Slashdot Mirror
Plugin Availability For Non-x86 Browsers?
Foredecker writes: "Many, many Internet appliances are being built with non-x86 processors such as Mips, ARM and PowerPC. Supposedly, one of the barriers to using such processors in Internet appliances is the notion that x86 has, by far, the advantage that many popular browser plug-ins are only available for x86. If they are available for non-x86 systems, then they are either available late (x86 first) or their are inferior to their x86 brethren. Is this a problem? Is it true? If it is true, is this going to make it harder for non-x86 based Internet appliances to win acceptance in the market?" Earlier this year, we talked about how the Web is now flooded with non-HTML content. Now I don't mind enhancing one's Web experience, but it would be nice if the folks who make these plug-ins realize that the Web is not only for those folks running Wintel or Macintosh systems. When will plug-in makers realize that there is a larger market out there who may also be interested in their product?
Hiya, There's a company called Insignia Solutions plc(Nasdaq: INSGY) who do write an optimized embedded VM (and plug-in) for strange processor platforms. I believe a Linux/ARM port will be ready for release around Christmas. It will also do WinCE/MIPS, VXWorks, etc...
It's a nice solution, but web developers will have to wait a few years before they can really use it. How many people on slow modem connections are really going to download the latest several-megabyte version of a browser just to get SVG? Many are still using version 3 or lower of the top two browsers.
Last time I checked some statistics, the various variants of Windows ran 95% of the browsers, Mac 4% and "Others" one single percent. Thats Linux, Solaris, BSD etc combined!
This means that unless you are specifically targeting the slashdot crowd, you have no reason to care wether your pages work outside the Windows/Mac world.
Most clueful webfolks stay away from the plugins anyway, since they add more maintenance cost to your site then they are worth. Nobody cares wether that single percent can view the site or not. If they can, its a bonus, if not: their problem.
All opinions are my own - until criticized
This reminds me of a Boston Globe review of the iBook where Simson Garfinkel (who is knowledgeable and not a zealot) loved its design but was extremely pessimistic about using MacOS. He noted that the stability has improved to the point that it's barely an issue for desktop users, but said he found himself locked out of all his favorite sites.
What am I missing here? I haven't touched Windows in months. I do all my browsing on PPC, either using IE or Mozilla for MacOS, and Konqueror in LinuxPPC. I haven't even bothered to install any of the available plugins besides Flash and Shockwave. And the only site I have trouble with is mynetscape.com which blocks IE connections. Am I just reading ancient, lame sites? Or has it been so long since I've used Windows that I don't realize every site now has holographic VRML that I'm just not seeing? Could someone point me to a site that will show me what I'm missing?
5% isn't a "tiny" market share. Neither is 1%.
If your marketing department doesn't get this, replace "1%" with "one MIIIIIIIILLLION affluent Internet users with the combined purchasing power of fifty BIIIIIILLION dollars". Preferably within earshot of the CEO.
Any technology which is distinguishable from magic is not sufficiently advanced.
I've been toying with this idea, and I might as well try it out here.
The free software answer to this question is, distribute source (typically in C). This works pretty well for us; although it is easy to include assumptions about the target architecture into the code, a port or two usually shakes them out. The Java answer is, distribute Java bytecode. However, most acknowledge that this in not a panacea either, as dependencies on particular implementations frequently arise.
Fundamentally, what is the difference between bytecode and C source code? Very little, except that most companies resist distributing the latter. Usually, this is jutstified as protecting company secrets.
But bytecode can be decompiled, and source can be obfuscated, with some degrees of success. In fact, early Java compilers produced code that could be easily decompiled with great results. I imagine that compilers are better now (due to more complicated techniques), but Java bytecode still retains a lot of information. How well do decompilers work today? Is it plausible that you could get more information out of bytecode than obfuscated C?
Imagine a GCC backend that, after parsing and machine-independent optimization, output C. Since this process is not substantially different from what a Java compiler does, I would expect both processes to result in similarly comprehensible output.
If this were true, it would be possible to hide "secrets", while gaining the benefits of C code, eg wide availability of compilers, interoperability with most API's (like plug-in API's), and native code speed.
