If you saw the doe-eyed little Microsofties on C-SPAN dutifully tossing warm and fuzzy bunny softball questions at Algore and telling him how they come to work every day just to make the world a better place, you know this must be an innocent error on Microsoft's part.
Does anyone else remember, I think it was in the mid-eighties, when a PC Mag. column published a rumor that Windows was doing something to progressively decrease the performance of Lotus 1-2-3 and eventually crash it? Deja vu?
Microsoft: Making the World a Better Place, one B.S.O.D. at a time.
Have you looked at the mozilla source? Your argument would some weight if you could cite specific problems with the source--but then, people would only say "Good job, you've found some problems; you're obviously smart enough, so go fix them." If you can't code, then use the nightly builds and report any bugs you find. There are all kinds of ways to help with the development effort. In the Open Source culture, there are complainers and there are contributors (a constructive criticism can be considered a contribution, by the way). Contributors are people who would rather make something happen than just sit back and carp about the state of the project.
Why do you think the mozpeople were so keen on re-designing from the ground up? Ease of maintenance. This means, making it less difficult for developers to dive in and work on the project, among other things like easier cross-platform implementation, i18n, etc. Unless the mozilla team have gone braindead, reasonably skilled developers should be able to break off small, digestible chunks of the code and work on them without having to grok the whole enchilada.
Take a deep breath and read my post again. I stated that "the future of Linux involves..." I did not say that the future of Linux is "dependent upon..." NASDAQ, etc. I did not, as you put it, "say otherwise."
Can you imagine a distribution that came with sources (as required by the GPL) yet couldn't be built except with the Borkland IDE?
You have got to be sleeping poorly to come up with any such scenario.
You came up with this scenario, not I. If anything, Borland have hinted that their compiler products will be very compatible, and why should't they be? There's no compelling reason I can think of, strategic or otherwise, for Borland to make their products incompatible or use non-standard switches, etc.
Maybe you're not old enough to remember the Turbo Pascal era. I am. It was an amazing time for folks who wanted to write their own software, and I think Borland can be credited with making PC software development a real cottage industry. Borland have proven that they can do compilers and can (occasionally) figure out where the marketplace is headed. Look at the Borland developer survey if you want to find out what developers are telling Borland about what they'd like them to do with development tools on Linux. If Borland are still as sharp as they used to be, they'll shrink wrap a drop-in replacement for egcs that (at least according to early benchmark reports) may very well produce tighter, faster code. If they give the Linux user a choice between a Free compiler that produces good code, and a $125.00 compiler that produces much smaller and faster code and requires no tweaking to build the kernel and/or projects packaged for autoconf, I think they may just repeat their compiler revolution.
Whence cometh the "hijacking" rhetoric? Who is hijacking whom? I see one company that is in business to make money, purportedly considering the acquisition of another company that is in business to make money.
This is a no-brainer. You're Red Hat, you've just been infused with tons of capital, you want to do things to improve the market penetration of Linux. How better to do this than to buy up the key Linux development tools vendor?
How many of you who are worried that RH are trying to become the Microsoft of Linux have considered what might happen if Borland repeat their 1980s DOS compiler coup on Linux? Before Borland's Turbo Pascal came along, a decent DOS compiler couldn't be had for less than several hundred to one thousand dollars; TP rolled out at ~$75 a pop, took the compiler market by storm. Suddenly, everybody was using TP to develop DOS apps. What if Borland's Linux compilers took the Linux market by storm and left Cygnus in the dust? How many of you would be complaining that RH weren't doing anything to preserve Cygnus and leaving us at the mercy of Borland?
Like it or not, the future of Linux involves NASDAQ, venture caps, and lots and lots of money. Technologies like "egcs" and embedded development tools that are important to the future of Linux will be acquired. What we should be asking ourselves is not, "Is it a good thing that someone may buy out Cygnus?" but "Which company would we prefer to take the reins at Cygnus? Turbo? Caldera? Red Hat? Microsoft?" I'd prefer that Red Hat end up with Cygnus since they've demonstrated that they are dedicated to the GNU concept. Some other vendors seem all too willing to market closed add-ons to Linux and would probably hurt Linux if they had control of Cygnus.
Right now, the Linux development tools space is where it's at, and this would seem like a wise move on Red Hat's part.
