DHTML: you can't dynamically modify page contents at all (other than repositioning elements) - and this a full 5 years after DHTML was introduced. Even NS4 in 1997 was ahead in that area (allowing you to at least overwrite layer contents independently). So doing anything after runtime is out.

4) doesn't support clip and overflow properties, or even tab-indexes, or even onDblclick. What price convenience?

5) though their "one-key" shortcuts are a good idea, as a user I really would like to be able to turn them off, or at least be able to reconfigure them.

6) lastly, I think Opera's ability to spoof other user-agents may be ill-advised: if a page is redirected based on UA, the browser ends up being fed statements which it can't process.

jkd
10-28-2002, 10:33 PM
Originally posted by zoobie
Google's toolbar only uses IE...nuf said :D

You can get one written in XUL for Mozilla. Look at www.mozdev.org for some nifty Mozilla projects.

Besides, using bookmark keywords, I just type this into my location bar whenever I want to search google:

google my search query

dauvm
10-28-2002, 10:40 PM
zoobie, opera come preconfigured with a google (and most others) search bar... 'nuf said ;)

realisis... those are all really good points, to which I only have a few comments. first, I am pretty sure that Opera is purposely not accepting dhtml because it is not a w3c standard. second, Opera's support forums on their website is their primary customer and user relations place and is monitored by opera staff all the time... I really think you should copy this message onto their forum also... give them something to chew on :) Opera 7 is due in not too long... and you might get some of the easier things to add, like disabling/changing 1key shortcuts, put in.

and lastly, operas user-agent spoofing is, in my opinion, well designed. It still contains "opera" in the string. It was designed to trick the old and outdated browser redirects that only accounted for IE and Netscape, while not being effected by newer ones that are well designed. A lot of people on the opera site complain this feature doesn't work, and opera people's answer to that was that it wasn't really designed to work on newer scripts, only the old and outdated ones. Konqueror has full user-agent spoofing...

-Doug

jkd
10-28-2002, 10:40 PM
Originally posted by dauvm
jkd, you have tabbed browsing and mouse gestures thanks to opera, who, as far as I know, introduced them as standard fare as far as popular browsers go.


Not entirely true. Tabbed browsing is completely unrelated to Opera's MDI. You still only have one window in Moz, while you have multiple, contained windows in Opera. Big difference as far as how the implementation works.

As for mouse gestures, Opera didn't originate them. They came from the olden days of GUI experiments, and I forgot which program implemented them, but years ago they could be found. Opera just made them popular :p.


Originally posted by dauvm

I use opera for all my testing... stuff. Chances are really good that I would use mozilla, but it takes 10 times longer to start up than opera. Yea you can make mozilla cache itself at startup or something... but that's kinda cheating.


It is what Internet Explorer does in Windows. And both IE6 and Mozilla 1.2 launch in 3 seconds for me (this is without the quick load enabled on Moz). Opera 6 was never much better or worse for me either.


Originally posted by dauvm
This is because those three browsers are bloated with support for poorly written code, whereas opera is not. :)
-Doug

Mozilla is just as strict as Opera, if not more so. If you've surfed the web with Opera, you must realize it can render a lot garbage.

dauvm
10-28-2002, 10:57 PM
I did mention this:
"... introduced them as standard fare as far as popular browsers go."

And this is funny, jkd. I am running a 2.2gHz p4 and 512 RAM... and when I first start windows, Mozilla will take maybe 6-7 seconds to load whereas Opera will take about 1 flat from a cold start. And yea, I know IE precaches itself... I never use it though... hmm... can I stop it from caching itself to improve startup time? anybody have a URL on how to do that?

And by opera loading "garbage" do you mean how it mangles poorly written code? Lets just say every time that a page looks like crap I use the validate feature, and it usually has about 50 or more errors in the code.

Ok ok... so opera isn't the say-all end-all of browsers by far... and i'm getting tired of defending it :), because I have a few problems w/ it too, namely frequent download file size mismatches. So here's the deal, i'll use mozilla for the rest of the week...long enough to figure out its features... and i'll see how I like it in comparison.

