ACTON, Mass August 30, 2002 OSDN today announced a second expansion to its recent "black market" entry -- selling links to porn sites. "With our new partners, hotassgirls.com and gianttits.com, we expect to enter a new era of market dominance," said Slashdot General Manager CmdrTaco, "We expect to reach profitability six years earlier than expected." OSDN stock rose to a 3 cent high on the pinksheets following the announcement.
When that firewall blows, who's neck is on the line?...should be enough reason to give them all your employers hard earnt money
Perhaps I'm understanding you wrong, but this seems a little bit unprofessional. You're willing to spend tons of extra company money so that *you* won't be the one to be responsible if something goes wrong?
Are there still people out there willing to take responsibility for their work?
Win32 is sort of an open standard, if incomplete. MS has published documentation on it, and while they may not have defined behavior for all situations, you can follow the documentation.
The problem is that occasionally the documentation differs from Windows, and if it does, everyone only cares whether software works correctly with Windows, not whether it follows the documentation.
Finally, there is no "frozen" version of the documentation that serves as a standard for MS to conform to and release patches to bring their products up to spec with.
I'd call Win32 about as open a standard as Postscript -- we know how to follow the docs, but there's only one implementation that 95% of the people out there care about.
The same is becoming true with HTML and Internet Explorer -- the HTML spec could say one thing, but authors are going to primarily care about compatibility with the leading implementation of that spec, rather than exact compliance with the spec.
People are just taking advantage of the medium, as they did with cartoons.
You have a good point.
If games with a message are infantile, then perhaps enjoyablebooks with a message suffer from the same complaint -- why aren't you reading a bland, straight political pamphlet if you want a "mature" medium?
Some of the (very little) bit of documentation on OS/400 that I could find on the Internet implied this -- not that one had to keep commands upper case, but that the system did so automatically. I'm used to an OS that distinguishes between case in commands, so this is a bit of a drawback for me.
I could be wrong -- this is some of the only OS/400 documentation that I could *find* freely available. It was a chapter excerpted from a book.
I was kind of interested in learning about OS/400, so I wanted to read up a bit about it. There was very little freely available information, though. This was terribly frusterating to me, because there's plenty of UNIX information available on the Internet.
I hold IBM somewhat responsible for this -- it costs them very little to give out information to the Internet at large, and yet they do not do so. Unless you're a shop prepared to drop quite a chunk of change on propriatary hardware, IBM salespeople aren't going to be very interested in giving you information -- and even if I was, I'd want my evaluation of the system to be based on my *own* interpretation of what I can read about the system, not on what I'm fed by the salespeople ("99.99% uptime! Foo Technology! IBM 24/7 support!"). Oh, yes. IBM does sell (*very expensive*) classes/seminars (oriented towards training to perform specific operations, from what I've seen) to teach people about OS/400. So if you're willing to drop quite a bit of money to learn about an OS that you may not want to use, you can get some information. However, others end up SOL. And I'm in the latter class.
Sorry, but it is impossible to pass judgement on OS/400 if you have no experience with it. It's a COMPLETELY different way of thinking about operating systems, and it's simply not comparable with Unix or Windows.
Fair enough.
I come from a Linux/Windows background, and I'm still trying to get my head around the/400 mindset after working with it for a few months.
That statement, though, also illustrates what I'm frusterated with./400 is *not* all that accessable to most people.
Also, there's plenty of good documentation and help for UNIX systems available on the Internet -- this is not true (unless I'm missing something) for OS/400....we don't have that sort of ignorance here on Slashdot...right?...right?
It's ignorance that gives us conversations. If we already all knew everything, there'd be no point in talking to each other -- Slashdot wouldn't exist.:-)
Do you think "ps -auxw" is anything less cryptic than the example you gave.
You make a good point -- I do take this for granted. However, this is also one of the more cryptic UNIX commands. Hmm. I think I should rephrase this. There are a lot of UNIX admins running around (compared to OS/400), so for better or for worse, a Solaris guy can get along pretty well on Linux. The commands are well-known by quite a few people.
On OS/400, however, the situation is worse, even assuming an equivalent level of "crypticness" because many people have to learn the OS for the first time.