The U.S.A. have little to fear in the way of invasion--we can still count on two large bodies of water to make a direct assault on our shores too costly to consider. In the cyberwar context, however, we cannot count on any natural geographical barriers, and we may be at a significant disadvantage because of our dependence on less-than-secure technology and our "Cover Your Ass at All Costs" corporate and government cultures. The Information Superdirtroad leads right to the back door of almost every mission critical institution and enterprise in this country and, as slashdot readers know, few of those doors are securely locked.
I think it would be wise of our military to refrain from cyberwar until the overall quality of security on corporate and government networks is improved. We can count on the military to defend us against an attacking force on the ground, but on the 'net, we're all on the front lines and it's every man for himself.
From the article: VDSL technology has an aggregate capacity of up to 60 Mbits/second over short distances...
DSL is a pipedream until the distance-sensitivity problem is solved. I also read (I think it was on C|Net) recently that there are a lot of complaints about poor implementations (braindead admins, most likely) and less-than-acceptable throughput.
For what it's worth, SWB just rolled out ADSL in our area (NE Oklahoma) and we're about 2k ft. too far from the C.O. SWB tell me that they will be upgrading to a technology that will allow them to move the service points out closer to the customers some time next year, but I'm not holding my breath. They keep promising cheap, fat pipes, but we're still stuck with $150 a month ISDN (128K) or $1,000+ a month T1 (yes, I know T1 is sym. therefore better) if we want bandwidth.
Is it just me, or are the telcos and telco/cable people (since AT&T swallowed up TCI) just stringing us along so they can squeeze every possible penny out of T1, etc. before they make consumer broadband a reality?
If you don't proclaim yourself moral, who else will do it for you?
Why, the arbiters of morality in the media, like Jon Katz, for instance.
(Jon: you need to relax, seriously.)
I'd rather have my children work on their Latin or a Perl script than play a video game, but if they're going to play a game, I'd rather it be one where they have the choice to play as an angel than one where they are limited to playing as some entity who can kill, maim and destroy without reason or consequence.
Old fashioned? Maybe, but aren't we supposed to teach our children that their actions have consequences and that we should strive to be "the good guys" in this life?
Our opinions of the software developers' evangelical bent (I only mention it because Mr. Katz seem to have a problem with it) aside, we should be happy that someone is trying to offer a choice. Isn't that what the OSS community is supposed to be about--choice?
What I find interesting is all those wall panels that don't seem to serve any useful purpose. Wouldn't it be interesting if those panels could be xterms?
The real watershed for BI was the release of Turbo Pascal 3.?? for $75.00 a pop. The TP 3.x distribution consisted of a shrinkwrapped paperback manual (I still have mine) with a single 360K floppy stuck in the middle. This came at a time when most PC compilers were going for $1,000.00 or so, and required royalty payments if you sold binaries linked against their libs.
I suspect that part of the reason that TPs sales were so high was the "cost of documentation" factor. Many otherwise honest folks, when faced with the decision whether to buy a license for a software package or "borrow" it, assume that they will have to pay at least $40-150 for third-party books if they really want to get any use out of the software. Since one had to spend roughly the same amount of money to get any value from TP no matter how it was acquired, I think a lot of people who would otherwise have copied a friend's diskette, decided to buy the package.
For what it's worth, BI's prices started to rise because PK got into a stupid OOP VaporWar with BG, which took a huge toll on BI. Buying Ashton-Tate and getting sidetracked into other things like an office suite competition didn't help either.
Personally, I didn't find learning the shell much more difficult than learning the Messy-DOS CLI. Maybe I'm just nutty like that. I wouldn't argue against temporary abstraction layers that would make it easier for newbies to get started, as long as they teach them how to use the CLI at the same time.
I'm not one of the type who think people shouldn't be allowed to touch UNIX if they don't know what they're doing, but I do think there is something to be said for an informed user base and the effect their input will have on the direction of future development. Growing a generation of folks who use Linux, but can't tell the difference between it and Windows or DOS would do little good for Linux.
IIRC, the problem when compiling with the latest gccs shows up as random lockups, not during the boot phase. FWIW, I've never had a problem with a kernel built with the latest gcc, maybe I'm just lucky.
The point is, Linux (and unices in general) gives you as much or as little control as you want over what users are allowed to do.
Based on my experiences supporting Winnders users ("Er, I deleted all the files in C:\lotus, now my typewriter thingy won't work!"), and with the potential for litigation if the wrong (politically incorrect) images appear on a workstation display, I'd lock down everything and only permit users to choose from an approved set of desktop configurations (e.g. benign tiled bitmaps, or company-oriented images).
Users would not be permitted to install their own programs, no way. This would open the door for too many potential problems.
