Basically, Netscape (for it is they!) have thrown away their old code, then used GTK+, and thrown that away (didn't have customisable look and feel, which is essential when none of the actual FEATURES work) and re-written everything again. The result doesn't work, and probably needs more work than just fixing GTK+, Xlib or even Motif, but it gave everyone something to do while pointing the blame desperately at each other. What a catastrophe.
A typical bug report in Bugzilla has commentry like this:
Bob: Mozilla breaks horribly when I do this. I do it a lot, and so does every other Mozilla user, someone should fix it.
Fred: Agreed. Marking this as M9, we need a fix soon
Pete: M9 is too close. Users can live without until M11
Dave: Put this down for D19 and FU2, I have two bodies on it now
Dave: Wait. Not our problem, this is Fred's.
Fred: I can't reproduce this. Marked Works for Me
Bob: Why isn't this fixed yet? It is easily reproducible, follow my URL
Dave: Nope. Can't reproduce
Bob: Did you click on the URL?
Dave: Oh, wait, now it does. I think this will be fixed when FlyingPig lands next week. Marking FP
FPTeam: Flying Pig is delayed, because it turns out we have to implement abstract classes not just pray. We didn't know that.
Bob: I don't understand, I have waited four months for this..?
Dave: This is now first priority for M15. We will definitely have this.
Fred: I can't commit to that. Call it M16 and we'll squeeze it in later
Harold: Just doing some admin. Nothing to see here, move along. (now M17)
FPTeam: Marked as fixed because FlyingPig landed today. FlyingPig is still pretty buggy. Please don't open more than one window or breathe near the computer. Sorry.
Bob: Nope, still not fixed as of M17, where is everyone? Why can't a simple bug be fixed in under twelve months?
---
Why is it so hard? I have lots of ideas, but none of them is very complementary toward the Mozilla team. I think an equally valid question is, "Why did Microsoft take nearly five years to complete NT 5?" and many of the answers are probably the same. Let this be a lesson to you, do not work on products which are supposed to solve everyones problems all at once.
Can someone fill in how far along *BSD's support for USB is? I see a lot of references to it, on Linux-USB, but everyone's rather vague about just how much of USB (which is a big wide spec) has landed in the various BSDs
Linux, I do know, will have HID (mouses, keyboards and stuff), modems, printers, some cameras, SCSI-like things and various other stuff in 2.4.0 when it arrives. Sound would be nice too, but that looks like a really eeevil specification.
ISTR That enough of Linux's USB support was available at 2.2.x time for USB based iMacs to be usable, but maybe my memory is blurry.
Oh wait, I remember, *BSD has more crazy fanatics, so you can feel more at home. And if someone asks "How do I delete a file?" you can call him lame, and tell him to buy a clue. The microkernel dig above from a supposed *BSD fan is hilarious "Our kernel has less features, and is therefore inherently better".
The *BSD kernels each have a lesson or two to teach Linux, but equally there are lessons they could learn from Linux, and from each other. It's not enough to be worth fighting over. IMHO today's *BSD userspace is poorer than the GNU userspace, and Linux is a better platform for bleeding edge kernel meddling.
I respect the OpenBSD crew for having the guys to actually audit code, and the adventurers doing "First Ports" to platforms like the old 68K Macs, but the trend towards fanatical hate of other Free Software projects leaves a very unpleasant taste, BSD users are pronouncing that a GPL'd branch is "Evil", while a MS proprietary branch is "Proof that BSD is truely free".
This inconsistent value set (and I don't clame it's present inside the holy temple of the core developers) is a rotten stinking thing inside the BSD community, and makes the "Anything but MS" sentiment found in some parts of the Linux camp seem healthy by comparison. While that rotten thing remains in the community I hold no hope of GNU/FreeBSD seeing the light of day.
The rationale is that they'd port GNU/Linux' userspace to BSD. Now since the userspace of GNU/Linux is GNU, that would definitely make GNU/FreeBSD.
A typical *BSD is mostly under BSD variant licenses and more recent 2 clause licenses, Debian's code would use the FreeBSD kernel, but mostly code from the GNU project. This is probably >50% GPL code, but even if it isn't, GNU/FreeBSD is a clear name for "What's in the box?"