The obvious attack against me is "Well, there are more MCSEs than UNIX admins, so UNIX also takes a penalty for being cryptic." I *agree* with this -- but UNIX is also being quite accomodating (because it has to be) to that majority of users, making available to them Windows-like environments like GNOME, and GUI administrative tools.
So OS/400 should have to do *more* work to be appealing to the average admin than Windows or even UNIX does...and it doesn't seem to be that it does.
Finally, I stand by my complaint that the programming environment is less than impressive -- the languages are really old. OS/400 programmers *might* just have have a deep love for COBOL or something, but it would really frusterate me.
You are right about the legacy custom apps, but that doesn't excite me about OS/400 -- it just makes me think "People *have* to use OS/400" rather than that "People *want* to use OS/400". It's the same frusteration I see when someone's written something for Win32 or the MacOS that I'd like to use on UNIX (or, occasionally, some other variant of UNIX).
Okay, my posts here are starting to sound like trolls, but I'm going to keep working at this point.
I've heard about a lot of systems billed as being "reliable". Windows folks call Windows 2k "reliable" and say they never have a problem with their servers, that they've been up for months. Linux folks call Linux more reliable than Windows, and say the same thing about their servers. Ditto for the BSD guys, the Sun guys, and evidently for the IBM guys.
So here's what I want to know -- I'll give you that the system is reliable, but that's only meaningful in the presence of unreliable alternatives. Are you claiming that an AS/400 box running OS/400 is more reliable than a BSD box, or a Solaris box, or a Linux box? I've never seen any of these just spontaneously fail or kernel panic...
I suspect that if AOL was ported to a PDA, people would be quite happy (smaller computer, same functionality), and AOL would avoid the must-run-on-Windows-so-we-must-suck-up-to-MS issue.
So I'm a little biased -- I've never sat down and used OS/400. But I did read through a bit of documentation, and I was thoroughly disgusted with what the OS sounds like. All commands are typed in in caps, in condensed combinations of verbs, then become lower case?
WRKSYSCTX is "work within the system context"?
The directory structure was something like an Apple II, the languages available were from the *Dark Ages*, and the prices obscene.
Why on earth would anyone *ever* want to use OS/400 over *any* variant of UNIX?
What on earth are you talking about? WINE isn't illegal in the least! What law do you think they're breaking?
Trademark? It'd be tough to show that WINE infringes on "Microsoft Windows".
Patent? I really doubt it. I can't imagine that there is any patent that could keep you from implementing an API.
Copyright? This was already tried by MS, arguing that they owned the header files, and duplicating the information in them was infringing. It didn't work.
EULA for reverse engineering? The WINE guys at least put up a pretense of clean-room engineering, and I think it'd be hard for MS to prove otherwise. You really don't need to disassemble Windows to implement the Win32 API as documented by MSDN. Finally, the right to reverse engineer is expressly granted in many locales for the purpose of ensuring compatibility. In the EU, this would make even a non-clean room impementation okay. In the US, it's a little more dicy, as I believe the laws only apply to compatibility over a network protocol between two different hosts, but there's still a general trend towards allowing people to produce compatible products.
Look-and-feel? Look-and-feel was an approach specifically shown by MS to not be a valid case for suit in the US legal system when they butted heads with Apple.
MS would have a *hell* of a time trying to prove that no one could implement a compatible product, which they'd have to do to nail the WINE guys. The antitrust guys would have a field day on MS.
Finally, WINE is not an emulator. "Emulator" has a specific meaning in the computer world -- it would reimplement the hardware that the software runs on. This would almost certainly cause a performance hit. WINE carries no such required overhead. At least in theory, WINE can run just as fast as (heck, faster) than Windows.
Given your AC nature, I'd almost say that you're trolling, but I can't quite be sure.
And screw obscurity, it was a good game. There are a lot of "obscure" games that are really, *really* good -- Total Annihilation is *still* better (much more powerful order-giving abilities) than the current crop of RTSes. Fallout is a top-notch RPG. The Last Express was beautifully done. There are a bunch of really good Q3 mods that were overshadowed by Counterstrike, likek Team Fortress and Weapons Factory.