-Doug

realisis
10-29-2002, 01:56 AM
Hey doug, thanks for your input.

"I really think you should copy this message onto their forum also... give them something to chew on (...) "

Okay, we'll see what comes of it.

http://my.opera.com/forums/showthread.php?threadid=5052

***

"I am pretty sure that Opera is purposely not accepting dhtml because it is not a w3c standard"

Well that might be true if using Microsoft's .innerHTML, but something like document.createElement is surely part of the standard DOM?

Anyway, even if that is the excuse, it wouldn't be internally consistent on Opera's part.

Forinstance, Opera openly admits on their website that their method of capturing mouse/key events is non-standard. They also mention implementing IE's (non-standard) "extensions" to DOM via the .style object

No I think Opera's engine is just sorely lacking in its ability to modify document structure after runtime. They openly admit as much on their website, in their specs pages.

Here's a little example for you, and yet it's fully compliant AND Opera tries to implement it, but fails:

<style>
<!--
a { font:bold 10px serif; color:red; }
a:hover { font-size: 15px; }
-->
</style>

Try that out on a block-level link in a webpage and notice that although Opera tries to resize the link when hovering, the browser is unable to redimension the link's boundaries, so that the extreme right of the link is entirely cut-off. (invisible).

This is not a bug. Instead it's a consequence of Opera's inability to reformat page contents after layout.

*********

"and lastly, operas user-agent spoofing is, in my opinion, well designed. It still contains "opera" in the string. It was designed to trick the old and outdated browser redirects that only accounted for IE and Netscape, while not being effected by newer ones that are well designed. A lot of people on the opera site complain this feature doesn't work, and opera people's answer to that was that it wasn't really designed to work on newer scripts, only the old and outdated ones."

Taken in that light (which I had not considered) you make a good point. But I still see very few people even on this board taking Opera into account when sniffing for document.all (Opera responds to this, but may or may not respond to the execution of subsequent statements or may throw up errors *because* it passed a document.all test). T-R-O-U-B-L-E.

"Konqueror has full user-agent spoofing..."

I still think that's ill-advised though. Consider your recent request for wanting to sniff out NS4. If Konqueror passes a test like that, and is brought to NS4's page... then what? What has Konqueror achieved other than letting your visitor view a deliberately lacklustre page (presumably without the user even knowing it was really intended for NS4?), when there's really a shiny new XML page waiting in the wings which the Kong user never gets to see?

And worse: what if you had NS4-specific javascript statements on the NS4 page? Kong would we throwing up errors all day - and your visitor (who may not be web-savvy) might just conclude that YOU don't know how to code... and never come back to your site.

I mention NS4 only as a example - any browser could be substituted for illustrative purposes. Regardless of standards, browsers will always have little bits of proprietary behaviours (which I welcome, actually), and normally we sniff for these, but if another browser is spoofing UAs without being able to subsequently interpret statements intended for that UA, who gains?

jkd
10-29-2002, 02:10 AM
Originally posted by realisis
Here's a little example for you, and yet it's fully compliant AND Opera tries to implement it, but fails:

<style>
<!--
a { font:bold 10px serif; color:red; }
a:hover { font-size: 15px; }
-->
</style>

Try that out on a block-level link in a webpage and notice that although Opera tries to resize the link when hovering, the browser is unable to redimension the link's boundaries, so that the extreme right of the link is entirely cut-off. (invisible).

This is not a bug. Instead it's a consequence of Opera's inability to reformat page contents after layout.