Clamping down on user configuration may not make one popular in the lunchroom, but corporations don't spend millions of dollars on IT to make their admins popular in the lunchroom.
Thanks for the correction. I was just grepping through the docs looking for more on this issue; I could swear I read something there that explained the whole problem of large images and what bzImage does to solve it, but I haven't found it yet. Maybe it was on the kernel list?
Your boot image may be too large. This is why we now have the option to build a (smaller) bzip-compressed image, instead of a gzip-compressed image. At the end of the final make stage (make boot, make Zimage, etc.) the last few lines of output should tell you the size of the image, e.g.:
Root device is (3, 1) Boot sector 512 bytes. Setup is 3436 bytes. System is 515 kB
I believe the max is 1024kB, give or take a few, since the image has to fit in one MB (correct me if I'm wrong). At any rate, if your image is close to 1024kB, try either making a bzip-compressed image, with "make bZimage," or make more of your drivers modular.
If that doesn't fix the problem, post your "make the kernel" command and the last few lines of output.
1. Command shell that doesn't involve a lot of learning. "move" should be the command to move a file, "copy" should be the command to copy a file, "delete" and "remove" should remove a file. Joe Blow doesn't care that when all we had was 6 letter commands, using "rm" for delete a good idea. We don't have those limits any more, we shouldn't be limited by them. (My suggestion is to call this DOS, for Dumb Old Shell, and make it work much like the MS-DOS command line.)
alias copy="cp" alias move="mv" alias delete="rm"
We call this a shell, and it's smart enough that it can accomodate just about any user.
Ah, true, for mom n pop desktops, but in the corporate setting, the admin configures the desktop once, puts all the homes on a server, and users don't have to configure squat.
In addition, this type of setup makes adding new workstations very quick and easy.
Maybe this could be done with Winnders, but at what cost?
I've never had any of the problems you mention with GIMP. I just installed RH 6.1 with GNOME/E on a P75 Toshiba 400CS laptop and it actually works most of the time--much better than I would have guessed it would on a P75, even with some of the E bells and whistles turned on. It's never crashed and locked up the system (cough, Windows?) but the session management did blow up on me once when I tried to save a session. I still think the GNOME folks need to make their stuff easier to install and maintain and clean up some of their code, but it'll do for now.
KDE's alright, I guess, but I'm not much of DE guy anyway--I prefer Window Maker on my primary PC.
Linux--problems? Maybe, but it works, it doesn't crash, and I get my work done with it. Sounds like maybe you have a few configuration problems to work out--just my guess.
I looked at ECCO, but thought it had too many bells and whistles. I'd much rather see Commence (used to be IBM Current) ported to Linux. Commence 2.x has/had a much cleaner interface and is programmable, to a certain extent. Commence seems like a good model/starting point for an Open Source project.
And besides, you may get something 'free', but then it's mostly 'cheap' and of inferior quality. 'Free' and Opera don't go together - and never will.
The Opera folks clearly don't grok "Free" (speech) software. Just wait 'til mozilla's been out for a year or two--they'll grok then. I may buy a license anyway if I think it will be worthwhile for testing my HTML for compatability, but I don't think Opera will ever become very important unless they free the source. Not because I object to paying for a license, but because I don't think they will be able to keep up with mozilla.
Netscape 5.0 will be based on a totally new code base, not a rehash of the old bloated 4.x code. These complaints and speculation that 5.0 will be "worse" than 4.x versions are baseless.
For those who haven't been paying attention, the browser is evolving beyond a pr0n viewing tool--developers want to use browsers as application platforms. This is what MS and Netscape (to a lesser degree, I guess) have been pushing for years. IE and NS have been too broken to be really useful for this purpose, but a standards-compliant browser is what's called for, and mozilla and NS 5.0 will (AFAIK) finally begin to deliver.
In order to build a better browser, not just another crappy 4.x build, and attempt to become compliant and stable, mozilla.org had to start with a clean slate. That's why it's taken so long to get this far.
Contrary to published reports (including the CNET and Time articles), the developers aren't bailing out, and outsiders are participating.
The URLs have been posted here, go download a milestone release or a nightly build and try it out. Don't take the media BS at face value.
Any mozilla build you download now is going to be Alpha (pre-Beta, not "Compaq Alpha") software--in other words, don't whine about it being "buggy," it's there so you can work with it and report bugs, Ok?
Finally, get yourself a slashdot account and add the mozilla.org and mozillazine slashboxes to your preferences so you can keep up with what's going on.