People should realize that a good chunk of games prosper based very much on their marketing. Blizzard, for instance, throws a *huge* amount of money into marketing their games -- it's very difficult for the competition (which I've often found to be superior) to compete. People buy games based on screenshots in magazines, based on the overwhelming number of ads they see, and on how compelling the box is. Few people play a game for a while before buying it.
I certainly see your point, but I don't think that fun and other people making money are entirely incompatible.
I send in patches to Linux software that are packaged and used to make lots of money for RH, Mandrake, and others. The attitude I have is "As long as other techies can grab this for free, I certainly don't have any problem with RH making a buck off of making it accessable to other users." It's kind of a trade, too -- I make patches, improve their product (and have fun in the process), and they give me an ever-improving free product for download.
I do want to say thank you, a big thank you, to isolation for his mingw and WINE work. I use both frequently (mingw in two commercial settings), and I deeply appreciate them. They've let me and others escape the MS monopoly and still get work done, and they are both technically impressive pieces of software.
I do sort of wish that WINE could have stayed BSD-licensed -- there were a lot of people pretty comfortable with it, and developer groups and companies were relying on that license. I, in general, prefer the GPL (well, the LGPL) to the BSD license, but in this case, I'm kind of sad about the switch. It fragmented the WINE developer community and started a lot of fights.
OTOH, I also have to take my cap off to Transgaming -- the gentlemen there are taking enormous personal financial risk (throwing their own money that they can't really afford) into trying to make a commercially viable company that gives away source to a product that lets you play Windows games. I know one Linux user who has only a single Windows game that he wants to play -- Max Payne -- and Transgaming has let him do that.
Anyway, every line of code put out there, by anyone, is helping an awful lot of people. It's also pulling off some darn impressive technical tricks -- WINE is one of the few things that really blows my mind -- I'm amazed that the developers pulled it off. Here's to more coding -- and less politics.:-)
The reason some things do not work is that you have to implement more than just the documented externals.
This is partly true, but not the whole story. Yes, there have been times when either MS bugs have to be reproduced and emulated, or undocumented behavior discovered. However, this really isn't the cause of most incompatibilities.
First, the WINE team is limited -- hardly as many man-hours per weak as the Windows team at MS. So they take a, as they put it, "product-driven" approach. They take a specific program, and implement just enough to get it working properly. Few programs use (or will use for several years) WinXP-specific features, because it would limit their potential market. Same goes for Win2k-specific features. So the WINE guys don't bother with emulating those. Also, less crucial and rarely used fuctions are often just stubs, meaning that software that uses lots of esoteric options/functions is much more likely not to work.
Last of all, rarely used chunks of Win32 are simply ignored. I believe that there is basically no CryptoAPI support, for instance, because implementing CryptoAPI would be a significant amount of work, and very, very few programs would actually use it.
A lead Nvidia hardware guy came and gave a talk at our university. I asked him whether Nvidia preferred using DirectX over OpenGL because of more supported features in DX8 vs OGL 1.x, and he surprised me by saying that, no, in fact they significantly preferred OGL.
Apparently, when Nvidia adds a feature, OGL has a standard way of adding an extension to the language to support extra features. MS, despite serious lobbying from Nvidia, strongly pushes against supporting extensions (for obvious reasons, but it still doesn't go over well with Nvidia). So there may be a number of performance-enhancing features or tricks that are used with OGL (because the game developer did a bit of extra work to support the extensions) that are not used with DirectX, or have to be done partly in software with DirectX.
Anyway, the gist of this is that if you want to use all the features of your Nvidia video card, you're likely better off using OGL modes in your games.
I'm not sure what the take on this is at ATI or Matrox, though.
I disagree with the overhead statement. Some software seems to run more snappily, others less.
For example, Windows menuy widgets seem to operate much more slowly, but I've played Starcraft on the same machine in both WINE and Windows NT, and if Starcraft wasn't faster in WINE, it was at least as fast (admittedly, WinNT's DirectX probably wasn't as tweaked as newer releases, but even so...).
Let me guess, you're one of those people who freaks out hardcore when you find telnet in use, right.