Are there technical reasons this couldn't work? I know that at least one company has been accused of distributing obfuscated hardware drivers and passing them off as free software. Political issues aside, people seemed to think this effectively reduced the usefulness of the source. I believe this case was "remedied" in short time; but has any company successfully used this strategy to achieve portability without exposing "secrets"?
The evaluation of an action as 'practical' . . . depends on what it is that one wishes to practice.
With more and more platforms having some sort of a web-browser, it is likely that a plugin is not available for the particular platform.
An important 'feature' of a plugin is its security hazard. Running native on the client system may be nice for performance, but also means that it has full access to all resources AND to the internet. This means that we have to scrutinize the quality/honesty for every plugin installed. I'd rather have just one thing to worry about: the browser.
And especially for embedded platforms (I'm working on an In-Flight Entertainment System) it is often not desirable or possible to run plugins. It's unlikely that third-party code will be executed, without some sort of _very_ rigorous verification.
I don't quite understand it either. You _can_ do quite a bit with a Java applet, or even with JavaScript. I know the sound/image synchronization of Macromedia is nice, but look what it's used for: marketing.
And I says: fsck marketing. The Web was designed to provide information on demand, not to watch advertisements (which is what most Flash content really is). It pisses me of when I have to sit through a frigging commercial before I get to see what I want.
In my opinion it's better to extend the capabilities of the browser, if new features are desired. This way it will be a more standardized solution (read NOT proprietary) and it's possible for _everyone_ to implement it in their browser.
Breace.
Most plugins are unnecessary gaudy crap glued onto the browser to create a market where none really exists. There are a handful of plugins which have "real" value. Depending on your point of view, it includes Flash, Shockwave, Real, and Acrobat. Most plugins aren't created to fill a need, they're created to make their authors money.
When you're trying to sell gaudy crap to someone, being multiplatform, stable, or well designed is irrelevant. The important thing is that it looks shiny on the computer of your end sucker... erm... customer. Joe Clueless Webmaster doesn't really know what Linux is, if the shiny crap looks good on his Windows 98 / MacOS 9 box, it must be good. Jane Clueful Webmistress does worry about Linux, WebTV, the Dreamcast, and other alternative web browsers, but Jane also realizes that good web sites are actually harmed by gaudy plugins.
Supporting another platform costs a great deal of money. You need developer time to support the new platform, an variety of hardware to test on and QA time to make sure it works well. It's a lot of money to try and sell product to a hostile audience.
So quoth Cliff:
Simply put, if you're selling snake oil, don't worry about convincing the real doctors, just talk fast enough to fool the marks. There is no "larger market."
For this same reason, open sourcing the plugin doesn't help. The open source and free software communities have no sympathy for unnecessary junk.
Search 2010 Gen Con events
The same process that caused you to drop Lynx for Netscape caused me to drop Cyberdog for Netscape, some years ago- several sites I commonly use (most notably Spamcop but to a lesser extent Thresh's Firing Squad)) make sensible use of Javascript.
It seems as if Javascript is of most utility on the Web, Java turns up occasionally in fancy banner ads, and Shockwave sites seem to focus so heavily on it that (like Camp Chaos) there's absolutely no point in going there unless you're setting out specifically to watch Shockwave movies and things- ideally Macromedia should accept this and continue to make heavier emphasis on standalone viewers, and sites should assume that some people will be content to download .swf and view it separately rather than assuming everyone wants to view it in a browser. Camp Chaos, I think, is a good example of what really happens in practice- loads of people have viewed Napster Bad etc. but that is not a navigation system, or webcandy- it's essentially a scripted cartoon, obviously standalone content. The standalone aspect is much more interesting than the 'interactive' aspects of the special edition, and in general the cartoons in the format are creative and entertaining (and very low bandwidth) and the games are just stupid :)
We all know Microsoft struck pure gold with DOS.
All their successful software follows a distinct pattern. First sell it at a loss or bundle it with another product so you get a lot of users. Then sell it at a price no commercial software shop can match without the MS user base. Monopolic? You bet! The catch is that every *single* user gains from using MS software, while users as a *collective* would benefit from choise.
