"Rich text format seems to be the preferred document format among open-source word processors"
"Heck, even MS-Word can read and save RTF!"
RTF is a Microsoft format. It's essentially a text version of DOC. Modern versions support the same macro and embedded COM object capabilities that DOC does.
It's true many independant vendors have implemented the Word2 or Word6 version of RTF, but that doesn't make it an open or completely documeted spec by any means.
Your post does highlight the issue that there are no standard formats in the OSS/Unix world, and nor are there 'standard' applications (as MS Office has become on Windows and Mac), and that OSS/Unix users have to fall back to Microsoft formats to interoperate with each other.
I like the idea of making IE behavior an optional pref. I've had users complain that forms are broken because they don't submit on Enter in Netscape, and then you have to explain that it's a browser issue, not your form.
The classic Netscape thing to do was only submit the form on Enter is when the form contained only 1 text field. That seemed to be how Moz worked up to 0.93.
IE generally submits all forms on Enter, including sometimes when you are trying to type in a textarea:P I haven't got this installed yet, but it would be really interesting if they went with IE's convention.
OK -- where is CORBA used (directly) for component development?
Instead Free Unixes have developed umpteen replacements for COM -- KDE's got (at least) one. Gnome's got one. Star Office has another. So does Mozilla. I've heard Motif has one too. Anyone else?
Great! 4+ competing specs which means what? Your interoperability is in the toilet. On Windows, I can embed 1-2-3 graphs in Word, or Excel graphs in WordPerfect. On Unix, I get single-vendor lock-in. (This is far worse than having your widgets look differently, BTW.)
Word had a decent enough WordPerfect import filter. Well, maybe it wasn't very good at preserving all formatting, but that was OK because WordPerfect's non-WYSISYG interface sucked so hard that very few WP docs had any formatting.
I somewhat agree with the point that import capabilities aren't the most important feature, but back in the WordStar/WordPerfect/WinWord2 days, transferring documents by e-mail was far far less common.
Thinking specifically of the 32-CPU Unisys boxes, which are being marketed with sweetheart deals specifically towards Sun customers. No, I don't know how many they've sold or where they line-up against Sun's product line.
First of all, Sun will be charging for support contracts, so not quite free for most corporate use.
Also, IBM tried offering SmartSuite essentially for free to shops they had a relationship with. They were also bundling it with their PCs and selling it very cheap at retail. The result was that they got very very few users -- I worked for a place that tried to standardize on it, but rampant MS Office piracy and document compatibility pretty much killed that idea.
My feeling is that Sun StarOffice exists because Microsoft is poking a stick in Sun's eye (big servers), so Sun is poking them back (office suites). If big name vendors such as IBM/Lotus and Corel/WordPerfect could field full featured suites and utterly fail to compete on price with Microsoft, it won't be any different with Sun.
That, and as an eat-your-own-dogfood shop, Sun probably felt having a piece of essential internal infrastructure under the control of a small company teetering on the edge of existence was probably a bad idea.
Now, when Microsoft's OEM licence practices are altered by the courts, StarOffice may well become a standard OEM freebie. However, that doesn't mean that many corporate users will or could switch.
Microsoft made a Z-80 coprocessor board for the Apple II back in the 70s. Much like the mice, the purpose was to sell more software (in this case Microsoft's branded version of CP/M and MBASIC).
5.1 Internet Media Type
As of the publication of this recommendation, the general recommended MIME labeling for XML-based applications has yet to be resolved.
However, XHTML Documents which follow the guidelines set forth in Appendix C, "HTML Compatibility Guidelines" may be labeled with the Internet Media Type "text/html", as they are compatible with most HTML browsers. This document makes no recommendation about MIME labeling of other XHTML documents.
Which implies that a text/html type would be more appropriate in this situation.
Although, I agree that there's something fishy about IE. Considering that it parses all sorts of strange MIME types (including text/plain) as HTML if the file starts with <HTML>, it ought to handle this, even if it can't groak the namespace.
Why is it 'correct' for a browser to take a text/xml stream and render it as HTML? If the intent is to transmit HTML to a browser, it would seem that MIME type is fairly misleading.