Not if it's Kerberized and encrypted or ssl-tunneled.:-)
This is a fairly reasonable thing to "freak out" about -- there *have* been major break ins because people are using unencrypted telnet and someone seizes a crucial machine.
to be honest, most of the benefit comes if you're writing documents with lots of mathematical formulae or tables or something in them. LaTeX is really, really nice for research papers or something, but a lot of office documents place more emphasis on formatting than on "meaning of the formatting". They want a letterhead here, and *this* font, not "put a header on this page and use the title font".
Finally, TeX's coolest features, like setting up automatic page counters and whatever, suffer from a really ugly, archaic language.
I used a combination of Paint Shop Pro and the Gimp.
I like the GIMP -- beautiful piece of software. But that being said, there are definitely things that it doesn't do that would be nice to have. There's a reason the parent poster had to use Paint Shop Pro as well.
The GIMP was modeled heavily after Photoshop -- as a photo retouching tool, though less work was put into for-print output features (like fancy color matching, duotons, color separations) and more into for-computer output (solid image compression and many file formats supported).
It does not, however, fit into the genre of program that existed during the 80s and the early 90s -- the "raster paint program". MacPaint or the Windows Paint program are good examples of this. They tend to emphasize per-pixel operations -- it's easy to, say, draw a single-pixel-thick line or polygon. It's usually not possible to do fine color adjustments. You usually used these things to make synthetic-looking art, diagrams, or game/UI art.
I'm kind of sad to see these things go. If I want to make a rectangle in the GIMP, yes, I can draw a rectangular selection and then fill it. I can simulate the MacPaint polygon tool by using non-smooth Bezier selections and then filling the area...but the old MacPaint style operations made many common computer graphics operations easier.
I wish the GIMP did have something like this -- it'd be easy to add, even if it led to a bit of feature duplication.
The GIMP also isn't much of a natural-media creation program. Anyone that's used MetaCreation's (now procreate's) Painter with a tablet, artist or not, realizes the kind of incredible, easy to make, natural-media output you can get from a computer. I've never painted with real oil paints in my life, but I can do decent-looking oil paintings with Painter and a tablet. The features to support this take some work (at the arcitectural level, unfortunately), but not tons of complex work. I think there are three major things that need to be added to the GIMP to support this. Layers support the image, and an alpha mask. They need a paper map, a dampness map, and a bump map. Layers must have bump maps attached (a la layer maps in the GIMP) which are used to recompute the bump map in real time. This allows paint accumulation (the "streaks" of paint buildup you see in real paintings. Various tools (chisels, scrapers) depress the bump map, and others (paint brushes, crayons) build up the bump map. The dampness map is used to determine bleeding in paper -- if a sponge is passed over paper and then a felt-tip pen drawn over the paper, the ink bleeds much more. Finally, the paper map determines amount of paint/crayon/whatever that comes off of the brush at any point -- it's used to give the paper "texture" and look like canvas, rough paper, wood, whatever.
The GIMP will probably never do this, and I think it's a bit of a shame.
The GIMP is not, despite admirable moves in that direction, a vector graphics program. The Bezier features are very nice, and enterprising artists can do really amazing things with them. Misery has some very good examples of what pseudo-3d work one can do with Bezier curves in a raster graphics program. For the GIMP to be able to compete with, say, Illustrator (which I think *is* possible without architectural changes -- programs like SuperPaint have combined vector and raster elements before), it needs objects, fitting to paths, and better numerical control over elements. You need to be able to say "set a constraint that this object is three inches to the left of this object" and be able to say that object A is two inches big and 300x200 pixels down from the top of the canvas.
It's possible to "stroke" a path -- make a brush draw along a path, but not possible to make text follow a path. This would be a really nice feature, and crucial for a vector graphics program.
Finally, the paths are the closest things to objects the GIMP has -- they are no more than outlines, however, and cannot contain color information, and do not have dependencies. For example, I cannot set a path to contain a gradient (and reshaping the path thus produce a mutated object with the same gradient). The GIMP also has no concept of dependencies -- I can't say "stroke along this path" or "make a drop shadow of the selection contained by this path", update the path, and then have the generated graphics automatically update, though this would be a *tremendous* boon to a graphic artist. The closest I've seen to this is using Illustrator in conjunction with Photoshop with Illustrator's few raster features (letting one, say, create an glass object that distorts objects behind it-- and has an updated distorted area if I reshape the glass object).
