I agree with your main point that this is competition. That being said the "nasty proprietary features" were incredible. Heck I'd love some of the features of I.E. 4 even now.
1) Almost native speed applications delivered over the web fully integrated. (And yes I understand the security reasons this doesn't exist anymore but it was really really nice before abuses).
2) Push technologies with full featured sites. Especially for things like cell phones this would be terrific. I can imagine youtube subscriptions using push / caching being far better than the experience we have today.
3) Any arbitrary HTML attached to any folder, essentially a folder programable desktop.
4) Speeds ups of important plugins by like 400% / yr. For example Microsoft's Java.
The browser wars were awesome. I was very sorry Netscape lost so quickly.
I had to use FF2 recently for a very old site. And I commented to friends how wonderfully speedy and light the browser felt.
Its going to be several years till the performance complaint dies with respect to FF. People like me who switched away because of performance aren't going to look at the performance again until we have to switch browsers for some other reason or have to use FF regularly and notice the change.
That's the general policy for all technology. As they get better the focus moves from features to reliability to price to convenience. At this point browsers render most stuff correctly, so the focus is on convenience.
Actually Macs were losing popularity at the time when ActiveX was dying. From 1997-2003 Mac market share was in the 2-7% range of desktop users. ActiveX was at its most popular when Netscape and AOL's homegrown browser were still being used quite a bit. When ActiveX died Firefox / Mozilla / Netscape was not a player.
ActiveX was popular because it was excellent in terms of functionality.
When changed the market share was the higher default security setting on Windows needed for any sense of security at all.
There really was no concept of security in the early 1990s. You just didn't use PCs for stuff that required security. PC Networking was all about LANs, Microsoft was focused in the networking direction on Microsoft LAN Manager. Winsock had been intended for BBS users, or commercial BBSes. Security wasn't an OS function at the time.
But all of Microsoft's plans sort of came together and their Network operating system became the dominant (almost exclusive) desktop client for home and corporate. I can understand why they didn't take security seriously until the crisis. What I can't understand was how they reacted once we were in a world of ActiveX exploits, popups, spam... They had built the Windows NT system to be secure. They could have ramed through a major security upgrade. And they had the leadership to have done a lot for spam email as well.
I'm listening to the 13:00 mark. Great speech, thanks for the link! It sounds to me like he's saying Microsoft did participate rather actively and then the company as a whole moved away from the focus on web applications (not surprisingly their focus was ActiveX and then "yeah back to rich clients!").
I think I can standby what I said which is that Microsoft had a moderate interest in JavaScript, not particularly hostile and somewhat supportive. It sounds like the main thing was a loss in 1999 of the vote for a major rewrite. Obviously in 1999 Microsoft had over 1000 people on the IE team during the days of 6. They wanted to get done that browser and slash the budget; those were US employees so you are talking $100m+ / yr; for a product they gave away for free.
I just don't see particular malevolence other than Microsoft's general hostility to the web, which I will grant.
It would still look like that to the customer. Just Oracle would be paying RedHat for 2nd or 3rd tier support on OEL. The customer gets a unified support experience by RedHat is doing the OS work. Besides they don't sell PC hardware so that' the very bottom layer of the stack easy to drop or do it 1/2 way or rebrand or...
I don't think Oracle wants to get into making an OS. Hell they don't even want to do much with Solaris.
That is a capability system. Application X has permission to do Y to Z, rather than Application X has permission to do Y. In which case you no longer need any sandboxing, just use capabilities.
I can understand the advantage of giving HTML higher permission for local. What I can't understand is why HTML5 is considered a good choice for applications.
iOS App store -- SDK, ability to create provisioning files, paying Apple for their cost to certify your applications OSX App store -- paying for advertising or something?
I'm having a hard time seeing where Oracle isn't multipolar. Their absolutely core technology is a database. All their business offerings on the next layer generally support databases other than Oracle. Oracle is usable by business products that conflict with their offerings. Going to their Sun acquisition it gives them a hardware platform they can control. The ability to buy an "Oracle box" which Oracle is responsible for maintaining, top to bottom.
As for OpenOffice I'm not sure how that fits with Oracle's model at all, it is a Sun asset they can't really make use of. MySQL they seem to be protecting fine keeping it focused on the low end, along with Berkley DB, which is also theirs.