I'm not saying it's wrong, just that it's a rather undefined situation and no better or worse that IE's approach of slapping the default XSLT sheet on it and displaying it.
I'm a little discouraged at lkcl's writeup, which manages to mention "Microsoft" at least 6 times. As pointed out, DCE/RPC was originally a UNIX protocol, developed at Sun and maintained by The Open Group.
While it's true that Microsoft reverse-engineered the protocol and uses it heavily in their network products, I imagine that an open source implementation would be useful for even a pure Unix shop.
The full licence for the real thing costs $100,000, although it appears to be free for internal and educational use. Has anyone asked TOG if they would consider an open source licence?
Re:I'll get hammered, but Internet Explorer 6 is o
on
KOffice 1.1 Rolls Out
·
· Score: 1
Probably the nastiest extention to the standards has been IE's client-side data-bound HTML elements, which allow you to specify a data source in the HTML and have IE make a ODBC reqeust to the server to get the data and populate the table or form or what-have-you. Big security hole on the server too.
Fortunately for the world, Interdev is such a POS that it's usually ignored. What's a little more scary is ASP.NET's HTML generation library. Right now it seems to produce straight-forward HTML, but looking at the API which simulates client-side events on the server (something like btnSubmit.onClick()), you could see how they could maybe move the thing wholesale over to an ActiveX applet and not break any code.
(partly I believe because NS was trying to keep up with all the proprietary changes that MS were making to HTML)
Others have already flamed you, but I'll pile on.
In 1995-6, Netscape had a 90% marketshare and essentially told the W3C to fuck themselves, and went ahead and implemented a completely proprietary Document Object Model (document.layers) and style sheet system (JavaScript Style Sheets). Meanwhile they were running around to their 'enterprise' customers and preaching the wonders of open standards and encouraging to replace their existing client-server apps with Netscape-based web applications.
Well, the W3C quite naturally ran to Microsoft, who managed to implement some of the W3C DOM and CSS specification in IE (at least until they started winning the marketshare battle, in which case they radically slowed down this standards support).
Meanwhile, Netscape's customers, many of whom had bought their quite expensive server products, figured out that Netscape had set a proprietary lock-in trap and were pissed. Netscape promised to make it all better and support the W3C standards.
At this point, however, their 'tag salad' renderer couldn't be hacked any further, and they felt the need to start over from scratch. While they were at it, they scrapped pretty much the entire Netscape 4.x codebase, thus starting the 3 year dev cycle that brings us to today.
As for 'HTML', everyone is pretty much aware of the proprietary extentions (Netscape has BLINK, IE has MARQUEE), and they aren't that big of a deal. 90% of the stuff that IE supports and NS4 doesn't is actually standards-based. Their approach to missing close tags is different too, but again, not that big of a deal for good markup.
It's the next generation stuff (DOM, CSS) that Netscape attempted to embrace-and-extend shittily and for that they are doomed to their miserable marketshare. I know you want to bash Microsoft, but until the release of Netscape 6.1 they've been playing the webstandard game far better than Netscape.
From what I've seen of the Netscape -> Mozilla progress it appears it's because Netscapes initial design was pretty poor and it wasn't easily maintainable as it progressed.
You bet. Netscape pounded on the 4.x codebase for at least 2 years trying to create 5.0 before chucking it as junk and starting over with a clean slate. That is the single reason that Mozilla is so far behind Microsoft.
And to disagree with the AC -- Selling out to AOL for $Billions was pure genius, especially when the IP of your flagship product was essentially worthless, and all the engineers bailed lickty-split. AOL is looking at a situation where tons of money *has* been poored into the browser for the last 5 years, and there's just now barely something to show for it and barely any userbase left that cares. Of course they're going to cut their losses at some point.
(Standard disclaimer: Mozilla is great, but it was essentially screwed market-wise by Netscape 4.x suckyness.)
Microsoft does some funny redirection with cookie placement in order to get centralized MSN cookies placed and logged while still not tripping their own privacy feature in IE6.