Anyway, those are my wish list for the GIMP. I'd like to see an end to statements like "I made this with Paint Shop Pro and the GIMP" or "I made this with Photoshop and the GIMP" and just see "I made this with the GIMP".
I like the fact that the OSS community has banded together enough to make something of the scale of Mozilla
Frankly, I really, really doubt if Mozilla would exist today if Netscape hadn't been such a strong, driving force behind it. They supplied many (if not most) of the engineers to write the code.
The largest "volunteerish" Open Source project I can think of is probably XFree86 or maybe Linux.
Succcess of such a large system requires extreme modularity, so that someone can drop in and write a patch. It's not that easy to just drop into the Moz source.
I get the feel that, while people *could* volunteer to write stuff for Moz, Moz was more of a Netscape project that happened to be very open to the public.
Slashdot Mirror
ACTON, Mass
August 30, 2002
OSDN today announced a second expansion to its recent "black market" entry -- selling links to porn sites. "With our new partners, hotassgirls.com and gianttits.com, we expect to enter a new era of market dominance," said Slashdot General Manager CmdrTaco, "We expect to reach profitability six years earlier than expected." OSDN stock rose to a 3 cent high on the pinksheets following the announcement.
When that firewall blows, who's neck is on the line?...should be enough reason to give them all your employers hard earnt money
Perhaps I'm understanding you wrong, but this seems a little bit unprofessional. You're willing to spend tons of extra company money so that *you* won't be the one to be responsible if something goes wrong?
Are there still people out there willing to take responsibility for their work?
Win32 is sort of an open standard, if incomplete. MS has published documentation on it, and while they may not have defined behavior for all situations, you can follow the documentation.
The problem is that occasionally the documentation differs from Windows, and if it does, everyone only cares whether software works correctly with Windows, not whether it follows the documentation.
Finally, there is no "frozen" version of the documentation that serves as a standard for MS to conform to and release patches to bring their products up to spec with.
I'd call Win32 about as open a standard as Postscript -- we know how to follow the docs, but there's only one implementation that 95% of the people out there care about.
The same is becoming true with HTML and Internet Explorer -- the HTML spec could say one thing, but authors are going to primarily care about compatibility with the leading implementation of that spec, rather than exact compliance with the spec.
People are just taking advantage of the medium, as they did with cartoons.
You have a good point.
If games with a message are infantile, then perhaps enjoyable books with a message suffer from the same complaint -- why aren't you reading a bland, straight political pamphlet if you want a "mature" medium?
Some of the (very little) bit of documentation on OS/400 that I could find on the Internet implied this -- not that one had to keep commands upper case, but that the system did so automatically. I'm used to an OS that distinguishes between case in commands, so this is a bit of a drawback for me.
I could be wrong -- this is some of the only OS/400 documentation that I could *find* freely available. It was a chapter excerpted from a book.
I was kind of interested in learning about OS/400, so I wanted to read up a bit about it. There was very little freely available information, though. This was terribly frusterating to me, because there's plenty of UNIX information available on the Internet.
I hold IBM somewhat responsible for this -- it costs them very little to give out information to the Internet at large, and yet they do not do so. Unless you're a shop prepared to drop quite a chunk of change on propriatary hardware, IBM salespeople aren't going to be very interested in giving you information -- and even if I was, I'd want my evaluation of the system to be based on my *own* interpretation of what I can read about the system, not on what I'm fed by the salespeople ("99.99% uptime! Foo Technology! IBM 24/7 support!"). Oh, yes. IBM does sell (*very expensive*) classes/seminars (oriented towards training to perform specific operations, from what I've seen) to teach people about OS/400. So if you're willing to drop quite a bit of money to learn about an OS that you may not want to use, you can get some information. However, others end up SOL. And I'm in the latter class.
Sorry, but it is impossible to pass judgement on OS/400 if you have no experience with it. It's a COMPLETELY different way of thinking about operating systems, and it's simply not comparable with Unix or Windows.
/400 mindset after working with it for a few months.
/400 is *not* all that accessable to most people.
...we don't have that sort of ignorance here on Slashdot...right? ...right?
:-)
Fair enough.