You're talking to the wrong guys. KOffice is a long way from having viable MS Office importers for any of their sub-products. If you just need Excel and Word you are better off with Gnumeric and AbiWord, which have been aiming for Office compatability (to the extent that it's not an oxymoron) from the word go.
Sure, Abiword's not very integrated, and you won't be able to draw charts in the stable Gnumeric this week, but you WILL be able to load and edit your MS Office documents, which in many cases is the SINGLE killer feature.
I see. So you are discounting the many many 3rd party Windows support operations? Are you really saying that HP's windows support is no good? Or that the many large resellers have no idea what they are doing? Are you saying that ICL doesn't support Windows when it uses it in projects? There are far, far more people able to support NT than Linux, especially when 'support' means support of large, complex developments, rather than simply supporting a distribution, or providing general Unix Q and A style help.
A lot of suits, and worse a lot of MSCEs don't understand this. I hope every serious admin/ support guy does though, because it can cost your job.
The only company in the world which can legally help you out when there's a show-stopping bug in MS Windows, is Microsoft. Even a major source licensee of which there are very few, is forbidden from helping customers by writing and providing OS fixes.
To put this clearly - when Linus' kernel breaks, my options are open. If I want to I can pay a specialist to work on nothing but my problem until it is fixed -- I can take $50K and spend it to fix a $1M problem.
When MS Windows breaks, I must put in a call to MS -- even if my support contract is through IBM or HP, they will put the call in for me, their hands are tied. If it's a doozy, I may get a fix the next day, or the one after. If no-one cares, it might take all year. Bye bye $1M.
Don't let this happen to you. Don't buy "support" for something the supporting company knows no more about than you do. Don't buy black box proprietary software. Don't permit single points of failure. Free Software gives you OPTIONS, and in business that agility is worth money.
Of course not, who could bare to not have the most cutting edge, up to the minute stuff in their stable release...
Oh wait, right now Debian is shipping stuff out of the Arc! There will ALWAYS be at least one vital, exciting brand new app sitting in "Incoming" just beggin for you to delay freeze. Experience tells you when to say "No more", and close the doors -- no matter who is left out in the cold. If you refuse to learn this lesson, cancel potato/ whatever and tell Debian users there will be no more stable releases.
Here (Southampton UK) they used to give people Slack, then later Red Hat, and now they don't include a Linux distro on the undegrad CD... ... because undergrads prefer to write the latest distros from FTP using the CDR machine. There's even an icon "Write latest Red Hat CD" on the desktop of that machine
Unfortunately, the CDR machine still runs WinNT, but you can't have everything (yet)
I don't have a DXR2, but looking at the source it seems Macrovision and similar protection is an on-off switch in the card firmware... I would be very surprised if there isn't a way to disable Macrovision on such old kit, but to be quite honest, who cares? Do you really want to wreck your crystal clear DVD video by putting it on nasty VHS?
"NT5... when it finally ships in the middle of 1998" Computing, 6 November 1997
"I really think we're going to hit the beginning part of next year" Steve Ballmer, quoted ZDNet, 8 June 1998
That's right, two years ago the press were telling readers that they'd see NT5 LAST YEAR Last year, Steve Ballmer told ZDNet he expected Windows 2000 on shelves early in 2000. It's all vapour until you get the CD in your sticky fingers.
Nah, no particular reason. Most people are using Apache, some people have a Windows machine, and they just run what came in the box.
"MS provides solutions" ? You must be thinking of some other company. Microsoft is one the VERY FEW software companies which has a significant role in the computer industry. All the other major players are working with hardware, people or true solutions. IBM, Compaq, Sun, etc. don't give a damn how much an OS costs -- the only people trying to jack up the price are MS.
"MS provides value" ? This one is definitely wrong, they're providing COST, not value, there's no question that what was decided here was the COST of W2K to the end-user. If you look at MS prices over the past 5 years, you'll see that they are IN NOW WAY linked to the value of the software, the cost is decided purely on maximising profit while preserving clear leadership in sales.
The CSS stuff (which is where this breakthrough was) is NOT PATENTED If they had PATENTED it, we would have had working (but illegal) stuff last year. Instead, they kept it a TRADE SECRET. That has no legal protection, but it means it took an extra 12 months or so to get free code which works.
For the player, no-one can legally make a FREE (as in Beer or as in Speech) player. You have to pay per-copy fees for at least some of the component technologies (AC3 comes to mind) in a player.