You are correct. It is not a bug. Read the CSS specs. Browsers are not required to reflow content.

http://www.w3.org/TR/REC-CSS2/selector.html#dynamic-pseudo-classes


User agents are not required to reflow a currently displayed document due to pseudo-class transitions. For instance, a style sheet may specify that the 'font-size' of an :active link should be larger than that of an inactive link, but since this may cause letters to change position when the reader selects the link, a UA may ignore the corresponding style rule.

realisis
10-29-2002, 02:20 AM
"a UA may ignore the corresponding style rule"

Yeah, but Opera doesn't ignore the corresponding rule, it still tries to resize the link without actually reflowing any content... which makes the browser look bad. It'd be better if OP just ignored the resizing altogether?

dauvm
10-29-2002, 04:31 AM
you're right... tuché! well I'm becoming increasingly interested in what changes release 7 will bring for opera... I can't find anything on their site detailing a list of changes... too bad :(

i see your point about content resizing. This is actually interesting to me because I noticed this before when I was designing a holding page located at www.jamphorums.com
at this page I defined a link style like:
a:hover {border: 1px dotted #000}
I noticed in opera that it displayed the border around the link without readjusting for the 2 pixels of width it would add, unlike mozilla. I liked how opera did it better because it didn't readjust my other content and it looked more natural, but i now understand better the implications of that, although i can't say that i've ever seen opera cut text off.

btw, jkd... methinks i like mozilla :thumbsup:

-Doug

brothercake
10-29-2002, 10:21 AM
Originally posted by dauvm
Opera may not support everrrything in standards, but they pretty much ONLY support standards, i.e. they don't put or support propriety code... which makes them good for testing.


Well ... you know that isn't really true. Part of Opera's mandate is to try to support as much existing code as possible, to gain market share. They do support some of IE's proprietary objects and methods.

The real thing, I think, with form elements is that it prevents them from being disguised as something else - making it impossible to trick users into interacting with forms without knowing it's a form.

realisis - actually .innerHTML is not MS proprietary, it's a standard. Opera's lack of support for this and other things is frustrating, but not for much longer. Check out http://www.brothercake.com/dropdown/ for some interesting hacks that make Opera work.

jkd
10-29-2002, 11:32 AM
Originally posted by brothercake
realisis - actually .innerHTML is not MS proprietary, it's a standard. Opera's lack of support for this and other things is frustrating, but not for much longer. Check out http://www.brothercake.com/dropdown/ for some interesting hacks that make Opera work.

It is not a standard. It just happens to work in IE and Moz. BIG difference.

brothercake
10-29-2002, 12:03 PM
I stand corrected. So is there a standard method for writing in html, or is there only innerText?

Pooh
10-29-2002, 12:44 PM
innerHTML will work or element.firstChild.nodeValue

Nothing compares to IE...............The rest is just Crap.

Strict standards are like the constraints of an over intrusive Gubbermint on freedom, and creativity. We know how everone wanted to drive those cars imported from the old Soviet Union. That's why all the tofu-wink Patchuli-smelling Birkenstock-clad bed-wetting Socialist geek ABM crowd want adherence to some Soviet style standard rather than stuff that is cutting edge and works........ As long as Microsoft doesn't come up with it.....it's OK.
Hating Microsoft is a cult with clout. A religion with a lot of electric koolaid drinkers.

dauvm
10-29-2002, 01:45 PM
Pooh... you ought to be a rapper... lol that was great.... although I totally disagree with almost every you said ;)
Strict standards are the actual coders fighting BACK against the over intrusive "gubberment" of browser makers who have, for years forced us to use their code and write within the constraints of their creative. And you talk about freedom? Try getting microsoft to impliment YOUR brilliant idea for the betterment of the internet... much easier to work with the semi-democratic body which is the w3c I gaurrantee. Standards are a way to let coders code in ONE language to be viewed well by all who see it. Ask anyone who codes in a language that is standardized (php, python, the Cs etc etc etc) if they'd rather have 3 propriety versions. IMHO, if you find standards confining, it's because you haven't taken the time to learn about them yet.
***

Part of Opera's mandate is to try to support as much existing code as possible, to gain market share. They do support some of IE's proprietary objects and methods.
Hmm, that is really interesting, I did not know that.

Since the topic has come up anyways, might I get one of you to explain a little about innerHTML?

-Doug

Pooh
10-29-2002, 03:12 PM
Rappers are ditch carp DEMONCRATS


Strict standards are the actual coders fighting BACK against the over intrusive "gubberment" of browser makers who have, for years forced us to use their code and write within the constraints of their creative.