As for Netscape, I must admit that I worked for a compny that provded me with payed-for Netscape at that time. I do recall that the free version lacked some features, but alas I never compared them.
Fire in the Valley is definitely on my reading list.
It really doesn't matter to a sheep if Microsoft has greener grass because they grow it well or because they torch their neighbors fields.
What *does* matter to the sheep is that the other sheep are in the MS field. Sheep like company. Wolves like lonely sheep.
Luckily we are not all sheep.
All opinions are my own - until criticized
AmigaDE is a completely different beast. With AmigaDE Amiga does all the hard work for you, they build a bunch of libraries and a runtime interpreter for different architectures under a single API. You then write apps for DE which get compiled to a P-code sort of state, they will not run without the Amiga runtime. You can load up the runtime from any OS and then run your Amiga apps natively on Linux or Win32 or MIPS-64 or PPC or any other microprocessor.
I'm a loner Dottie, a Rebel.
because it's not a 'larger' market. The target market is the home consumer, which think Linux plays a piano.
The Game Guy
Hell. Non-x86 must be a giant problem. I have problems finding the darn plugins in need for Non Windows / mac platforms. Even if they are x86 based. I would give my left foot for Quicktime or ASF working nicely under Linux. What needs to be done is a standard plugin format which does not depend on Machine Code. *cough* bytecode *cough* cough* Java. Im just bitter
--------========+++Dont Feed The Lab Techs+++========--------
A quantitative determination of the answer for any question of the pattern ``when will x be done by y company'' is now possible after many centuries of theoretical thought regarding the flow of resources in human society. The answer is: exactly when doing X is economically advantageous to company Y. No sooner, but maybe later if the company is run by idiots...
:-)
On the topic of plugins, well, the appliance writers should either partner with plugin companies in order to port the plugins needed or write ones of their own if the file format is an open standard. This solves the ``now'' problem but it probably precludes any $29.95 devices any time soon...
--
News for Geeks in Austin, TX
Java is the solution for internet appliances. Projects like Mozilla will bridge the platform gap, and special plugins and browser extensions could be written in java. Java's UI has come a long way with JDK 1.3.
Someone you trust is one of us.
"the web is now flooded with non-HTML content"
This is a web designers choice, i.e. he/she may or may not use plugins. Most non x86 devices don't support newer web standards anyway, so it is still a must for some designers to create different versions of a web site no matter how undesriable. The fact that plugins are not supported is irrelevant as there are still bigger issues to be addressed....
"Ceilean Súil an ní ná feiceann..."
Heh, having a blown motherboard and having to use my Dreamcast as my only means of net access for the past week, I've quickly learned the web basicly sucks if you're not using the latest version of IE on the latest version of Windows.
You should tell this to the people who develop web pages. It is their fault if they only program for _one_ "platform" "plugin" "browser" or whatever. There always should be an alternative view. I can see it often for flash pages...where you can choose the "non-flash" version. This should be the case for every non-html content. It is the fault of the developers.
Last time I checked, over 95% of people were using Windows or Mac to access the web.
I'm not sure that the corps care about the tiny market share increase enough to have a site that doesn't look as nice.
Sure, it's nice to have a 'flash/non-flash' option on the first screen, but that means doing two websites. How many corps want to pay for that?
_____
My Journal
This is exactly the kind of problem that Java was created to solve. Unfortunately, Java as we know it tends to be buggy as all hell, and runs with the speed and stability of a 3-legged cow.
Lots of embedded devices run on things like ARM processors or other chips that aren't found on the desktop market at all, so we're not just talking about resource optimization. Since many embedded devices also use alternative operating systems, the APIs are different, so we're talking about more than just a recompile there.
The only way to get around inconsistent APIs and processor calls that I can see having long-term portability is to standardize to the Java Virtual Machine. Just-in-time compilers and other sorts of wonderfulness are still in their infancy, and I've been told that Apple's Java VM runs bytecodes as fast as native compiled C code.
I know the whole system is kinda bloated at the moment, but the write-once-run-anywhere method does fall into line with the "worse is better" mantra of Unix programming, which has been incredibly successful over the years. Unfortunately, we are still it the point where worse is still worse, and have yet to see the really good benefits that portability and standardization promise.