However, just because it is ILLEGAL to make a free player, doesn't mean no-one has:) Anyway, the key breakthrough (CSS) means there's no major difference between "DVD Video on Linux" and "MPEG 2 on Linux", except a few silly add-on features. I can live without subtitles if I have to.
Wrong freedoms people. GPL is freedoms for the USER - the programmer already had all their freedom when they wrote and licensed the code...
The "I'm alright Jack", attitude from the *BSD camp is a _developer's_ attitude, and a snobby one at that (most particularly for BSD-licensed Motif apps, what a hypocritical decision) By using GPL I give ALL MY USERS FREEDOM. Even if I never meet them, even if they don't realise the software was mine, they will benefit from my freedom. This is the best possible freedom I can give software.
Why should I help BSDi and others of similar ilk to make money for their shareholders when users are crying out for free software?
Don't get me wrong, X-style licenses have their place, but it is not in applications programming. It doesn't preserve enough freedom for the eventual USER of the software. When I release X/BSD code for a library, the sample implementation is ALWAYS a GPL application. Proprietary programmers can write their own damn examples, Users don't necessarily have that choice.
SSL is an "opened" standard, it was developed by Netscape, but they recognised many moons ago that to get wide acceptance you need Open Standards. So they told everyone how to do SSL, went through the process and got the standard out there. It's a good standard (in comparison to a lot of stuff on the web) so it won. As patent problems go, this is far from the worst: RSA have reasonable terms, the patent runs out soon, and it's not valid in most of the world anyway.
If SSL had been designed from scratch as an open standard, I'm sure SSL wouldn't include RSA but rather an equivalent but free algorithm. Still, as MPEG members would tell you a non-free standard is better than no standard at all.
1. It is FREE SOFTWARE, fit your own menu system if you really think that's what Gimp needs, and if others agree they'll use your version not the main tree.
2. The "normal menu system" of which you speak doesn't fit Gimp's way of working. Something like it MAY be an option in a later Gimp, but I for one won't be using it. There's just no space on a 200x300 image to have a menu with 10 headings in it. So that means either the menu shrinks/ is truncated, or it floats separately. That gets you MDI, which is probably the worst thing Microsoft ever gave the GUI.
KDE folks floated this when they proposed a Qt2.0-based Gimp. As expected they spent a week or so slagging of Gimp, then started a project (now dead) called KImageShop to prove us wrong. What a waste of everyone's time:(
PS No offense to KDE developers actually _developing_ something, it's the vapourware KImageShop people who annoyed me.
Here's the deal with Sorenson, and why it IS fair to whine at Apple, and perhaps even whoever made this video available...
XAnim's author has always made NDA deals with codec technology owners to create DECODERS for their formats on Linux. The existence of these decoders promotes the format, makes it easier to justify using it, and COSTS NOTHING
But with Sorenson, Apple _REFUSED_ to permit a deal of this sort. They feel that Sorenson is SO GREAT that people will throw away their SGI, their E450 and their Linux Alpha machine, and buy a PowerMac just so they can watch video. What's even MORE stupid IMHO is that there are people out there who PAID Apple good money to create videos which Apple WILL NOT LICENSE for use by non-Mac/ non-Windows users.
That's right -- someone paid money (probably as part of some other software) to MASSIVELY REDUCE THEIR AUDIENCE If that's not stupid I don't know what is.
I like RH a lot, but I'd heard over the last six months or so that other distributions had much better out-of-box experiences for new users...
I saw someone installing Caldera 2.2 (and Petreley doesn't ever make it clear that its improved for 2.3) and they were having a hell of a time. The GUI looked nice, but he wasn't making progress. But I can't make an in-depth criticism of COL 2.2 because I didn't take time out to walk them through it, maybe they would've had just as much trouble with Win98
SuSE OTOH I really must warn people about. If you've installed SuSE 6.2 and you think you got a "Smooth" install process, you need to take a better look at the competition. Twelve months ago, I installed SuSE and it had a pretty awful install process, the rediculous Yast tool making every task a tangle of fingers and thumbs. What a mess! I installed RH5.x soon after, and it was hardly better, although at least it didn't have Yast. Since then though, Red Hat's RH6.1 makes it all better with a sweet GUI installer -- but when I helped install SuSE for a friend last week, it was STILL JUST AS BAD AS IN 1998. SuSE can't afford to sit on their asses like this. Throw away Yast and use one of the BETTER, FREE installation tools from another vendor.