Sorry for the rant, but I'm fed up with all the disinformation on mozilla. Even our key advocates have been quoted as trashing it recently, which really bothers me.
Slashdot Mirror
If you saw the doe-eyed little Microsofties on C-SPAN dutifully tossing warm and fuzzy bunny softball questions at Algore and telling him how they come to work every day just to make the world a better place, you know this must be an innocent error on Microsoft's part.
Does anyone else remember, I think it was in the mid-eighties, when a PC Mag. column published a rumor that Windows was doing something to progressively decrease the performance of Lotus 1-2-3 and eventually crash it? Deja vu?
Microsoft: Making the World a Better Place, one B.S.O.D. at a time.
Have you looked at the mozilla source? Your argument would some weight if you could cite specific problems with the source--but then, people would only say "Good job, you've found some problems; you're obviously smart enough, so go fix them." If you can't code, then use the nightly builds and report any bugs you find. There are all kinds of ways to help with the development effort. In the Open Source culture, there are complainers and there are contributors (a constructive criticism can be considered a contribution, by the way). Contributors are people who would rather make something happen than just sit back and carp about the state of the project.
Why do you think the mozpeople were so keen on re-designing from the ground up? Ease of maintenance. This means, making it less difficult for developers to dive in and work on the project, among other things like easier cross-platform implementation, i18n, etc. Unless the mozilla team have gone braindead, reasonably skilled developers should be able to break off small, digestible chunks of the code and work on them without having to grok the whole enchilada.
Take a deep breath and read my post again. I stated that "the future of Linux involves..." I did not say that the future of Linux is "dependent upon..." NASDAQ, etc. I did not, as you put it, "say otherwise."
Can you imagine a distribution that came with sources (as required by the GPL) yet couldn't be built except with the Borkland IDE?
You have got to be sleeping poorly to come up with any such scenario.
You came up with this scenario, not I. If anything, Borland have hinted that their compiler products will be very compatible, and why should't they be? There's no compelling reason I can think of, strategic or otherwise, for Borland to make their products incompatible or use non-standard switches, etc.
Maybe you're not old enough to remember the Turbo Pascal era. I am. It was an amazing time for folks who wanted to write their own software, and I think Borland can be credited with making PC software development a real cottage industry. Borland have proven that they can do compilers and can (occasionally) figure out where the marketplace is headed. Look at the Borland developer survey if you want to find out what developers are telling Borland about what they'd like them to do with development tools on Linux. If Borland are still as sharp as they used to be, they'll shrink wrap a drop-in replacement for egcs that (at least according to early benchmark reports) may very well produce tighter, faster code. If they give the Linux user a choice between a Free compiler that produces good code, and a $125.00 compiler that produces much smaller and faster code and requires no tweaking to build the kernel and/or projects packaged for autoconf, I think they may just repeat their compiler revolution.
Whence cometh the "hijacking" rhetoric? Who is hijacking whom? I see one company that is in business to make money, purportedly considering the acquisition of another company that is in business to make money.
This is a no-brainer. You're Red Hat, you've just been infused with tons of capital, you want to do things to improve the market penetration of Linux. How better to do this than to buy up the key Linux development tools vendor?
How many of you who are worried that RH are trying to become the Microsoft of Linux have considered what might happen if Borland repeat their 1980s DOS compiler coup on Linux? Before Borland's Turbo Pascal came along, a decent DOS compiler couldn't be had for less than several hundred to one thousand dollars; TP rolled out at ~$75 a pop, took the compiler market by storm. Suddenly, everybody was using TP to develop DOS apps. What if Borland's Linux compilers took the Linux market by storm and left Cygnus in the dust? How many of you would be complaining that RH weren't doing anything to preserve Cygnus and leaving us at the mercy of Borland?
Like it or not, the future of Linux involves NASDAQ, venture caps, and lots and lots of money. Technologies like "egcs" and embedded development tools that are important to the future of Linux will be acquired. What we should be asking ourselves is not, "Is it a good thing that someone may buy out Cygnus?" but "Which company would we prefer to take the reins at Cygnus? Turbo? Caldera? Red Hat? Microsoft?" I'd prefer that Red Hat end up with Cygnus since they've demonstrated that they are dedicated to the GNU concept. Some other vendors seem all too willing to market closed add-ons to Linux and would probably hurt Linux if they had control of Cygnus.
Right now, the Linux development tools space is where it's at, and this would seem like a wise move on Red Hat's part.
