I'm very interested in the work that Athena has done using Mozilla to do real-time updating of the client. Is there a way to do this without having to modify the Mozilla source code? I am interested in finding a standards-compliant way of doing this type of 'Real-Time HTML' in the browser. If you notice more and more the browser is kind of turning into a server, in which the HTML server itself can periodically check on to see how things are going and to give and request information. There are many diverse applications of real-time HTML, or any browser application that can dynamically update the nodes of a DOM tree: * chat clients * real-time paging * real-time collaboration

It even gets more interesting when you stop thinking of the nodes on a DOM tree as being HTML elements and rather think of them as being entire components, so that entire components can be moved back and forth from the server to the client. Suddenly, you can create JavaScript mobile-agents merely by having them be a node on the DOM tree, and then moving this node to a server or another browser! Talk about something similar to Java's JINI....

Thanks, Brad Neuberg VP of Technology, BaseSystem Inc. Steward at the OpenPortal Initiative, http://www.openportal.org OpenPortal, where websites become discussions, and discussions become websites

This may be the most important new feature in browsers in years. Hopefully, a standard way of doing this can be quickly agreed upon.

Right now with mozillazine (or slashdot or any other website with changing content): 1) you have to manually reload the page in order to get the latest changes 2) you don't know when to reload 3) you're reloading the _entire_ document even if one change is made instead of just the actual change If this solves these 3 problems then the web won't be such a big pain in the neck. :)

"I am interested in finding a standards-compliant way of doing this type of 'Real-Time HTML' in the browser."

Should be possible with JavaScript and an invisible Layer. First, store any info in the invisible layer through JavaScript variables. Then JavaScript fires off a timed sequence to refresh the layers content (using a CGI program or ColdFusion perhaps). If there are changes, then JavaScript re-aranges things as need be. The trick is to ensure that the Layer reload only contains the bare essentials needed (text only versions of the latest changes) not the whole page again.

Combine this with XUL and your on your way to creating some wack web-apps.

Wouldn't work in older browsers of course, but hey... thats what [CHOOSE Netscape] buttons are for.

This could really come in usefull for caching pages that one visits often. If the browser could figure out which pages a person views on a regular basis it could just grab the delta when the page was changed and keep the cached page up to date then the page would really load fast. The only problem i see is bandwidth.

bandwidth is going to be much less of a problem in the very near future. @ home and other companies like them are going to make it a non-issue via cable modems, etc. Once everyone with a cable hookup can get 50 times the speed of of a modern 56k connection for $40 a month, real-time is going to be standard. This sounds like a real tomorrow feature, and an absolute necessity for the Moz.

Despite what the article says, IE4 and IE5 have very flexible ways to automatically update portions of the DOM without reloading the entire document. Take a look at the "databinding" stuff on Microsoft's site.

Microsoft submitted this stuff to the w3c html committe, but it was squashed because Netscape did not have a chance in hell of implementing it at the time.

This is what java was made to do. In Java a multithreaded tcp/ip server is easy. The rendering engine should be written in java and provided to the programmer as an API. Both renderers are used the same and you can setup any kind of client/server or better - peer/peer relationship.

The problem seems to be the FIREWALL more than the protocol per-se, but I suppose a wolf in sheeps clothing.. Anyway, we need to make people turn off their stupid firewalls. If everyone didn't use windows, firewalls wouldn't seem so important - nor viruses, but I digress.

I know I've seen a java, peering chat server out there, and jazilla, of course, as well as other java html rendering projects. Don't have the URLs..

-=Julian=-

You can use java through firewalls. It's just that everyone opens up port 80 on their firewall, if they had another standardized port for this sort of network communication protocol, then the java applets would work just fine (It's not even a java problem!)

Some many comments to reply too. First, I don't see cable modems taking over. Say you live in a subdivision w/ 100 homes and everyone has cable. Now say everyone gets a cable modem. Those 100 homes will be feeding off each others bandwidth. DSL offers more, but currently its too expensive. AOL is going to offer a special version of their product with special DSL deal ($40/month for 256 Kbs connect). Its a start plus a DSL can be used by a computer and a phone at the same time and only requires a small adapter(not quite sure).

Second, databinding was rejected because it was a poorly written purposal/standard. If this DDR is written better then the W3C might accually recommend it.

Third, Java wasn't meant to do this. It can do this but it was orginally targetted for making cross platform apps.

Now enough of me picking on other peoples comments.