This version of Gnome is very much better than 1.0 Certainly that is in part because 1.0 was premature, but it's also because more people have had an opportunity to contribute now.
Packaging is an issue of flexibility versus simplicity. The KDE crowd believe (rightly or wrongly) that most users will be happy with "All in the box" packaging, while Gnome believes in exposing individual optional parts.
So, I'd certainly say KDE is easier to install manually, but GNOME wins for me because I happen to be picky, and I don't _want_ zillions of unnecessary components included in one huge package. I'll probably install about half of October Gnome, and never miss anything I don't install.
Users getting all this on CD won't care either way (In RH6.1 it was just one button press Gnome vs KDE)
Actually PotatoShop's requirements on the System Swap don't increase when working with large images, IIRC PotatoShop, like Gimp on Unix use a separate disk file to provide their own virtual memory for image tiles. I suspect that since PotatoShop is relying solely on WinNT they have a slight performance edge there.
The point in general is well made, it's just that big image editing apps don't expect their 240Mb of swap to come from the OS, especially on a 32Mb computer:)
PATHs in NT don't behave like the ones in Unix, for a variety of reasons. The result is that you can't have the complete control you'd like (and sometimes need) over the search path for executables and libraries.
Most of this isn't Microsoft's fault, except in as much as you could condemn them for writing a single user, non-networked OS which still uses Real Mode in 1998... If software written in the late 90s had been designed for NT rather than 95 you'd see a lot less of these odd dependency problems. Unix (at least the Unices I've used) still have better shared object support than NT though.
No, it means you don't understand the first thing about system security. There ARE options for Linux to have encrypted file store protected by a password on startup, but like their equivalents in NT they are rarely used. The true way to prevent direct physical attacks (like inserting a boot floppy) is to prevent unauthorised people from having physical access to the machine. The second level of defence (which is used in the Undergrad lab below me) is to deactivate the non-HD boot facility and alarm the cases.
Regardless this is off-topic, the AC was talking about how a LEGITIMATE administrator would access a machine. Physical security doesn't apply to the legitimate admins
I'm sure Bruce already knows this, but there ARE reasons to choose not only the LGPL, but also BSD and other licenses.
Of course if Free Software is your one and only goal, GPL is the way to go, but some projects have more complicated goals. The PNG development team want first and foremost to promote the use of PNG instead of GIF or BMP or whatever. This purpose is best served by making it legal for EVERYONE to use their code.
I'd encourage anyone seeking wide-ranging support for a new protocol or format to release under LGPL or 2-clause BSD. If you support Free Software you can make a GPL'ed application which demonstrates the potential of your format, but your reference implementation should not be GPL if you want industry-wide support.
NB Despite FREE availability of libpng, Adobe spent many developer hours on an inferior PNG codec for PotatoShop. Sometimes proprietary software strives to be WORSE:)
I'm waiting to be stomped, SniggleBunny. We've sure got a sackful of features (the alternative I suppose would be to hang around in #warez all week scrounging Kai's add-ons, since PS ships with hardly any good add-ons) The interface is only "hostile" if you're insistent on treating Gimp like MS Paint. IF you want a kiddy finger-paint package, write your own, this is a serious job and it has a serious tool. Much of the more glaring inconsistency is gone in the forthcoming Gimp 1.2 (try 1.1.10 now to see what's coming), and If you're one of the zillions who can't find their right mouse button (doesn't Windows have one of those these days?) Gimp 1.2 adds an explicit GUI menu selector too.
For those of us who've tried PS recently after using Gimp, it's annoying to have the One-Picture, One-Menu, One-Process method. I'd like to get some WORK done here. Of course, since it's Free Software you're free to fix all of your perceived problems and see if TigerT et al prefer your "easy" version.
Humbug. I heard this story five years ago, and I'll be hearing it five years from now...
"You Free Software people can't build a usable UNIX replacement" Oh, wait, we did - GNU/Linux, a modern cross-platform POSIX OS
"You Free Software people can't build GUI tools for end users" Oh, wait, we did - Gimp, used by plenty of experienced digital artists and newbies alike.
There's always someone who says the next mountain is an impossible climb, and they've been wrong EVERY TIME so far. Why would it be different this time?