I come from a Linux/Windows background, and I'm still trying to get my head around the
That statement, though, also illustrates what I'm frusterated with.
Also, there's plenty of good documentation and help for UNIX systems available on the Internet -- this is not true (unless I'm missing something) for OS/400.
It's ignorance that gives us conversations. If we already all knew everything, there'd be no point in talking to each other -- Slashdot wouldn't exist.
Do you think "ps -auxw" is anything less cryptic than the example you gave.
You make a good point -- I do take this for granted. However, this is also one of the more cryptic UNIX commands. Hmm. I think I should rephrase this. There are a lot of UNIX admins running around (compared to OS/400), so for better or for worse, a Solaris guy can get along pretty well on Linux. The commands are well-known by quite a few people.
On OS/400, however, the situation is worse, even assuming an equivalent level of "crypticness" because many people have to learn the OS for the first time.
The obvious attack against me is "Well, there are more MCSEs than UNIX admins, so UNIX also takes a penalty for being cryptic." I *agree* with this -- but UNIX is also being quite accomodating (because it has to be) to that majority of users, making available to them Windows-like environments like GNOME, and GUI administrative tools.
So OS/400 should have to do *more* work to be appealing to the average admin than Windows or even UNIX does...and it doesn't seem to be that it does.
Finally, I stand by my complaint that the programming environment is less than impressive -- the languages are really old. OS/400 programmers *might* just have have a deep love for COBOL or something, but it would really frusterate me.
You are right about the legacy custom apps, but that doesn't excite me about OS/400 -- it just makes me think "People *have* to use OS/400" rather than that "People *want* to use OS/400". It's the same frusteration I see when someone's written something for Win32 or the MacOS that I'd like to use on UNIX (or, occasionally, some other variant of UNIX).
Okay, my posts here are starting to sound like trolls, but I'm going to keep working at this point.
I've heard about a lot of systems billed as being "reliable". Windows folks call Windows 2k "reliable" and say they never have a problem with their servers, that they've been up for months. Linux folks call Linux more reliable than Windows, and say the same thing about their servers. Ditto for the BSD guys, the Sun guys, and evidently for the IBM guys.
So here's what I want to know -- I'll give you that the system is reliable, but that's only meaningful in the presence of unreliable alternatives. Are you claiming that an AS/400 box running OS/400 is more reliable than a BSD box, or a Solaris box, or a Linux box? I've never seen any of these just spontaneously fail or kernel panic...
Heck, I have a swap partition > 2 GB.
I suspect that if AOL was ported to a PDA, people would be quite happy (smaller computer, same functionality), and AOL would avoid the must-run-on-Windows-so-we-must-suck-up-to-MS issue.
So I'm a little biased -- I've never sat down and used OS/400. But I did read through a bit of documentation, and I was thoroughly disgusted with what the OS sounds like. All commands are typed in in caps, in condensed combinations of verbs, then become lower case?
WRKSYSCTX is "work within the system context"?
The directory structure was something like an Apple II, the languages available were from the *Dark Ages*, and the prices obscene.
Why on earth would anyone *ever* want to use OS/400 over *any* variant of UNIX?
I'm impressed that an accountant is successfully poking around with WINE. Very heartening.
What on earth are you talking about? WINE isn't illegal in the least! What law do you think they're breaking?
Trademark? It'd be tough to show that WINE infringes on "Microsoft Windows".
Patent? I really doubt it. I can't imagine that there is any patent that could keep you from implementing an API.
Copyright? This was already tried by MS, arguing that they owned the header files, and duplicating the information in them was infringing. It didn't work.
EULA for reverse engineering? The WINE guys at least put up a pretense of clean-room engineering, and I think it'd be hard for MS to prove otherwise. You really don't need to disassemble Windows to implement the Win32 API as documented by MSDN. Finally, the right to reverse engineer is expressly granted in many locales for the purpose of ensuring compatibility. In the EU, this would make even a non-clean room impementation okay. In the US, it's a little more dicy, as I believe the laws only apply to compatibility over a network protocol between two different hosts, but there's still a general trend towards allowing people to produce compatible products.
Look-and-feel? Look-and-feel was an approach specifically shown by MS to not be a valid case for suit in the US legal system when they butted heads with Apple.