Except they haven't forced you do do anything. You have a choice to use any piece of crap browser out there and code for whatever browser you want. You may have to code for IE because most people choose to use a superior browser and you'd rather not eat alpo.


And you talk about freedom? Try getting microsoft to impliment YOUR brilliant idea for the betterment of the internet... much easier to work with the semi-democratic body which is the w3c I gaurrantee.

The W3C is like the UN.....It's a consortium of disgruntled pus chancres who hate Microsoft. In the UN's case, it's hate America. If Microsoft comes up with any innovative idea then they won't adopt it. They'll let someone else take the idea, morph it and then approve it. Btw----Microsoft is bettering the internet.


Standards are a way to let coders code in ONE language to be viewed well by all who see it. Ask anyone who codes in a language that is standardized (php, python, the Cs etc etc etc) if they'd rather have 3 propriety versions. IMHO, if you find standards confining, it's because you haven't taken the time to learn about them yet.


They wouldn't want three versions but it probably started out that way and one won out. Like Microsoft won out in the browser wars. Competition is a good thing........

There's nothing wrong with guidelines.....but having a bureaucratic gubberning body stifling and shooting down ideas before they get off the ground leaves you driving the Soviet made automobile. Tell me why innerHTML and outerHTML is not a standard?

btw----are you saying those languages php, python, etc are not evolving and or replacing other languages which were once 'standards'? In PHP's case, if that's standardized there's an ungrade or bug fix issued every other day. The web evolves, programming languages evolve, and 'standards' don't last too long.

dauvm
10-29-2002, 03:59 PM
The W3C is like the UN.....It's a consortium of disgruntled pus chancres who hate Microsoft.well not ALL w3c hates microsoft, considering microsoft is a member!

I don't hate microsoft... ....ok I do!*laughs* but why shouldn't I? Their products are over-price, over-rated, and security? Don't get me started...

Ok back on topic... you say it's alright to have guidelines...
"There's nothing wrong with guidelines.....but having a bureaucratic gubberning body stifling and shooting down ideas before they get off the ground..."
you sound really threatened by the w3c. They're not the bad guys trying to squash good ideas... in fact ALL they do is set guidelines. Is microsoft being persecuted for breaking the w3c's rules? I'd like to think not. The only reason the w3c's work is gaining so much momentum against the propriety based system that runs the internet now is because of people (like me) that welcome a change and try hard to improve the system we have now.

-Doug

dauvm
10-29-2002, 04:02 PM
Let me just clear some stuff up... I don't ACTUALLY "hate" microsoft... and not ALL of their stuff is crap... for instance I am not too familiar with it, but I've heard their new .NET stuff is a novel idea that is changing a lot of things... hurray!

-Doug

jeorg
10-29-2002, 04:17 PM
I am working with .NET since more than one year, beta 1, beta 2, release... it is incredible !!! the tool I was waiting for ! :)))

fo rthe browsers I like very much Mozilla, but sometimes you cannot see a website because Mozilla is too strict, the same site is perfect readable with IE... only a very small part of web sites is actually using standarts of W3, and Opera for me is totally OUT ! what we want is to surf, not to get an headhake with the fantasy of each one

jkd
10-29-2002, 04:29 PM
Originally posted by dauvm
Since the topic has come up anyways, might I get one of you to explain a little about innerHTML?

-Doug

Until DOM3, there really isn't a good way of invoking a browser's builtin DOM parser and XML serializer. That is primarily why innerHTML is so popular, and supported in Mozilla.

But why doesn't the standards have innerHTML? Simply because it is too specific, the DOM3 Load and Save provides a much superior way of running a string through the XML parser, and serializing a Node into a string of XML. With these interfaces, and a standards-compliant browser, you could implement your own innerWHATEVER properties by prototyping whatever interface.

I don't know why it wasn't considered until DOM3, but you can't change the past.

Bosko
10-29-2002, 05:41 PM
Originally posted by realisis
3) Opera still doesn't do DHTML: you can't dynamically modify page contents at all (other than repositioning elements) - and this a full 5 years after DHTML was introduced. Even NS4 in 1997 was ahead in that area (allowing you to at least overwrite layer contents independently). So doing anything after runtime is out.