Slashdot Mirror
Basically, Netscape (for it is they!) have thrown away their old code, then used GTK+, and thrown that away (didn't have customisable look and feel, which is essential when none of the actual FEATURES work) and re-written everything again. The result doesn't work, and probably needs more work than just fixing GTK+, Xlib or even Motif, but it gave everyone something to do while pointing the blame desperately at each other. What a catastrophe.
A typical bug report in Bugzilla has commentry like this:
Bob: Mozilla breaks horribly when I do this. I do it a lot, and so does every other Mozilla user, someone should fix it.
Fred: Agreed. Marking this as M9, we need a fix soon
Pete: M9 is too close. Users can live without until M11
Dave: Put this down for D19 and FU2, I have two bodies on it now
Dave: Wait. Not our problem, this is Fred's.
Fred: I can't reproduce this. Marked Works for Me
Bob: Why isn't this fixed yet? It is easily reproducible, follow my URL
Dave: Nope. Can't reproduce
Bob: Did you click on the URL?
Dave: Oh, wait, now it does. I think this will be fixed when FlyingPig lands next week. Marking FP
FPTeam: Flying Pig is delayed, because it turns out we have to implement abstract classes not just pray. We didn't know that.
Bob: I don't understand, I have waited four months for this..?
Dave: This is now first priority for M15. We will definitely have this.
Fred: I can't commit to that. Call it M16 and we'll squeeze it in later
Harold: Just doing some admin. Nothing to see here, move along. (now M17)
FPTeam: Marked as fixed because FlyingPig landed today. FlyingPig is still pretty buggy. Please don't open more than one window or breathe near the computer. Sorry.
Bob: Nope, still not fixed as of M17, where is everyone? Why can't a simple bug be fixed in under twelve months?
---
Why is it so hard? I have lots of ideas, but none of them is very complementary toward the Mozilla team. I think an equally valid question is, "Why did Microsoft take nearly five years to complete NT 5?" and many of the answers are probably the same. Let this be a lesson to you, do not work on products which are supposed to solve everyones problems all at once.
Nick.
Can someone fill in how far along *BSD's support for USB is?
I see a lot of references to it, on Linux-USB, but everyone's rather vague about just how much of USB (which is a big wide spec) has landed in the various BSDs
Linux, I do know, will have HID (mouses, keyboards and stuff), modems, printers, some cameras, SCSI-like things and various other stuff in 2.4.0 when it arrives. Sound would be nice too, but that looks like a really eeevil specification.
ISTR That enough of Linux's USB support was available at 2.2.x time for USB based iMacs to be usable, but maybe my memory is blurry.
The advantages were that great, right?
Like, um, errr, um, nope I'm stuck already.
Oh wait, I remember, *BSD has more crazy fanatics, so you can feel more at home. And if someone asks "How do I delete a file?" you can call him lame, and tell him to buy a clue. The microkernel dig above from a supposed *BSD fan is hilarious "Our kernel has less features, and is therefore inherently better".
The *BSD kernels each have a lesson or two to teach Linux, but equally there
are lessons they could learn from Linux, and from each other. It's not enough
to be worth fighting over. IMHO today's *BSD userspace is poorer than the
GNU userspace, and Linux is a better platform for bleeding edge kernel
meddling.
I respect the OpenBSD crew for having the guys to actually audit code, and the adventurers doing "First Ports" to platforms like the old 68K Macs, but the trend towards fanatical hate of other Free Software projects leaves a very unpleasant taste, BSD users are pronouncing that a GPL'd branch is "Evil", while a MS proprietary branch is "Proof that BSD is truely free".
This inconsistent value set (and I don't clame it's present inside the holy temple of the core developers) is a rotten stinking thing inside the BSD community, and makes the "Anything but MS" sentiment found in some parts of the Linux camp seem healthy by comparison. While that rotten thing remains in the community I hold no hope of GNU/FreeBSD seeing the light of day.
Nick.
The rationale is that they'd port GNU/Linux' userspace to BSD.
Now since the userspace of GNU/Linux is GNU, that would definitely make GNU/FreeBSD.
A typical *BSD is mostly under BSD variant licenses and more recent 2 clause licenses,
Debian's code would use the FreeBSD kernel, but mostly code from the GNU project. This is probably >50% GPL code, but even if it isn't, GNU/FreeBSD is a clear name for "What's in the box?"