Even in 96, the writing was on the wall for LD. "It's like selling water", one reseller told me back in 92 or so, "everyone has the same product, only the brandnames are different".
I think the comment about the bells being 10 years behind is right on the mark. Don't forget that these guys are born-and-bred AT+T end-to-end monopolists, and controlling things end-to-end is the only way they know how to think.
Fiber is never going "to the curb" in the US, except maybe in brand new developments or high-density, high-dollar townhouses. The cable companies almost bankrupted themselves wiring America in the 80s, and in the modern competitive landscape, nobody is going to there again -- AT+T, once the biggest blue chip, is now laden with gazillions of dollars of cable build-out debt.
(And DSL is extremely cost effective technology, especially for neighborhoods like mine where density is high, and good 80 year old copper goes everywhere. Essentially free money for the telcos, although the whole Covad-type middleman thing turned out to be bogus. It's true that DSL is not effective for the more prevelant 70s-style suburban development. Within time, there will be much cheaper wireless solutions, tho.)
Well most popular games are ported to the Mac eventually (call it the other 7.5%, although it's probably higher than that for the home market).
When you have a game that the publisher knows is going to be a big hit, like say Quake 3, it probably makes financial sense to build a portable engine up front and release simultaniously on Windows and Mac.
The problem is that most games aren't sure hits at all - something like 90% of titles lose money, in which case additional portability/porting costs are pointless.
There was this thing called "Project Monterey" which was YA grand UNIX unification plan. Essentially, AIX and UNIXWare were to be combined and ported to IA64.
IBM got smart and moved to a Linux strategy, and apparently so did SCO, so the thing was dropped. When AIX5L shipped, IBM said "ummmm, *This* is Project Monterey".
Slashdot Mirror
RTF 1.0 Specification "Copyright © Microsoft Corporation"* right here
:)
*or it would say that if Word 2000 could properly read RTF 1.0 files
Some of this stuff is in CSS-3, but that's not done yet.
"Rich text format seems to be the preferred document format among open-source word processors"
"Heck, even MS-Word can read and save RTF!"
RTF is a Microsoft format. It's essentially a text version of DOC. Modern versions support the same macro and embedded COM object capabilities that DOC does.
It's true many independant vendors have implemented the Word2 or Word6 version of RTF, but that doesn't make it an open or completely documeted spec by any means.
Your post does highlight the issue that there are no standard formats in the OSS/Unix world, and nor are there 'standard' applications (as MS Office has become on Windows and Mac), and that OSS/Unix users have to fall back to Microsoft formats to interoperate with each other.
I like the idea of making IE behavior an optional pref. I've had users complain that forms are broken because they don't submit on Enter in Netscape, and then you have to explain that it's a browser issue, not your form.
The classic Netscape thing to do was only submit the form on Enter is when the form contained only 1 text field. That seemed to be how Moz worked up to 0.93.
:P I haven't got this installed yet, but it would be really interesting if they went with IE's convention.
IE generally submits all forms on Enter, including sometimes when you are trying to type in a textarea
"under free OSen its called CORBA"
OK -- where is CORBA used (directly) for component development?
Instead Free Unixes have developed umpteen replacements for COM -- KDE's got (at least) one. Gnome's got one. Star Office has another. So does Mozilla. I've heard Motif has one too. Anyone else?
Great! 4+ competing specs which means what? Your interoperability is in the toilet. On Windows, I can embed 1-2-3 graphs in Word, or Excel graphs in WordPerfect. On Unix, I get single-vendor lock-in. (This is far worse than having your widgets look differently, BTW.)
Word had a decent enough WordPerfect import filter. Well, maybe it wasn't very good at preserving all formatting, but that was OK because WordPerfect's non-WYSISYG interface sucked so hard that very few WP docs had any formatting.
I somewhat agree with the point that import capabilities aren't the most important feature, but back in the WordStar/WordPerfect/WinWord2 days, transferring documents by e-mail was far far less common.
The average individual office drone only uses 5% of MS Office, but across an organization of some size, 90% of the features are being used?