I think Sun should be applauded for their vision. Let's face it, they've done wonders for the web, and Java is going to take it a lot further. Sure the tech is buggy at the moment, and they have been a little more reluctant to opening the standards than most here would like, but I think we should give credit where credit is due, and then kindly ask them if they'd turn their toy into a tool.
WARNING: there is a trojan on your
Leaving aside Java for the moment, I see plugins and applications for Sparc and Intel Linux that handle:
- All the ms file formats
- Adobe Portable Document Format
- Postscript
- Frame maker
- ULAW Audio
- MIDI Audio
- AIFF Audio
- RealAudio
- MPEG Video
- Soundtracker audio
- Microsoft wave files
- Commodore 64 audio
- PNG image
- Sun raster image
- TIFF image
- IFF animation
- DL animation
- FLI animation
- MPEG Animation
- AVI animation
- Quicktime animation
- SGI animation
I haven't looked for flash, so I suspect it's the popular missing link this week(;-))I collected that list this summer, as part of the research to see if writing a book (with this as one chapter) was worthwhile. Then I got hit by lightening...
Ok, folks, should I write a book on this? (I was the second author on O'Reilly's "Using Samba"). If so, send me mail!
--dave
davecb@spamcop.net
The problem with this 'larger' market, is that it doesn't really exist. The number of people surfing the net and using linux/bsd on the desktop is miniscule, even compared to the Macintosh, let alone wintell.
Java applets show some promise, but write-once, debug everywhere is still a problem.
--
From a business standpoint, it's wonderful to have flashy content (and I don't mean only Macromedia's stuff)
But it's also a bad thing to leave a large portion of users behind.
Internet Appliances along with mobile phones and other devices now getting on the internet are lumped under the heading of *PERVASIVE DEVICES*, or pervasive computing.
There are currently two options available to address these devices.
Option one is to follow the model of Phone.Com. Sell the browser to the device companies and collect a license fee for each device sold. Then sell the access point for those devices a proxy software that converts html to code for their browser.
Option two is to use IBM technology. WebSphere Transcoding Publisher (ok, I'm biased, I know people on this team) runs on an LDAP, proxy, or reverse-proxy model, and checks to see what device-type is accessing a page. If it has a device type profile for the device making the http request, it translates the code (transcodes) the html into code suitable for the device, along with altering the images to meet the device's needs.
If a transcoder device type profile were written for the various internet appliances and other non-Wintel/Mac browsers then non-html content requiring plug-ins would no longer be an issue.
Of course, this is kind of a dodge- No phone that I know of attempts to be a desktop computer browser (yet), and while we might expect that an internet appliance that hopes to be the desktop browser should play flash and shockwave content, it's not a desktop computer, and we shouldn't expect it to be one.
Yes, I like my alternative opsys' browsers to be supported, but no one writes flash plugins for NetPositive, and my Netscape Communicator on OS/2 relies on flash plug-ins for win3.1.
I don't think the fault for this situation lies with the makers of plug-ins, it lies with the page designers for not providing a stripped down version that doesn't require plug-ins, and with the internet appliance marketeers for inspiring false expectations.
A host is a host from coast to coast, but no one uses a host that's close
That is the problem. Ask any x86 Linux user who couldn't view a particular page because of required plugins available only for Windows.
As someone who used to be involved with producing a plugin for a particular alternate image file format, I can honestly say there are several issues associated with trying to produce and maintain plugins for multiple platforms -- not the least of which is maintainability and keeping them all current. Add to that the hardware costs of releasing a binary for linux-libc6-strongARM and every other os/hardware combination you can come up with, and you wind up with exactly the same cost-benefit analysis situation you do when producing any other proprietary software. We were among the best in terms of cross-platform support at our peak (fall '96), with:
- Win 3.1
- Win 95/NT
- Mac 68k
- Mac PPC
- Sparc Solaris 2.4
- Digital Unix 3.2
all as supported platforms, but invariably there'd either be somebody grumbling about the fact that the "latest" version wasn't out for Solaris yet (hey, at least we had one), or "can you make one for Irix?" or whatever; the fact remains, most plugins are made by small companies who are doing the best they can with the resources they have to support the most people.Don't chime in about open-sourcing the codec either: compression codecs are exactly one of the r&d-intensive products for which retaining closed-source is the only viable revenue model that ESR talks about in "The Magic Cauldron." If the only thing you have of value is your algorithm, you really can't distribute source that everyone can implement.