You're talking to the wrong guys. KOffice is a long way from having viable MS Office importers for any of their sub-products.
If you just need Excel and Word you are better off with Gnumeric and AbiWord, which have been aiming for Office compatability (to the extent that it's not an oxymoron) from the word go.
Sure, Abiword's not very integrated, and you won't be able to draw charts in the stable Gnumeric this week, but you WILL be able to load and edit your MS Office documents, which in many cases is the SINGLE killer feature.
A lot of suits, and worse a lot of MSCEs don't understand this. I hope every serious admin/ support guy does though, because it can cost your job.
The only company in the world which can legally help you out when there's a show-stopping bug in MS Windows, is Microsoft. Even a major source licensee of which there are very few, is forbidden from helping customers by writing and providing OS fixes.
To put this clearly - when Linus' kernel breaks, my options are open. If I want to I can pay a specialist to work on nothing but my problem until it is fixed -- I can take $50K and spend it to fix a $1M problem.
When MS Windows breaks, I must put in a call to MS -- even if my support contract is through IBM or HP, they will put the call in for me, their hands are tied. If it's a doozy, I may get a fix the next day, or the one after. If no-one cares, it might take all year. Bye bye $1M.
Don't let this happen to you. Don't buy "support" for something the supporting company knows no more about than you do. Don't buy black box proprietary software. Don't permit single points of failure. Free Software gives you OPTIONS, and in business that agility is worth money.
Nick
Of course not, who could bare to not have the most cutting edge, up to the minute stuff in their stable release...
Oh wait, right now Debian is shipping stuff out of the Arc!
There will ALWAYS be at least one vital, exciting brand new app sitting in "Incoming" just beggin for you to delay freeze.
Experience tells you when to say "No more", and close the doors -- no matter who is left out in the cold. If you refuse to learn this lesson, cancel potato/ whatever and tell Debian users there will be no more stable releases.
Here (Southampton UK) they used to give people Slack, then later Red Hat, and now they don't include a Linux distro on the undegrad CD...
... because undergrads prefer to write the latest distros from FTP using the CDR machine.
There's even an icon "Write latest Red Hat CD" on the desktop of that machine
Unfortunately, the CDR machine still runs WinNT, but you can't have everything (yet)
I don't have a DXR2, but looking at the source it seems Macrovision and similar protection is an on-off switch in the card firmware...
I would be very surprised if there isn't a way to disable Macrovision on such old kit, but to be quite honest, who cares?
Do you really want to wreck your crystal clear DVD video by putting it on nasty VHS?
Oops. Of course, Steve is saying early NEXT YEAR in 1998, he means 1999 not 2000 (and so did I, must learn to proof read)
"NT5 ... when it finally ships in the middle of 1998"
Computing, 6 November 1997
"I really think we're going to hit the beginning part of next year"
Steve Ballmer, quoted ZDNet, 8 June 1998
That's right, two years ago the press were telling readers that they'd see NT5 LAST YEAR
Last year, Steve Ballmer told ZDNet he expected Windows 2000 on shelves early in 2000.
It's all vapour until you get the CD in your sticky fingers.
Nah, no particular reason. Most people are using Apache, some people have a Windows machine, and they just run what came in the box.
"MS provides solutions" ?
You must be thinking of some other company. Microsoft is one the VERY FEW software companies which has a significant role in the computer industry.
All the other major players are working with hardware, people or true solutions. IBM, Compaq, Sun, etc. don't give a damn how much an OS costs -- the only people trying to jack up the price are MS.
"MS provides value" ?
This one is definitely wrong, they're providing COST, not value, there's no question that what was decided here was the COST of W2K to the end-user.
If you look at MS prices over the past 5 years, you'll see that they are IN NOW WAY linked to the value of the software, the cost is decided purely on maximising profit while preserving clear leadership in sales.
You're way off :)
:)
The CSS stuff (which is where this breakthrough was) is NOT PATENTED
If they had PATENTED it, we would have had working (but illegal) stuff last year. Instead, they kept it a TRADE SECRET. That has no legal protection, but it means it took an extra 12 months or so to get free code which works.
For the player, no-one can legally make a FREE (as in Beer or as in Speech) player. You have to pay per-copy fees for at least some of the component technologies (AC3 comes to mind) in a player.