Oracle Linux is silly. I think Oracle will likely start licensing RedHat as it gets more difficult to support. Once they start writing checks their problems with RedHat will be over.
When ActiveX was popular IE was very very dominant. Platforms weren't the problem with ActiveX, no one cared too much about the other platforms.
As for sandbox, you always have the key problem: If the application is sandbox by itself you can't have platform interactivity. If the application is sandboxed with other applications then effectively the stuff the user cares about is inside not outside the sandbox.
The security problems on Microsoft ware not an inability to sandbox (NT 3.51 was excellent there) but an inability to have the sandbox small enough by default.
And these "apps" are HTML5? So there is an HTML5 for the web and an extended HTML5 for installed applications? Sort of like Microsoft gadgets or Apple's Dashboard as the primary application language?
OK we agree on most stuff. Your revised version makes sense and my "Microsoft better performance" is an example of the "slow startup" issue you mentioned with respect to Netscape Navigator.
Your dates on Java are wrong, the HotJava browser (supported applets) was '94. I was using Java in Mosaic in '94. The Microsoft JVM which was 3rd generation I believe was in I.E. 3 which came out Aug '96. Scripting engine (at least on IE) was part of the NT 4 Options pack and so later. I agree not by much.
As for applets being influential I think the influence was ActiveX, which was also in I.E 3.0. Microsoft at that point was following a dual strategy of trying to implement Java at all different levels of the Windows stack including a compiled version. ActiveX was marketed as a substitute for applets, but one with a much more mature API. And ActiveX was incredibly heavily used. It was well on its way to being the dominant application platform were it not for the security problems and Microsoft's general lack of enthusiasm for the web once it no longer was a strategic threat.
As for Java not living up to the hype. I agree it became the COBOL of the 1990s, the days of internet appliance with Oracle and Sun were still quite a bit off. But I do think Java forced progress in that direction. It prevented Microsoft from establishing the same sort of strangle hold on server based applications it had on desktop applications at a time when the big Unixes and the last of the mini computer systems (AS400, OpenVMS) were failing. The Web is not filled with ActiveX, Microsoft proprietary protocols today in part because of Java (though LAMP is likely the most important reason).
We are talking about the web so lets give an example, the end users for webservers, "webmasters" the earliest web developers were the people that developed CGI. They needed scripting but didn't want to write full fledged network applications.
Another example is macro languages in Excel. Accountants do quite a bit of programming. Another example are creative production languages and automated document processing. End users script.
Adobe made this worse by selling Flash to advertisers and not offering controls like "disable this plugin". I agree that HTML5 will have the same problems. But... it is open which means browsers, that have no particular interest in marketing, will be implementing the balance between end users and advertisers.
As for Adobe dropping Flash... they likely will push the technology into other products. They still offer a great vector manipulation system.
1) I don't think Flash killed applets. The Microsoft / Sun war did. Java under Netscape was miserably slow. Microsoft was creating a faster usable Java but adding proprietary extensions killing "write once, run anywhere"; Sun consider that the core feature and certainly didn't want to allow for Windows lock in....
Because most people used I.E. 6 and I.E. used an old Java and a bad Java the "run anywhere" started to fall apart along with the speed. Sun got neither and Java became server only. Flash never had that problem since a) Flash always worked best on Windows browsers. b) Flash had lower ambitions that didn't threaten Microsoft
2) I think applets were a major success. They broke the hold of HTML/CGI. Prior to applets there was no way to even having something like a variable menu. The web was static information + forms + response. It simply didn't offer anything like interactivity. Think about something like.pdfs today.
The hardware differences go beyond just screen size. Moreover the software differences go further. Apple wants people to use Cocoa and thus have integration with desktop versions. Google isn't pushing that.
Why would you do this? You can just as easily write the code on an old Mac and compile on a server. Apple has all kinds of great tools for using bad computers and fast servers for a compile so people can share.
Slashdot Mirror
I agree with your main point that this is competition. That being said the "nasty proprietary features" were incredible. Heck I'd love some of the features of I.E. 4 even now.
1) Almost native speed applications delivered over the web fully integrated. (And yes I understand the security reasons this doesn't exist anymore but it was really really nice before abuses).
2) Push technologies with full featured sites. Especially for things like cell phones this would be terrific. I can imagine youtube subscriptions using push / caching being far better than the experience we have today.