The point is, there is no good solution: if you are building a site that contains essential content, don't use non-standard technologies (I'd argue this includes Java applets). Conversely, if you're not using the same technology as 80% of the people out there, and I too am in that outer 20%, prepare to fall victim to the 80-20 rule when vendors of everything, including browser plugins, do cost-benefit analysis.
MOO;IANAL.
MOO;IANAL.
There used to be a picture linked here.
IIRC, the early test versions were actually released for Linux first.
The full release for both OSs was actually released at the same time (Mac might've been out at the same time too, I don't recall). However, there were problems actually getting the Linux version because it comes in a fancy tin box and the availability of that box was not so good.
Also, there were problems in that many of the retail outlets simply didn't stock the Linux version. As you said, "A majority of their targeted user base were Windows users". Id software, however, seems to be very interested in Linux.
WebTV uses SpyGlass, Device Mosaic. It doesn't support plugins, unless WebTV has modified DM heavily. I know this because I'm porting DM to an In-Flight entertainment system. We will not support plugins either.
Pity those of us who want to create innovative content using a "specialized plug-in" such as Chime. (Chime allows 3d molecular display in a web page. Yes, there are Java applets that do this. When one of them can do 1/10th of what Chime can at better than 1/10th the speed I might look into it again.)
Chime lets me write pages like these: Atomic orbitals and Crystal unit cells Very useful in my line of work (Chemical education) But as soon as I do this I lock out all "alternative OSs"
Chime's actually more cross platform than most: Windows, Mac and an older IRIX version. It even is anti-MS: it runs like crap in MSIE on both Mac and Windows. (The Mozilla team fixed the compatibility problems when I sent the bug in.) But I can't do this stuff and make it truly cross-platform.
This is a problem for me: in fact, it's one of the main reasons I don't use Linux on the desktop.
"Seven Deadly Sins? I thought it was a to-do list!"
Just looking at the latest server logs from the site I manage explains the lack of plugins for alternate OS's. Out of 1.7 million requests, only around 3500 came from the various Unix variants. Even Slashdot has said that Windows browsers make up a majority of their traffic, and this is a Unix stronghold. It doesn't take a braindead marketroid to figure out the numbers here...
The new Scalable vector Graphics will help as it provides a way to develop interactive web pages, ala Flash, in a standard way. The real answer is for the W3C to define more such standards covering all the current extensions that now require plugins.
You mean an open standard like Macromedia's Shockwave Flash format, complete with an open implementation. It's not a W3C standard, and hence sites shouldn't use it exclusively (there should always be a non-flash option), but the specs and a reference implementation are freely available. Ideally, the W3C would adopt it as an official standard, but it would need work on accessibility issues first (e.g., for those using text-only, or braille or voice browsers).
"The invisible and the non-existent look very much alike." -- Delos B. McKown
Netscape 4.71 on IRIX (so that includes all 64-bit MIPS processors and the 32-bit ones who think they're 64... anything from R4K on up) runs exactly like it does on a PC with Win or Linux. I can run flash, realaudio, and most other major plugin content.
:)
Yes, that also means that Netscape is bloated and crashes arbitrarily every 30 minutes or so, but you gotta take the lumps with the bad.
If only MSIE were ported to every platform... sigh.
-Chris
...More Powerful than Otto Preminger...
I have used Alpha-Linux for many years now and have constantly had problems getting software. It has gotten far worse lately as more and more binary only software is coming out for Linux. Not only are vendors who claim to support Linux only supporting Linux/i386 but an awfull lot of code out there is sloppily i386 or at least 32 bit specific.