Speaking of cross-platform StarOffice, how is the Mac port going? That would probably rank higher on the average IT manager's mind than desktop Unix.
Thinking specifically of the 32-CPU Unisys boxes, which are being marketed with sweetheart deals specifically towards Sun customers. No, I don't know how many they've sold or where they line-up against Sun's product line.
But, but, but, SO is FREE.
First of all, Sun will be charging for support contracts, so not quite free for most corporate use.
Also, IBM tried offering SmartSuite essentially for free to shops they had a relationship with. They were also bundling it with their PCs and selling it very cheap at retail. The result was that they got very very few users -- I worked for a place that tried to standardize on it, but rampant MS Office piracy and document compatibility pretty much killed that idea.
My feeling is that Sun StarOffice exists because Microsoft is poking a stick in Sun's eye (big servers), so Sun is poking them back (office suites). If big name vendors such as IBM/Lotus and Corel/WordPerfect could field full featured suites and utterly fail to compete on price with Microsoft, it won't be any different with Sun.
That, and as an eat-your-own-dogfood shop, Sun probably felt having a piece of essential internal infrastructure under the control of a small company teetering on the edge of existence was probably a bad idea.
Now, when Microsoft's OEM licence practices are altered by the courts, StarOffice may well become a standard OEM freebie. However, that doesn't mean that many corporate users will or could switch.
Microsoft made a Z-80 coprocessor board for the Apple II back in the 70s. Much like the mice, the purpose was to sell more software (in this case Microsoft's branded version of CP/M and MBASIC).
Well, here's what the W3C says:
5.1 Internet Media Type
As of the publication of this recommendation, the general recommended MIME labeling for XML-based applications has yet to be resolved.
However, XHTML Documents which follow the guidelines set forth in Appendix C, "HTML Compatibility Guidelines" may be labeled with the Internet Media Type "text/html", as they are compatible with most HTML browsers. This document makes no recommendation about MIME labeling of other XHTML documents.
Which implies that a text/html type would be more appropriate in this situation.
Although, I agree that there's something fishy about IE. Considering that it parses all sorts of strange MIME types (including text/plain) as HTML if the file starts with <HTML>, it ought to handle this, even if it can't groak the namespace.
Why is it 'correct' for a browser to take a text/xml stream and render it as HTML? If the intent is to transmit HTML to a browser, it would seem that MIME type is fairly misleading.
I'm not saying it's wrong, just that it's a rather undefined situation and no better or worse that IE's approach of slapping the default XSLT sheet on it and displaying it.
I'm a little discouraged at lkcl's writeup, which manages to mention "Microsoft" at least 6 times. As pointed out, DCE/RPC was originally a UNIX protocol, developed at Sun and maintained by The Open Group.
While it's true that Microsoft reverse-engineered the protocol and uses it heavily in their network products, I imagine that an open source implementation would be useful for even a pure Unix shop.
The full licence for the real thing costs $100,000, although it appears to be free for internal and educational use. Has anyone asked TOG if they would consider an open source licence?
Probably the nastiest extention to the standards has been IE's client-side data-bound HTML elements, which allow you to specify a data source in the HTML and have IE make a ODBC reqeust to the server to get the data and populate the table or form or what-have-you. Big security hole on the server too.
Fortunately for the world, Interdev is such a POS that it's usually ignored. What's a little more scary is ASP.NET's HTML generation library. Right now it seems to produce straight-forward HTML, but looking at the API which simulates client-side events on the server (something like btnSubmit.onClick()), you could see how they could maybe move the thing wholesale over to an ActiveX applet and not break any code.
(partly I believe because NS was trying to keep up with all the proprietary changes that MS were making to HTML)
Others have already flamed you, but I'll pile on.
In 1995-6, Netscape had a 90% marketshare and essentially told the W3C to fuck themselves, and went ahead and implemented a completely proprietary Document Object Model (document.layers) and style sheet system (JavaScript Style Sheets). Meanwhile they were running around to their 'enterprise' customers and preaching the wonders of open standards and encouraging to replace their existing client-server apps with Netscape-based web applications.