Ehm,surf to www.dhtmlcentral.com and be amazed.All the cool DHTML works in Opera 6 too;resizing & moving of those boxes,dynamic menu's etc.So I don't understand how you can make such a silly statement.
Opera 6 doesn't support the DOM,but Opera 7 does.

brothercake
10-29-2002, 05:57 PM
realisis' statement is still correct - Opera 6 does not allow for any kind of dynamic element modification after load, other than very simple things like the dimensions of a DIV. You can't write into layers; you can't even resize an image.

The fact that some people have managed to make advanced dhtml work in Opera 6 is a tribute to their coding skills (including myself :D) ... or is it just bloodymindedness ...? I've always been of the "if you build it they will come" school of thought. I've supported Opera since O4, and continue to do so. One day it will all seem worth it :)

In its day, Netscape 4 was streets ahead of anything IE4 or its contempraries could do. But you should try coding DHTML for Opera 5 if you really want a challenge ...

jkd - that's interesting. You got any refs for that?

dauvm
10-29-2002, 06:22 PM
"Opera 6 doesn't support the DOM,but Opera 7 does."

Where are you finding specs about Op7? I've been lookin' for them... very interested to find out what the changes will be...

"But why doesn't the standards have innerHTML? Simply because it is too specific..."

That is a really good point. I read somewhere that the way that html changes all the time now would be like changing the english language every 5 years... you lose so much important information and history. The w3c's true goal is to give the internet a language that is brood enough to provide a stable base to always build upon, and never detract from. It must be a near impossibly tast to balance the "give it now!" attitude from developers against the true goal of allowing pages written in xml to still retain their ability to be displayed correctly many years down the road.

-Doug

Roy Sinclair
10-29-2002, 06:57 PM
Originally posted by brothercake
...
In its day, Netscape 4 was streets ahead of anything IE4 or its contempraries could do. ...

Not really true. Netscape 4 was released before IE 4 but the day IE 4 came out it lost the "most capable browser" title and the "most compliant" browser title. Netscape documented it's capabilities better at first but IE 4.0 was a much more capable browser.

Roy Sinclair
10-29-2002, 07:19 PM
Originally posted by jeorg
I am working with .NET since more than one year, beta 1, beta 2, release... it is incredible !!! the tool I was waiting for ! :)))

fo rthe browsers I like very much Mozilla, but sometimes you cannot see a website because Mozilla is too strict, the same site is perfect readable with IE... only a very small part of web sites is actually using standarts of W3, and Opera for me is totally OUT ! what we want is to surf, not to get an headhake with the fantasy of each one

I would bet that 90% of the sites you can't see in Mozilla that work fine in IE are the result of bad html code that IE incorrectly forgives written by developers who are too lazy to double check their code in aother browser much less run it through a validator.

jeorg
10-29-2002, 07:45 PM
" would bet that 90% of the sites you can't see in Mozilla that work fine in IE are the result of bad html code that IE incorrectly forgives written by developers who are too lazy to double check their code in aother browser much less run it through a validator."

>>> you are 100% right ! but you and me will not change the world, and everybody is allowed to do a simple web site without caring too much about the "perfection" of opera

IE is very permissive and for this reason it is the best browser, even if for me the best is Mozilla

brothercake
10-29-2002, 08:00 PM
Originally posted by Roy Sinclair
Netscape documented it's capabilities better at first but IE 4.0 was a much more capable browser.

I disagree. Netscape 4's <layer> setup is infinitely more powerful than anything IE4 could do. Admittedly, ie4 was a lot more stable and supported better CSS and javascript methods ... but those <layer>s ... they rocked.

dauvm
10-29-2002, 08:14 PM
"you and me will not change the world, and everybody is allowed to do a simple web site without caring too much about the "perfection" of opera"

That's very true, but how many people that write "simple" websites do you think pop open a text editor and go at it? Very few I'd imagine. More than ever people with very little html knowledge are writing webpages using WYSIWYG editors like frontpage and dreamweaver and things like geocities site designer etc... in those situations you can't expect people to know the difference between good code and bad... but is it too far fetched to expect the editor makers to?