3) Any arbitrary HTML attached to any folder, essentially a folder programable desktop.
4) Speeds ups of important plugins by like 400% / yr. For example Microsoft's Java.
The browser wars were awesome. I was very sorry Netscape lost so quickly.
I had to use FF2 recently for a very old site. And I commented to friends how wonderfully speedy and light the browser felt.
Its going to be several years till the performance complaint dies with respect to FF. People like me who switched away because of performance aren't going to look at the performance again until we have to switch browsers for some other reason or have to use FF regularly and notice the change.
Memory leaking is why I switched away from Safari to Firefox and then from Firefox to Chrome. So throw me in that group.
That's the general policy for all technology. As they get better the focus moves from features to reliability to price to convenience. At this point browsers render most stuff correctly, so the focus is on convenience.
Actually Macs were losing popularity at the time when ActiveX was dying. From 1997-2003 Mac market share was in the 2-7% range of desktop users. ActiveX was at its most popular when Netscape and AOL's homegrown browser were still being used quite a bit. When ActiveX died Firefox / Mozilla / Netscape was not a player.
ActiveX was popular because it was excellent in terms of functionality.
When changed the market share was the higher default security setting on Windows needed for any sense of security at all.
Why would Oracle want to bleed RedHat which makes the distribution their Linux is based on?
There really was no concept of security in the early 1990s. You just didn't use PCs for stuff that required security. PC Networking was all about LANs, Microsoft was focused in the networking direction on Microsoft LAN Manager. Winsock had been intended for BBS users, or commercial BBSes. Security wasn't an OS function at the time.
But all of Microsoft's plans sort of came together and their Network operating system became the dominant (almost exclusive) desktop client for home and corporate. I can understand why they didn't take security seriously until the crisis. What I can't understand was how they reacted once we were in a world of ActiveX exploits, popups, spam... They had built the Windows NT system to be secure. They could have ramed through a major security upgrade. And they had the leadership to have done a lot for spam email as well.
I'm listening to the 13:00 mark. Great speech, thanks for the link! It sounds to me like he's saying Microsoft did participate rather actively and then the company as a whole moved away from the focus on web applications (not surprisingly their focus was ActiveX and then "yeah back to rich clients!").
I think I can standby what I said which is that Microsoft had a moderate interest in JavaScript, not particularly hostile and somewhat supportive. It sounds like the main thing was a loss in 1999 of the vote for a major rewrite. Obviously in 1999 Microsoft had over 1000 people on the IE team during the days of 6. They wanted to get done that browser and slash the budget; those were US employees so you are talking $100m+ / yr; for a product they gave away for free.
I just don't see particular malevolence other than Microsoft's general hostility to the web, which I will grant.
It would still look like that to the customer. Just Oracle would be paying RedHat for 2nd or 3rd tier support on OEL. The customer gets a unified support experience by RedHat is doing the OS work. Besides they don't sell PC hardware so that' the very bottom layer of the stack easy to drop or do it 1/2 way or rebrand or...
I don't think Oracle wants to get into making an OS. Hell they don't even want to do much with Solaris.
That is a capability system. Application X has permission to do Y to Z, rather than Application X has permission to do Y. In which case you no longer need any sandboxing, just use capabilities.
How many people want to rent access to a mac and don't own or are unwilling to buy a cheap one? I have a hard time imagining an actual paying client.
I can understand the advantage of giving HTML higher permission for local. What I can't understand is why HTML5 is considered a good choice for applications.
It is essentially a different service for now.
iOS App store -- SDK, ability to create provisioning files, paying Apple for their cost to certify your applications
OSX App store -- paying for advertising or something?
I'm having a hard time seeing where Oracle isn't multipolar. Their absolutely core technology is a database. All their business offerings on the next layer generally support databases other than Oracle. Oracle is usable by business products that conflict with their offerings. Going to their Sun acquisition it gives them a hardware platform they can control. The ability to buy an "Oracle box" which Oracle is responsible for maintaining, top to bottom.
As for OpenOffice I'm not sure how that fits with Oracle's model at all, it is a Sun asset they can't really make use of. MySQL they seem to be protecting fine keeping it focused on the low end, along with Berkley DB, which is also theirs.