However, just because it is ILLEGAL to make a free player, doesn't mean no-one has
Anyway, the key breakthrough (CSS) means there's no major difference between "DVD Video on Linux" and "MPEG 2 on Linux", except a few silly add-on features. I can live without subtitles if I have to.
Wrong freedoms people. GPL is freedoms for the USER - the programmer already had all their freedom when they wrote and licensed the code...
The "I'm alright Jack", attitude from the *BSD camp is a _developer's_ attitude, and a snobby one at that (most particularly for BSD-licensed Motif apps, what a hypocritical decision)
By using GPL I give ALL MY USERS FREEDOM. Even if I never meet them, even if they don't realise the software was mine, they will benefit from my freedom. This is the best possible freedom I can give software.
Why should I help BSDi and others of similar ilk to make money for their shareholders when users are crying out for free software?
Don't get me wrong, X-style licenses have their place, but it is not in applications programming. It doesn't preserve enough freedom for the eventual USER of the software.
When I release X/BSD code for a library, the sample implementation is ALWAYS a GPL application. Proprietary programmers can write their own damn examples, Users don't necessarily have that choice.
SSL is an "opened" standard, it was developed by Netscape, but they recognised many moons ago that to get wide acceptance you need Open Standards.
So they told everyone how to do SSL, went through the process and got the standard out there. It's a good standard (in comparison to a lot of stuff on the web) so it won.
As patent problems go, this is far from the worst: RSA have reasonable terms, the patent runs out soon, and it's not valid in most of the world anyway.
If SSL had been designed from scratch as an open standard, I'm sure SSL wouldn't include RSA but rather an equivalent but free algorithm. Still, as MPEG members would tell you a non-free standard is better than no standard at all.
1. It is FREE SOFTWARE, fit your own menu system if you really think that's what Gimp needs, and if others agree they'll use your version not the main tree.
:(
2. The "normal menu system" of which you speak doesn't fit Gimp's way of working. Something like it MAY be an option in a later Gimp, but I for one won't be using it.
There's just no space on a 200x300 image to have a menu with 10 headings in it.
So that means either the menu shrinks/ is truncated, or it floats separately. That gets you MDI, which is probably the worst thing Microsoft ever gave the GUI.
KDE folks floated this when they proposed a Qt2.0-based Gimp. As expected they spent a week or so slagging of Gimp, then started a project (now dead) called KImageShop to prove us wrong. What a waste of everyone's time
PS No offense to KDE developers actually _developing_ something, it's the vapourware KImageShop people who annoyed me.
Here's the deal with Sorenson, and why it IS fair to whine at Apple, and perhaps even whoever made this video available...
XAnim's author has always made NDA deals with codec technology owners to create DECODERS for their formats on Linux.
The existence of these decoders promotes the format, makes it easier to justify using it, and COSTS NOTHING
But with Sorenson, Apple _REFUSED_ to permit a deal of this sort. They feel that Sorenson is SO GREAT that people will throw away their SGI, their E450 and their Linux Alpha machine, and buy a PowerMac just so they can watch video.
What's even MORE stupid IMHO is that there are people out there who PAID Apple good money to create videos which Apple WILL NOT LICENSE for use by non-Mac/ non-Windows users.
That's right -- someone paid money (probably as part of some other software) to MASSIVELY REDUCE THEIR AUDIENCE
If that's not stupid I don't know what is.
I like RH a lot, but I'd heard over the last six months or so that other distributions had much better out-of-box experiences for new users...
I saw someone installing Caldera 2.2 (and Petreley doesn't ever make it clear that its improved for 2.3) and they were having a hell of a time. The GUI looked nice, but he wasn't making progress.
But I can't make an in-depth criticism of COL 2.2 because I didn't take time out to walk them through it, maybe they would've had just as much trouble with Win98
SuSE OTOH I really must warn people about. If you've installed SuSE 6.2 and you think you got a "Smooth" install process, you need to take a better look at the competition.
Twelve months ago, I installed SuSE and it had a pretty awful install process, the rediculous Yast tool making every task a tangle of fingers and thumbs. What a mess!
I installed RH5.x soon after, and it was hardly better, although at least it didn't have Yast.
Since then though, Red Hat's RH6.1 makes it all better with a sweet GUI installer -- but when I helped install SuSE for a friend last week, it was STILL JUST AS BAD AS IN 1998.
SuSE can't afford to sit on their asses like this. Throw away Yast and use one of the BETTER, FREE installation tools from another vendor.