-Doug

beetle
10-29-2002, 08:19 PM
Originally posted by dauvm
but is it too far fetched to expect the editor makers to?No! Check this out! (http://www.webstandards.org/act/campaign/dwtf/)

I can vouch for it, DWMX is awesome. All I need is DWMX and Textpad, and I'm ready to rock.

jkd
10-29-2002, 09:02 PM
Originally posted by brothercake
jkd - that's interesting. You got any refs for that?

You can reason out the innerHTML issue. Obviously innerHTML isn't the only markup language, and with XHTML, using innerHTML could possibly invalidate the document on the fly.

As for references on DOM3 Load and Save, it is in working draft status at W3C:
http://www.w3.org/TR/2002/WD-DOM-Level-3-LS-20020725/

It is kind of complicated, but when any browser has a rudimentary implementation of it, I intend to write a detailed tutorial on it. This provides ways of building XML documents from strings, remote locations, etc - all in a standardized fashion! It also allows you to serialize XML documents. Picture sending a string back to the server of the modifications made dynamically to the page.
This is very important for the future of DHTML, and since I tell everyone not to use innerHTML, I feel obligated to offer an alternative, and make it almost as simple as innerHTML.

Bosko
10-29-2002, 09:03 PM
Originally posted by dauvm
"Opera 6 doesn't support the DOM,but Opera 7 does."

Where are you finding specs about Op7? I've been lookin' for them... very interested to find out what the changes will be...


I read it in a couple of interview with the Opera 7 team.

beetle
10-29-2002, 09:26 PM
Here's my thoughts on the whole 'innerHTML' issue.

I used to think innerHTML was all that. I loved what it could do for my web pages, and heralded myself for the fantastic results.

Until, one day, I needed to make changes. Ugh. It took TOO long, and the data was all intermingled with the formatting and my DOM scripts couldn't properly interact with the written HTML and BLECH! Enuf! I had enough.

Since my abandonment of innerHTML, I have not yet come across an implementation of it that couldn't be duplicated with the proper DOM methods, and be more flexible to boot.

Roy Sinclair
10-29-2002, 09:30 PM
Originally posted by brothercake


I disagree. Netscape 4's <layer> setup is infinitely more powerful than anything IE4 could do. Admittedly, ie4 was a lot more stable and supported better CSS and javascript methods ... but those <layer>s ... they rocked.

Nope, nada. <layer> stinks as a tag and <ilayer> never did get fixed enough to make it very usable. The object model for Netscape 4 is incomplete and very poorly designed. IE 4 supported position:absolute and position:relative on most tags (usage was usually limited to <div> and <span> for compatability with Netscape), the innerHTML ability to rewrite the content, a heck of a lot more CSS and a reasonable selection of events on various objects. IE 4 also allowed updating non-positioned elements and reflowed the page after such an update, something that's only come to the Netscape world very recently. Netscape 4 is why I see the word Netscape and think Netscum, Netscam, Netscab and other less polite variants.

beetle
10-29-2002, 09:38 PM
******************************
Beetle's Deep Thought of the Day

Why are people in this forum talking about/discussing/arguing about what version 4 browsers DID. Does it really matter? Does a browser's history have any REAL relevance on it's capabilities today?

******************************

brothercake
10-29-2002, 09:58 PM
possibly not :o

But anyway ... it was the ability to go <layer onload .. that I really loved. It was so cool, and made up for everything else in my book


notices the original thread title

Ah yes. Well then I guess the answer is, test in the browsers you care about, based on the your visitors stats and your personal preferences. Code to standards, but make sure it works in IE first and foremost, because that's what everyone actually uses.

While doing which ... dream of a future when cross-browser issues are a thing of the past. In my learning PHP, I've become so happy with having a predictable environment, that browser issues just more and more irritate me. I guess we all have different ideas about why and what to do about it ... we can only hope that sense and logic prevails.