Oracle Linux is silly. I think Oracle will likely start licensing RedHat as it gets more difficult to support. Once they start writing checks their problems with RedHat will be over.
I don't agree with the author.
When ActiveX was popular IE was very very dominant. Platforms weren't the problem with ActiveX, no one cared too much about the other platforms.
As for sandbox, you always have the key problem:
If the application is sandbox by itself you can't have platform interactivity.
If the application is sandboxed with other applications then effectively the stuff the user cares about is inside not outside the sandbox.
The security problems on Microsoft ware not an inability to sandbox (NT 3.51 was excellent there) but an inability to have the sandbox small enough by default.
And these "apps" are HTML5? So there is an HTML5 for the web and an extended HTML5 for installed applications? Sort of like Microsoft gadgets or Apple's Dashboard as the primary application language?
Why?
OK we agree on most stuff. Your revised version makes sense and my "Microsoft better performance" is an example of the "slow startup" issue you mentioned with respect to Netscape Navigator.
Your dates on Java are wrong, the HotJava browser (supported applets) was '94. I was using Java in Mosaic in '94. The Microsoft JVM which was 3rd generation I believe was in I.E. 3 which came out Aug '96. Scripting engine (at least on IE) was part of the NT 4 Options pack and so later. I agree not by much.
As for applets being influential I think the influence was ActiveX, which was also in I.E 3.0. Microsoft at that point was following a dual strategy of trying to implement Java at all different levels of the Windows stack including a compiled version. ActiveX was marketed as a substitute for applets, but one with a much more mature API. And ActiveX was incredibly heavily used. It was well on its way to being the dominant application platform were it not for the security problems and Microsoft's general lack of enthusiasm for the web once it no longer was a strategic threat.
As for Java not living up to the hype. I agree it became the COBOL of the 1990s, the days of internet appliance with Oracle and Sun were still quite a bit off. But I do think Java forced progress in that direction. It prevented Microsoft from establishing the same sort of strangle hold on server based applications it had on desktop applications at a time when the big Unixes and the last of the mini computer systems (AS400, OpenVMS) were failing. The Web is not filled with ActiveX, Microsoft proprietary protocols today in part because of Java (though LAMP is likely the most important reason).
Sure they do.
We are talking about the web so lets give an example, the end users for webservers, "webmasters" the earliest web developers were the people that developed CGI. They needed scripting but didn't want to write full fledged network applications.
Another example is macro languages in Excel. Accountants do quite a bit of programming. Another example are creative production languages and automated document processing. End users script.
OK Flashpaper.
The ability to view a large document one page at a time rendered correctly. Essentially combing the web's ability to surf with .pdf accuracy.
Adobe made this worse by selling Flash to advertisers and not offering controls like "disable this plugin". I agree that HTML5 will have the same problems. But... it is open which means browsers, that have no particular interest in marketing, will be implementing the balance between end users and advertisers.
As for Adobe dropping Flash... they likely will push the technology into other products. They still offer a great vector manipulation system.
What you are describing is ActiveX. A heavily used API/ABI with a web interface that didn't require install.
Hmmm good comment. 2 points.
1) I don't think Flash killed applets. The Microsoft / Sun war did. Java under Netscape was miserably slow. Microsoft was creating a faster usable Java but adding proprietary extensions killing "write once, run anywhere"; Sun consider that the core feature and certainly didn't want to allow for Windows lock in....
Because most people used I.E. 6 and I.E. used an old Java and a bad Java the "run anywhere" started to fall apart along with the speed. Sun got neither and Java became server only. Flash never had that problem since
a) Flash always worked best on Windows browsers.
b) Flash had lower ambitions that didn't threaten Microsoft
2) I think applets were a major success. They broke the hold of HTML/CGI. Prior to applets there was no way to even having something like a variable menu. The web was static information + forms + response. It simply didn't offer anything like interactivity. Think about something like .pdfs today.
The hardware differences go beyond just screen size. Moreover the software differences go further. Apple wants people to use Cocoa and thus have integration with desktop versions. Google isn't pushing that.
Why would you do this? You can just as easily write the code on an old Mac and compile on a server. Apple has all kinds of great tools for using bad computers and fast servers for a compile so people can share.
If it is small just do it all on an old mac.
Microsoft has fantastic server solutions that are expensive. Their interest is to promote things like Sharepoint which create Office lock in.