I think DDR may be a better than any current way of doing this for one simple reason, it will be part of the DOM. All you need to know ishow to access the DOM, which is easier than learning Java.

I know in Seattle by March 2000 over half of the home are going to have access to cable modems. and my point was not that cable modems would take over the world, but that Net infrastructure is quickly coming out of the mud and that bandwidth does will not be a problem for a large portion of us for much longer. AOL is also going to offer access through @ Home in Seattle, at least. There is also DSL and other methods coming into the fray and with a combination of them, I think by 2001 56k will be considered slow.

forgive my grammar botchings, I'm in a hurry. Also, I don't think with dual fiber-optic cable that the subdivision thing is going to be a problem, but granted, they have to lay the cable first.

When, i stayed in Rochester, NY for a few months I was able to get RoadRunner. It was great the entire time I had it, but my neighbors started getting it too. well, I have a pretty good Idea of what they were doing because my connection speed dropped considerably, mind you that it was still fast (~300K) but I used to get it up to 1 Mbs. I think it might have been the poor wiring of the area, it was a dump with lead paint, but Time Warner said that was normal for cable modems.

Yes, I agree that 56k will be called really slow by 2000. Its slow now. Damn I wish I had my ethernet connect from college.

It's a very sad day when a third party startup has to implement something in Mozilla for it to be competitive with what IE4 had almost 2 years ago.

I am NOT a Microsoft fan. But the fact of the matter is, I was able to do DDR one year ago with IE4 and JavaScript or Java.

Using Java, I can even do it in real time by using Java to poll for external data and then automatically update the DOM tree.

Microsoft also has a "data binding" spec which allows HTML to be bounded to data sources and change in real time.

What I see in this community is a bunch of closed eyes. People continue to bash MS, say its code sucks, with eyes closed, cheering open-source slogans, and totally ignoring the fact that they need to get their shit together and code something better.

Basher, I think you're missing the point. Microsoft proposed it as a standard and the W3C totally rejected it. Obviously, it had some major problems.

I haven't used DDR, but if it turns out that this exceptional well implemented and thoughtout, more so than databinding then it might become a standard. If not they too bad.

MS code does suck, here is why: IE and NS have basically the same functionality. Now compare their size. Now compare their size against Mozilla. There is a big different. IE is about 2x of NS and much larger compared to Mozilla.

Compare Windows CE to Windows 98. Win 98 is about 150 MBs for an average install. CE is like a few Mbs, granted it doesn't do a lot but comparing size Win CE is more effecient. Windows 9x is also written in mostly Ada, back from Windows 1.

Compare WordPerfect Suite 8 w/ SP5 to Office 97 and 2000. WordPerfect comes on one CD with quite a bit of space for other crap like fonts and bitmaps. I think Office 97 has 5 CDs and I heard from News.com the 2000 will have 7, it better not need Gbs of space to install. What do all of those CDs have on it to constitue that much space.

I'll admit MS do make good products once in a while, I love Outlook 98. In fact if you read this group a lot of people here use IE and MS stuff. I code in Visual C++, I think its terrible but I have to for what I'm doing.

Yes, Windows is a bloated nightmare, and so is IE, and so is Office 2000. I have the beta and it took me like half an hour to install just the basics. And the point is not that you can do something--eventually--with IE, but that you can do it ONLY with IE. Anyone lunatic enough to program for only IE will not be programming for NS users, and that is not good business. It won't happen.

Did you hear they're releasing not only Win98 SE but also another 9x version next year, a 32bit Win2000 and then a 64bit Win2000 a bit later, and THEN a couple years later a non-DOS consumer OS?! Jesus, there's going to be Windows OSs coming out of our EARS!

Then, Dell announced it is going to offer Linux desktops, and the next gen. Amigas OE is scheduled for beta by quarter3 1999 and for release by q4 1999. Please God let there be competition for MS in this, please oh please. They would not be such a bad company if they weren't so big. Do you think a company that really relied on Office would not be developing for Linux if they weren't worried about Windows supremacy?

Do you think Win98 would be more stable, and smaller, without IE clogging it up with over 27MB of "slimmed down" space? I do. Come on, maybe you don't care about your hard drive but I do. And as for IE5 being a shade faster, that is not going to matter in a world with higher speed connections. And you want to tell me IE can seriously run a hand-held at that size?

As for your whining about a third party adding to Mozilla, ISN'T THAT THE POINT? HELLO?! To involve talent outside the parent company? The thing about MSs talent is that it's homogenuous, everybody is in the same environment--like inbreeding, especially as closed as they are.