Mind you ... I don't think it will. I suspect that (as usual) the dollar will determine the reality, and we'll all carry on playing the game, while we privately pursue interests that the dollar hasn't poisoned yet.

jeorg
10-30-2002, 06:20 AM
but is it too far fetched to expect the editor makers to?

yes of course, even the very expensive Visual studio net doesn't close <img/> and use a few old html when you ask a xhtml compatibility... dreamweaver mx is a bit better now ....

for me I use editplus and only code....

jeorg
11-14-2002, 11:54 AM
now with OPERA 7 all works ! .... very well

Bosko
11-14-2002, 04:01 PM
yeah,Opera 7 is released.It now has full DOM 2 support.

jkd
11-14-2002, 04:03 PM
Originally posted by Bosko
yeah,Opera 7 is released.It now has full DOM 2 support.

I intend to experiment with it extensively this weekend, but I've heard it falls short (right behind Moz actually) on DOM2 support, but it's CSS2 support is supposedly better.

I intend to put up a definitive document with my tests sometime soon.

Bosko
11-14-2002, 04:06 PM
Originally posted by jkd


I intend to experiment with it extensively this weekend, but I've heard it falls short (right behind Moz actually) on DOM2 support, but it's CSS2 support is supposedly better.

I intend to put up a definitive document with my tests sometime soon.

Hmmm,it's just what I have heard.I tried it for a couple of minutes today at school and I noticed that most scripts at WebFx now work in Opera too.Ofcourse it's really hard to beat Mozilla in DOM support,but at least its much better than the previous versions.

brothercake
11-14-2002, 04:36 PM
Originally posted by jkd
I've heard it falls short (right behind Moz actually) on DOM2 support, but it's CSS2 support is supposedly better.


That's about the size of it. Very good, but not perfect DOM implementation; there are compromises there where supporting existing (proprietary) scripting is considered in order for Opera to gain market share over IE.

I wish I could offer more specific comments, but 90% of the discussion during beta was over my head :o

jkd
11-14-2002, 07:49 PM
After a few quick tests, any DOM2 they didn't mention they don't support.

This includes the immensely useful Traversal-Range interface. Despite not mentioning this on their DOM2 Style support, they also neglect to support the also very powerful getComputedStyle() method on the AbstractView interface (which they do implement as document.defaultView).

I haven't yet to even try out their DOM2 Events implementation, but since they only mention Events and UIEvents, I fear they don't include MutationEvents and even possibly HTMLEvents. More on this to come.

Also, one thing I am sorely missing is an equivalent to the CSS3 user-focus: normal property/value declaration. I've been doing neat stuff with this lately (like creating drop down menus without Javascript or DHTML for example while maintaining accessibility). Perhaps it is a little too much to ask for, but Mozilla has -moz-user-focus :D.
Other than that, initial CSS support seems excellent.

And the GUI is terrible! ughh

jeorg
11-14-2002, 09:42 PM
jkd you'r asking a lot :-))) I am allready happy when my site look nearlly normal with opera 7

jkd
11-14-2002, 10:36 PM
Originally posted by jeorg
jkd you'r asking a lot :-)))

But Mozilla offers just about everything I currently want. Therefore, if another browser doesn't offer it, obviously they didn't try hard enough ;).

brothercake
11-15-2002, 10:48 AM
Or perhaps they have different - dare I say more realistic - priorities. Being a poster-browser for standards support is not what Opera is about.

re - getComputedStyle() - i believe "currentStyle" will work

Kylena
04-13-2004, 04:03 AM
Supported Browsers (http://css.maxdesign.com.au/listamatic2/browser-support.htm)

liorean
04-17-2004, 02:04 AM
Have a look at op7.50 instead. It's miles ahead of op7.11, almost as much as op7 was an improvement on op6. It's CSS is still top notch, though it's falling behind when it comes to CSS3 as a trade off for being the best CSS2.1 browser. It's DOM is not at Mozilla's level yet, but it's getting close, and is now far ahead of KHTML/KJS in Konqueror and Safari, and Tasman in MSN/OSX.



EZ Archive Ads Plugin for vBulletin Copyright 2006 Computer Help Forum