MS would have a *hell* of a time trying to prove that no one could implement a compatible product, which they'd have to do to nail the WINE guys. The antitrust guys would have a field day on MS.
Finally, WINE is not an emulator. "Emulator" has a specific meaning in the computer world -- it would reimplement the hardware that the software runs on. This would almost certainly cause a performance hit. WINE carries no such required overhead. At least in theory, WINE can run just as fast as (heck, faster) than Windows.
Given your AC nature, I'd almost say that you're trolling, but I can't quite be sure.
And screw obscurity, it was a good game. There are a lot of "obscure" games that are really, *really* good -- Total Annihilation is *still* better (much more powerful order-giving abilities) than the current crop of RTSes. Fallout is a top-notch RPG. The Last Express was beautifully done. There are a bunch of really good Q3 mods that were overshadowed by Counterstrike, likek Team Fortress and Weapons Factory.
People should realize that a good chunk of games prosper based very much on their marketing. Blizzard, for instance, throws a *huge* amount of money into marketing their games -- it's very difficult for the competition (which I've often found to be superior) to compete. People buy games based on screenshots in magazines, based on the overwhelming number of ads they see, and on how compelling the box is. Few people play a game for a while before buying it.
I certainly see your point, but I don't think that fun and other people making money are entirely incompatible.
:-)
I send in patches to Linux software that are packaged and used to make lots of money for RH, Mandrake, and others. The attitude I have is "As long as other techies can grab this for free, I certainly don't have any problem with RH making a buck off of making it accessable to other users." It's kind of a trade, too -- I make patches, improve their product (and have fun in the process), and they give me an ever-improving free product for download.
I do want to say thank you, a big thank you, to isolation for his mingw and WINE work. I use both frequently (mingw in two commercial settings), and I deeply appreciate them. They've let me and others escape the MS monopoly and still get work done, and they are both technically impressive pieces of software.
I do sort of wish that WINE could have stayed BSD-licensed -- there were a lot of people pretty comfortable with it, and developer groups and companies were relying on that license. I, in general, prefer the GPL (well, the LGPL) to the BSD license, but in this case, I'm kind of sad about the switch. It fragmented the WINE developer community and started a lot of fights.
OTOH, I also have to take my cap off to Transgaming -- the gentlemen there are taking enormous personal financial risk (throwing their own money that they can't really afford) into trying to make a commercially viable company that gives away source to a product that lets you play Windows games. I know one Linux user who has only a single Windows game that he wants to play -- Max Payne -- and Transgaming has let him do that.
Anyway, every line of code put out there, by anyone, is helping an awful lot of people. It's also pulling off some darn impressive technical tricks -- WINE is one of the few things that really blows my mind -- I'm amazed that the developers pulled it off. Here's to more coding -- and less politics.
The reason some things do not work is that you have to implement more than just the documented externals.
This is partly true, but not the whole story. Yes, there have been times when either MS bugs have to be reproduced and emulated, or undocumented behavior discovered. However, this really isn't the cause of most incompatibilities.
First, the WINE team is limited -- hardly as many man-hours per weak as the Windows team at MS. So they take a, as they put it, "product-driven" approach. They take a specific program, and implement just enough to get it working properly. Few programs use (or will use for several years) WinXP-specific features, because it would limit their potential market. Same goes for Win2k-specific features. So the WINE guys don't bother with emulating those. Also, less crucial and rarely used fuctions are often just stubs, meaning that software that uses lots of esoteric options/functions is much more likely not to work.
Last of all, rarely used chunks of Win32 are simply ignored. I believe that there is basically no CryptoAPI support, for instance, because implementing CryptoAPI would be a significant amount of work, and very, very few programs would actually use it.
A lead Nvidia hardware guy came and gave a talk at our university. I asked him whether Nvidia preferred using DirectX over OpenGL because of more supported features in DX8 vs OGL 1.x, and he surprised me by saying that, no, in fact they significantly preferred OGL.
Apparently, when Nvidia adds a feature, OGL has a standard way of adding an extension to the language to support extra features. MS, despite serious lobbying from Nvidia, strongly pushes against supporting extensions (for obvious reasons, but it still doesn't go over well with Nvidia). So there may be a number of performance-enhancing features or tricks that are used with OGL (because the game developer did a bit of extra work to support the extensions) that are not used with DirectX, or have to be done partly in software with DirectX.