Well, the W3C quite naturally ran to Microsoft, who managed to implement some of the W3C DOM and CSS specification in IE (at least until they started winning the marketshare battle, in which case they radically slowed down this standards support).
Meanwhile, Netscape's customers, many of whom had bought their quite expensive server products, figured out that Netscape had set a proprietary lock-in trap and were pissed. Netscape promised to make it all better and support the W3C standards.
At this point, however, their 'tag salad' renderer couldn't be hacked any further, and they felt the need to start over from scratch. While they were at it, they scrapped pretty much the entire Netscape 4.x codebase, thus starting the 3 year dev cycle that brings us to today.
As for 'HTML', everyone is pretty much aware of the proprietary extentions (Netscape has BLINK, IE has MARQUEE), and they aren't that big of a deal. 90% of the stuff that IE supports and NS4 doesn't is actually standards-based. Their approach to missing close tags is different too, but again, not that big of a deal for good markup.
It's the next generation stuff (DOM, CSS) that Netscape attempted to embrace-and-extend shittily and for that they are doomed to their miserable marketshare. I know you want to bash Microsoft, but until the release of Netscape 6.1 they've been playing the webstandard game far better than Netscape.
I'm actually genuinely interested in why mozillaquestquest.com doesn't work in IE and not interested in the stupid Moz sucks versus IE sucks flamewar.
Does anyone know what content-type header mozillaquestquest is sending?
From what I've seen of the Netscape -> Mozilla progress it appears it's because Netscapes initial design was pretty poor and it wasn't easily maintainable as it progressed.
You bet. Netscape pounded on the 4.x codebase for at least 2 years trying to create 5.0 before chucking it as junk and starting over with a clean slate. That is the single reason that Mozilla is so far behind Microsoft.
And to disagree with the AC -- Selling out to AOL for $Billions was pure genius, especially when the IP of your flagship product was essentially worthless, and all the engineers bailed lickty-split. AOL is looking at a situation where tons of money *has* been poored into the browser for the last 5 years, and there's just now barely something to show for it and barely any userbase left that cares. Of course they're going to cut their losses at some point.
(Standard disclaimer: Mozilla is great, but it was essentially screwed market-wise by Netscape 4.x suckyness.)
Microsoft does some funny redirection with cookie placement in order to get centralized MSN cookies placed and logged while still not tripping their own privacy feature in IE6.
Info here.
Even in 96, the writing was on the wall for LD. "It's like selling water", one reseller told me back in 92 or so, "everyone has the same product, only the brandnames are different".
I think the comment about the bells being 10 years behind is right on the mark. Don't forget that these guys are born-and-bred AT+T end-to-end monopolists, and controlling things end-to-end is the only way they know how to think.
Fiber is never going "to the curb" in the US, except maybe in brand new developments or high-density, high-dollar townhouses. The cable companies almost bankrupted themselves wiring America in the 80s, and in the modern competitive landscape, nobody is going to there again -- AT+T, once the biggest blue chip, is now laden with gazillions of dollars of cable build-out debt.
(And DSL is extremely cost effective technology, especially for neighborhoods like mine where density is high, and good 80 year old copper goes everywhere. Essentially free money for the telcos, although the whole Covad-type middleman thing turned out to be bogus. It's true that DSL is not effective for the more prevelant 70s-style suburban development. Within time, there will be much cheaper wireless solutions, tho.)
Well most popular games are ported to the Mac eventually (call it the other 7.5%, although it's probably higher than that for the home market).
When you have a game that the publisher knows is going to be a big hit, like say Quake 3, it probably makes financial sense to build a portable engine up front and release simultaniously on Windows and Mac.
The problem is that most games aren't sure hits at all - something like 90% of titles lose money, in which case additional portability/porting costs are pointless.
There was this thing called "Project Monterey" which was YA grand UNIX unification plan. Essentially, AIX and UNIXWare were to be combined and ported to IA64.
IBM got smart and moved to a Linux strategy, and apparently so did SCO, so the thing was dropped. When AIX5L shipped, IBM said "ummmm, *This* is Project Monterey".