Proprietary gibberish killed Commodore and it will hurt MS if they're not smart--and MS has already said they are thinking of licensing their code to certain individuals. MS isn't stupid. Don't you be either.

Netscape blocked the Microsoft submission to the w3c because Netscape could not implement it. It was not rejected because the submission did not have merit.

So, it was rejected. What should Microsoft do? Intranet customers found it useful. Should Microsoft have sat on it or should it have met the needs of customers?

Microsoft elected to meet the needs of a customer group, and now you dump all over them. If you don't understand why delivering value to customers is good, then you probably don't understand why IE is beating Nav.

Thats pure crap. Data Binding was submitted to the W3C. It was reject because ti depended on the use of ActiveX Data Objects (ADO). The W3C liked the idea of data binding but decide not to continue with the idea, they wanted more proposals that avoided the use of ActiveX. Theey want it to be part of the DOM.

The idea has merit, it was the way to implement it that didn't. You can do that now with Java.

Its a nice feature so it wasn't made a standard that doesn't mean it was useful. Many things that aren't standards have uses. As you said MS found a use for it good for them. I think Ms should have released their implementation of it, like they did, and rework their proposal.

I'm not dumping all over them because of that. There are other and more important reasons for that.

Oh please NS still hold the majority of the browser market. If you include AOL's native browser into that IE and NS are almost tied. And when ever AOL switches to mozilla then netscape gains 20+% more users and IE loses that much.

yep, and the most new users IE sees are IE4 users--meaning the percentage won't grow much. MS leveraged Windows to get IE where it is now--PERIOD. They've done about as much of that as they can, and now it will be NSs turn. Don't worry, AOL will include Moz with AOL 5.0--but not until then. If they did it today NS would be back up into the 70s share.

The best counter argument that you can come up with is that MS apps are under some metric "bloated"?

Netscape4.5 is no slim-chicken either, and IE5, without Netshow/ComicChat/Java/Windows Media Player/etc is < 10mb.

Office2000 may be large, but Office2000 offers more functionality than *any* other office package, and, in addition to supporting everyone else's file formats, it now supports saving as HTML/CSS/XML.

Nevertheless, face the facts: IE4 had CSS, XML, PICS, PNG, OSD, and Java1.1 (minus RMI and JNI) support way before Netscape did. It had dynamic styles and object model access too. It also had the coveted incremental layout that the Gecko engine is laying claim to.

And look at IE5: Support for XSL, Scripts, Behavioral Style Sheets, Voyager, Namespaces, etc. For instance, CSS behaviors are a good idea that should be copied. (Yes, Netscape has their own Action Style Sheets proposal, but it is semanatically identical with inferior syntax and way behind in deployment. Why not just support MS's? Too much Not-invented-here)

I am still a Netscape user, but I am sick of tired of people defending Netscape. They fucked up. Bad management, bad execution. They sat on a 3 year old decaying code base. They stopped innovating. They changed their focus from browser, to server, and then to application servers, and then finally to portals.

I really can't fault MS for developing a superior browser. It is patently obvious that IE4's rendering engine and object model support is superior to Netscape 4's.

Mozilla may be the answer, but all I'm saying is, you can't just IGNORE MS and say their code sucks and leave it at that. You have to be willing to admit that once in a while they have a good idea, and have the guts to copy it, just like they have copied everyone else.

Anything else is just irrational immature behavior.

I've admitted that MS does make a few good products. Add IE currently supports more than Netscape.

I was comparing the average installs for NS 4 and IE 4/5. If you want to go to the smallest possible size then the stand-alone of NS is less than < 10 MB too.

WordPerfect 9 (part of WordPerfect Office 2000) suppports tons more file formats than Office does. In WordPerfect 8 I can open up WordStar 1.0 documents and even Word 1.0 docs. When, I was using my friends computer and Office 95 it couldn't even read Word 1.0 docs correctly, it hashed up crap on the screen. WP is also adding support for XML in this release, it has supported HTML for a while.

NS4 has CSS, PICS, and PNG support and it came out before IE4.

You said Ie5 supports namespaces and it does, sort of. There was an article on News.com I believe about the poor way it was implemented. MS co-authored the spec with the W3C yet they didn't support it, in fact it was said it was totally wrong.

As for code bases, everyone suffers from this. IE 5 has legacy code from all the previous IEs as does NS. Word is in the same boat as is Windows and MacOS. Developers don't have a lot of time before release dates to do a serious rewrite or write the code with foresight.