Slashdot on Mainstream Media on Slashdot and Microsoft
The U.S.A. have little to fear in the way of invasion--we can still count on two large bodies of water to make a direct assault on our shores too costly to consider. In the cyberwar context, however, we cannot count on any natural geographical barriers, and we may be at a significant disadvantage because of our dependence on less-than-secure technology and our "Cover Your Ass at All Costs" corporate and government cultures. The Information Superdirtroad leads right to the back door of almost every mission critical institution and enterprise in this country and, as slashdot readers know, few of those doors are securely locked.
I think it would be wise of our military to refrain from cyberwar until the overall quality of security on corporate and government networks is improved. We can count on the military to defend us against an attacking force on the ground, but on the 'net, we're all on the front lines and it's every man for himself.
see "The Terminal Man," an early Michael Crichton novel.
Also made into a film starring George Segal.
Scary stuff, I read it when I was a kid. It's one of those "technology is evil" plots and seems eerily germane.
From the article:
VDSL technology has an aggregate capacity of up to 60 Mbits/second over short distances...
DSL is a pipedream until the distance-sensitivity problem is solved. I also read (I think it was on C|Net) recently that there are a lot of complaints about poor implementations (braindead admins, most likely) and less-than-acceptable throughput.
For what it's worth, SWB just rolled out ADSL in our area (NE Oklahoma) and we're about 2k ft. too far from the C.O. SWB tell me that they will be upgrading to a technology that will allow them to move the service points out closer to the customers some time next year, but I'm not holding my breath. They keep promising cheap, fat pipes, but we're still stuck with $150 a month ISDN (128K) or $1,000+ a month T1 (yes, I know T1 is sym. therefore better) if we want bandwidth.
Is it just me, or are the telcos and telco/cable people (since AT&T swallowed up TCI) just stringing us along so they can squeeze every possible penny out of T1, etc. before they make consumer broadband a reality?
http://www.opendesk.com/Login/vslicens e.html
In a nutshell, they haven't decided on their final license yet. They state that whatever license they decide on, it will be OSI-friendly.
If you don't proclaim yourself moral, who else will do it for you?
Why, the arbiters of morality in the media, like Jon Katz, for instance.
(Jon: you need to relax, seriously.)
I'd rather have my children work on their Latin or a Perl script than play a video game, but if they're going to play a game, I'd rather it be one where they have the choice to play as an angel than one where they are limited to playing as some entity who can kill, maim and destroy without reason or consequence.
Old fashioned? Maybe, but aren't we supposed to teach our children that their actions have consequences and that we should strive to be "the good guys" in this life?
Our opinions of the software developers' evangelical bent (I only mention it because Mr. Katz seem to have a problem with it) aside, we should be happy that someone is trying to offer a choice. Isn't that what the OSS community is supposed to be about--choice?
su = iddqd
What I find interesting is all those wall panels that don't seem to serve any useful purpose. Wouldn't it be interesting if those panels could be xterms?
The real watershed for BI was the release of Turbo Pascal 3.?? for $75.00 a pop. The TP 3.x distribution consisted of a shrinkwrapped paperback manual (I still have mine) with a single 360K floppy stuck in the middle. This came at a time when most PC compilers were going for $1,000.00 or so, and required royalty payments if you sold binaries linked against their libs.
I suspect that part of the reason that TPs sales were so high was the "cost of documentation" factor. Many otherwise honest folks, when faced with the decision whether to buy a license for a software package or "borrow" it, assume that they will have to pay at least $40-150 for third-party books if they really want to get any use out of the software. Since one had to spend roughly the same amount of money to get any value from TP no matter how it was acquired, I think a lot of people who would otherwise have copied a friend's diskette, decided to buy the package.
For what it's worth, BI's prices started to rise because PK got into a stupid OOP VaporWar with BG, which took a huge toll on BI. Buying Ashton-Tate and getting sidetracked into other things like an office suite competition didn't help either.
Personally, I didn't find learning the shell much more difficult than learning the Messy-DOS CLI. Maybe I'm just nutty like that. I wouldn't argue against temporary abstraction layers that would make it easier for newbies to get started, as long as they teach them how to use the CLI at the same time.
I'm not one of the type who think people shouldn't be allowed to touch UNIX if they don't know what they're doing, but I do think there is something to be said for an informed user base and the effect their input will have on the direction of future development. Growing a generation of folks who use Linux, but can't tell the difference between it and Windows or DOS would do little good for Linux.