It is not a big deal to provide binaries for simple things like plug-ins for multiple platforms under Linux so I don't understand why the vendors don't. At the very least, if they don't intend to provide anything but Lintel software than they should not say they have Linux support but Linux/i386 support.
I think unless a push is made to force vendors to support the major alternative Linux platforms (Alpha,Sparc,Mips,Arm,PowerPC) then there soon will be little to no support for them. Remember what happened with Mips and Alpha Windows NT!
This is one of Crusoe's big selling points. The fact that it's X86-compatible (yet low power etc. for internet appliances) means that standard browser plugins work with it.
Gerv
If I can't visit a site without having to load a plug-in or activating Javascript then I will tend to steer clear. I know of one site that requires you to install shockwave to view their site and you can't see the html because there isn't any, even worse the company's market is Unix based, predominently AIX and there isn't even a plug-in for that platform - this is a serious problem. Before adding plugins to a website the following question should be asked:
- Can your audience run the plug-in?
- Do they have the authority to install it?
- Is their connection fast enough to download the plug-in?
- Why should the user go through the hassle just to see your site?
- Does your audience have the knowledge to install a plug-in?
- Why can't you make do with standard HTML?
- Is your connection really that fast? - go and try seeing your website from elsewhere
Yep, I am anti-plug-in and with good reason, though I will allow for a movie player that plays either MPEG, AVI or Quicktime - partly because movies are a missing element in the standard web file formats.
Jumpstart the tartan drive.
I think when it comes to plugins (and, come to think of it, Quicktime) the designers are in love with what's possible, and the suits are unaware of the compatibility issues. Face it: a designer shows a net-illiterate suit two designs, one with Flash animation, and one without, and he'll choose the best looking one, even if it only really works on the designer's G3 (the suit doesn't know otherwise).
Let's assume 5% of browsers can't handle (some) plugins...
No store owner would be satisfied if 5% of all customers were turned away at the door; likewise I'm sure no e-commerce website owner would be satisfied that 5% of visitors were being turned away. I'm sure banner-advertisers would be disappointed to discover how many potential impressions they may be losing because a site is unviewable to a (growing, with the advent of set top boxes etc) proportion of the target audience.
Make the guys with the purse-strings aware of this problem, and it'll go away. 5% more turnover is the kind of figure these people like to hear about.
--
Make the plug-ins portable between platforms. Use Java for the Plug-Ins. That way they can be ported between platforms. Make the code in the browser to load the Plug-In as native code or Java code.
So is this a good idea or not? Let me know.
yeah, I can see it pretty easily really. The Java2D classes in JDK 1.2.x and 1.3 are nice. Think the GIMP and PostScript having a love child, with a dab of Illustrator. Since the .swf file format is an open one, you could write a Java-based parser class (ByteInputStream anyone?) and use it's output to feed a Java2D-based ``rendering engine''...
That would actually be pretty sweet. If I were more of a Java God I might do it... :-)
WRT doing it on an ARM or other embedded processor, why not? StrongARMs are, what, 200-odd MHz? That's faster than the machine I'm using to type this (p166), and this guy handles Flash and JDK 1.3 applets pretty well. Now what if that internet app is using a Crusoe, eh?
This is not to say that your company's product doesn't sound nifty, it does...
WRT to VRML, I think that standard is bloated enough already irregardless of the plugin playing it... ;-).
--
News for Geeks in Austin, TX
I don't have any plug-ins installed on my system, and I've never had a problem - the only one I'd consider is one for PDF, as that is accually used. I've only seen a couple web sites that even have any other plug-in requested (shockwave or flash - or are thsoe the same?) and every one allows easy bypass. I might miss some fancy animation once in a while, but I've never found a web site where that was critical. I've seen far to many websites where java script or graphics were critical (forcing me to abandon lynx for netscape), something I mind when I'm after information that shouldn't require either, but don't mind when it is accually useful. Case in point: My bank requires javascript, but I once figgured out how to bypass that check and I couldn't find a thing that didn't work.
When netscape asks if I want to download a plug-in I hit no. I don't trust random binaries on my system, I'm uncomfortable enough with pre-compiled netscape (someday I'll compile mozilla and solve this), and I compile everything else.