I think its not the " Not-invented here" thingie but Invented by Ms thingie. MS has lock out people before who followed MS lead. DR-Dos for example couldn't run Windows 3.x well. Betas for that Windows used to say DR-DOS may crash this system. NS and Marmbia(SP?) were going to support CDF, but MS changed CDF slightly so that neither company could do it without doing a major rewrite.

Nobody defended Netscape. But I'M sick of people defending Microsoft. Anything good they do is ruined by their evil tactics and it is hard to give them credit for ANYTHING. Prick #@$@#$# Steve Ballmer screaming "WHAT THE WORLD NEEDS IS MORE, MORE, MORE, MORE WINDOWS!!." and Gates asking AOL how much he could pay them to screw Netscape. I honestly don't care what the hell they do. When they don't have competition they don't have any incentive to do anything right.

And superior to NS 4 IE may be in SOME WAYS, but superior to NS 5 it most definitely will NOT be. I don't care how bad they fucked up, they eventually made the decision FINALLY to can the old code and go with the Moz AT THE EXPENSE OF RELEASING 5.0 EARLIER THAN IE5. Why you or JWZ thinks WHINING about it now does any good is beyond me. Who cares what they DID it's what they're DOING.

And actually, Office2000 doesn't do that much more that I need than Office 97 and Frontpage 98, which I choose to use over 2000 because it does everything I need it to and it fits on TWO CDs and takes five minutes to install both. It certainly doesn't do anything worth SEVEN CDs and more than a half an hour to install--and uninstall.

And YES, Windows is BLOATED. I don't want to be able to surf the web from my Windows Explorer. It is not an OS function, not a function I want, and I can't get rid of it because fucking with Windows is like slitting your own tires. They make everything unnecesarily complicated so that you have to pay THEIR sorry asses to fix it. And WORSE, they don't even make sure their software is DONE before they ship it. They throw something out, let everybody else find the bugs, and then release a service pack six months later.

Sorry, I use MS products for now while I have to, and they have some good programmers and YES, even some good programs. But if they don't want to be ignored they can FEEL FREE TO CONTRIBUTE TO MOZILLA. They'll certainly make use of it after it comes out, and try to patent it at that, I bet.

If you want to harp on Netscape's past mistakes, fine. It's their future that matters.

I would like to see this added to the Mozilla code, but not at the expense of shipping Navigator 5 as soon as possible (say 5.0.1).

MS didn't build a better browser. the built a MS-centric, buggy, **partially** compliant browser. They added features and impimantations that no one asked for, or required, and got the stuff people wanted bass-ackwards. They initially FORCED web-designers who wanted to add the "GET IE Now" buttons to use IE proprietary tags. The dicked aroiund with Java to make it more "Windows Freindly" (should have gone the other way round, but that's a whole different rant). IE5 isn't much better. instead of getting the currently approved standerdisation down right, they added more non-approved, non-standard crap, made half-assed attempts at fixing the buggy implimentation that was already there; then started pointing fingers and getting all hot under the collar when people called them on it.

As For Netscape, you right, they screwed up on alot of stuff (the complete divergence in CSS/DHTML impementation from initial IE3 for one), but this isn't Netscape; it's Mozilla. This is attempting to fix the problem. IE5 fixed squat, it furthered the problem.

I'm not quite sure what Office2000 has to do with things, but lets keep in mind it is all that extras functionality that allows for fun little diversions like the Mellisa Macro virus to make our Windows experience oh so Fun (personally wasn't affects, don't use Outlook, or Word).

They've been adding features for features sake, not for consumer need.

Their inintial driving goal was to have a computer in every office and household. and have Misrosoft Software running on it.

Now they want ONLY Microsoft software running. They released NT as a server platform, even though it (STILL) has massive security holes. They keep trying to sell Access as an enterprise DB. When they found out that VBA Macros could make viruses, they did little to stop that*. They got into the browser war, not to make a better product, but to kill the lizard; plain and simple. And that, my friend, is why we hate them.

* re: Macro Virus Security; one checkbox, deep within the app preferences doesn't count. I want dialogs popping up telling me that Macros are editing files in the background, and e-mailing them tomy 50 best friends; and i want those dialog to let me stop it if I want. Yes that functionality might be handy, if I had a monthly, reader customizable, newsletter to mail out (which I don't); but I still want to know its happening.