This version of Gnome is very much better than 1.0
Certainly that is in part because 1.0 was premature, but it's also because more people have had an opportunity to contribute now.
Packaging is an issue of flexibility versus simplicity. The KDE crowd believe (rightly or wrongly) that most users will be happy with "All in the box" packaging, while Gnome believes in exposing individual optional parts.
So, I'd certainly say KDE is easier to install manually, but GNOME wins for me because I happen to be picky, and I don't _want_ zillions of unnecessary components included in one huge package.
I'll probably install about half of October Gnome, and never miss anything I don't install.
Users getting all this on CD won't care either way (In RH6.1 it was just one button press Gnome vs KDE)
Actually PotatoShop's requirements on the System Swap don't increase when working with large images, IIRC
:)
PotatoShop, like Gimp on Unix use a separate disk file to provide their own virtual memory for image tiles. I suspect that since PotatoShop is relying solely on WinNT they have a slight performance edge there.
The point in general is well made, it's just that big image editing apps don't expect their 240Mb of swap to come from the OS, especially on a 32Mb computer
PATHs in NT don't behave like the ones in Unix, for a variety of reasons.
The result is that you can't have the complete control you'd like (and sometimes need) over the search path for executables and libraries.
Most of this isn't Microsoft's fault, except in as much as you could condemn them for writing a single user, non-networked OS which still uses Real Mode in 1998...
If software written in the late 90s had been designed for NT rather than 95 you'd see a lot less of these odd dependency problems.
Unix (at least the Unices I've used) still have better shared object support than NT though.
"This makes a Linux box unsecure"
No, it means you don't understand the first thing about system security.
There ARE options for Linux to have encrypted file store protected by a password on startup, but like their equivalents in NT they are rarely used.
The true way to prevent direct physical attacks (like inserting a boot floppy) is to prevent unauthorised people from having physical access to the machine.
The second level of defence (which is used in the Undergrad lab below me) is to deactivate the non-HD boot facility and alarm the cases.
Regardless this is off-topic, the AC was talking about how a LEGITIMATE administrator would access a machine. Physical security doesn't apply to the legitimate admins
I'm sure Bruce already knows this, but there ARE reasons to choose not only the LGPL, but also BSD and other licenses.
:)
Of course if Free Software is your one and only goal, GPL is the way to go, but some projects have more complicated goals.
The PNG development team want first and foremost to promote the use of PNG instead of GIF or BMP or whatever. This purpose is best served by making it legal for EVERYONE to use their code.
I'd encourage anyone seeking wide-ranging support for a new protocol or format to release under LGPL or 2-clause BSD.
If you support Free Software you can make a GPL'ed application which demonstrates the potential of your format, but your reference implementation should not be GPL if you want industry-wide support.
NB Despite FREE availability of libpng, Adobe spent many developer hours on an inferior PNG codec for PotatoShop. Sometimes proprietary software strives to be WORSE
I'm waiting to be stomped, SniggleBunny.
We've sure got a sackful of features (the alternative I suppose would be to hang around in #warez all week scrounging Kai's add-ons, since PS ships with hardly any good add-ons)
The interface is only "hostile" if you're insistent on treating Gimp like MS Paint. IF you want a kiddy finger-paint package, write your own, this is a serious job and it has a serious tool.
Much of the more glaring inconsistency is gone in the forthcoming Gimp 1.2 (try 1.1.10 now to see what's coming), and
If you're one of the zillions who can't find their right mouse button (doesn't Windows have one of those these days?) Gimp 1.2 adds an explicit GUI menu selector too.
For those of us who've tried PS recently after using Gimp, it's annoying to have the One-Picture, One-Menu, One-Process method. I'd like to get some WORK done here.
Of course, since it's Free Software you're free to fix all of your perceived problems and see if TigerT et al prefer your "easy" version.
Humbug. I heard this story five years ago, and I'll be hearing it five years from now...
"You Free Software people can't build a usable UNIX replacement"
Oh, wait, we did - GNU/Linux, a modern cross-platform POSIX OS
"You Free Software people can't build GUI tools for end users"
Oh, wait, we did - Gimp, used by plenty of experienced digital artists and newbies alike.
There's always someone who says the next mountain is an impossible climb, and they've been wrong EVERY TIME so far. Why would it be different this time?