IIRC, the problem when compiling with the latest gccs shows up as random lockups, not during the boot phase. FWIW, I've never had a problem with a kernel built with the latest gcc, maybe I'm just lucky.
The point is, Linux (and unices in general) gives you as much or as little control as you want over what users are allowed to do.
Based on my experiences supporting Winnders users ("Er, I deleted all the files in C:\lotus, now my typewriter thingy won't work!"), and with the potential for litigation if the wrong (politically incorrect) images appear on a workstation display, I'd lock down everything and only permit users to choose from an approved set of desktop configurations (e.g. benign tiled bitmaps, or company-oriented images).
Users would not be permitted to install their own programs, no way. This would open the door for too many potential problems.
Clamping down on user configuration may not make one popular in the lunchroom, but corporations don't spend millions of dollars on IT to make their admins popular in the lunchroom.
If I may correct myself, it's "make bzImage," not "make bZimage."
Still suffering from early-morning caffeine deficiency syndrome.
Thanks for the correction. I was just grepping through the docs looking for more on this issue; I could swear I read something there that explained the whole problem of large images and what bzImage does to solve it, but I haven't found it yet. Maybe it was on the kernel list?
Your boot image may be too large. This is why we now have the option to build a (smaller) bzip-compressed image, instead of a gzip-compressed image. At the end of the final make stage (make boot, make Zimage, etc.) the last few lines of output should tell you the size of the image, e.g.:
Root device is (3, 1)
Boot sector 512 bytes.
Setup is 3436 bytes.
System is 515 kB
I believe the max is 1024kB, give or take a few, since the image has to fit in one MB (correct me if I'm wrong). At any rate, if your image is close to 1024kB, try either making a bzip-compressed image, with "make bZimage," or make more of your drivers modular.
If that doesn't fix the problem, post your "make the kernel" command and the last few lines of output.
1. Command shell that doesn't involve a lot of learning. "move" should be the command to move a file, "copy" should be the command to copy a file, "delete" and "remove" should remove a file. Joe Blow doesn't care that when all we had was 6 letter commands, using "rm" for delete a good idea. We don't have those limits any more, we shouldn't be limited by them. (My suggestion is to call this DOS, for Dumb Old Shell, and make it work much like the MS-DOS command line.)
alias copy="cp"
alias move="mv"
alias delete="rm"
We call this a shell, and it's smart enough that it can accomodate just about any user.
Ah, true, for mom n pop desktops, but in the corporate setting, the admin configures the desktop once, puts all the homes on a server, and users don't have to configure squat.
In addition, this type of setup makes adding new workstations very quick and easy.
Maybe this could be done with Winnders, but at what cost?
I've never had any of the problems you mention with GIMP. I just installed RH 6.1 with GNOME/E on a P75 Toshiba 400CS laptop and it actually works most of the time--much better than I would have guessed it would on a P75, even with some of the E bells and whistles turned on. It's never crashed and locked up the system (cough, Windows?) but the session management did blow up on me once when I tried to save a session. I still think the GNOME folks need to make their stuff easier to install and maintain and clean up some of their code, but it'll do for now.
KDE's alright, I guess, but I'm not much of DE guy anyway--I prefer Window Maker on my primary PC.
Linux--problems? Maybe, but it works, it doesn't crash, and I get my work done with it. Sounds like maybe you have a few configuration problems to work out--just my guess.
I looked at ECCO, but thought it had too many bells and whistles. I'd much rather see Commence (used to be IBM Current) ported to Linux. Commence 2.x has/had a much cleaner interface and is programmable, to a certain extent. Commence seems like a good model/starting point for an Open Source project.
From the Opera URL listed in the parent:
And besides, you may get something 'free', but then it's mostly 'cheap' and of inferior quality. 'Free' and Opera don't go together - and never will.
The Opera folks clearly don't grok "Free" (speech) software. Just wait 'til mozilla's been out for a year or two--they'll grok then. I may buy a license anyway if I think it will be worthwhile for testing my HTML for compatability, but I don't think Opera will ever become very important unless they free the source. Not because I object to paying for a license, but because I don't think they will be able to keep up with mozilla.
Sorry for the rant, but I'm fed up with all the disinformation on mozilla. Even our key advocates have been quoted as trashing it recently, which really bothers me.
Start with the Access HOWTO:
http://metalab.unc.edu/Linux/H OWTO/Access-HOWTO.html
Then the Keyboard and console HOWTO:
http://metalab.u nc.edu/Linux/HOWTO/Keyboard