Anyway, the gist of this is that if you want to use all the features of your Nvidia video card, you're likely better off using OGL modes in your games.
I'm not sure what the take on this is at ATI or Matrox, though.
I disagree with the overhead statement. Some software seems to run more snappily, others less.
For example, Windows menuy widgets seem to operate much more slowly, but I've played Starcraft on the same machine in both WINE and Windows NT, and if Starcraft wasn't faster in WINE, it was at least as fast (admittedly, WinNT's DirectX probably wasn't as tweaked as newer releases, but even so...).
I just wanna see Close Combat work fully...sigh.
They use Lua, but it's not like they're handing out a collection of Lua source...also, it wouldn't surprise me if they've modified their Lua.
This is exactly what I was saying. When there were other PPC desktop manufacturers, there was a market with multiple vendors.
Let me guess, you're one of those people who freaks out hardcore when you find telnet in use, right.
:-)
Not if it's Kerberized and encrypted or ssl-tunneled.
This is a fairly reasonable thing to "freak out" about -- there *have* been major break ins because people are using unencrypted telnet and someone seizes a crucial machine.
So why don't you have WordPerfect available?
to be honest, most of the benefit comes if you're writing documents with lots of mathematical formulae or tables or something in them. LaTeX is really, really nice for research papers or something, but a lot of office documents place more emphasis on formatting than on "meaning of the formatting". They want a letterhead here, and *this* font, not "put a header on this page and use the title font".
Finally, TeX's coolest features, like setting up automatic page counters and whatever, suffer from a really ugly, archaic language.
I used a combination of Paint Shop Pro and the Gimp.
I like the GIMP -- beautiful piece of software. But that being said, there are definitely things that it doesn't do that would be nice to have. There's a reason the parent poster had to use Paint Shop Pro as well.
The GIMP was modeled heavily after Photoshop -- as a photo retouching tool, though less work was put into for-print output features (like fancy color matching, duotons, color separations) and more into for-computer output (solid image compression and many file formats supported).
It does not, however, fit into the genre of program that existed during the 80s and the early 90s -- the "raster paint program". MacPaint or the Windows Paint program are good examples of this. They tend to emphasize per-pixel operations -- it's easy to, say, draw a single-pixel-thick line or polygon. It's usually not possible to do fine color adjustments. You usually used these things to make synthetic-looking art, diagrams, or game/UI art.
I'm kind of sad to see these things go. If I want to make a rectangle in the GIMP, yes, I can draw a rectangular selection and then fill it. I can simulate the MacPaint polygon tool by using non-smooth Bezier selections and then filling the area...but the old MacPaint style operations made many common computer graphics operations easier.
I wish the GIMP did have something like this -- it'd be easy to add, even if it led to a bit of feature duplication.
The GIMP also isn't much of a natural-media creation program. Anyone that's used MetaCreation's (now procreate's) Painter with a tablet, artist or not, realizes the kind of incredible, easy to make, natural-media output you can get from a computer. I've never painted with real oil paints in my life, but I can do decent-looking oil paintings with Painter and a tablet. The features to support this take some work (at the arcitectural level, unfortunately), but not tons of complex work. I think there are three major things that need to be added to the GIMP to support this. Layers support the image, and an alpha mask. They need a paper map, a dampness map, and a bump map. Layers must have bump maps attached (a la layer maps in the GIMP) which are used to recompute the bump map in real time. This allows paint accumulation (the "streaks" of paint buildup you see in real paintings. Various tools (chisels, scrapers) depress the bump map, and others (paint brushes, crayons) build up the bump map. The dampness map is used to determine bleeding in paper -- if a sponge is passed over paper and then a felt-tip pen drawn over the paper, the ink bleeds much more. Finally, the paper map determines amount of paint/crayon/whatever that comes off of the brush at any point -- it's used to give the paper "texture" and look like canvas, rough paper, wood, whatever.
The GIMP will probably never do this, and I think it's a bit of a shame.
The GIMP is not, despite admirable moves in that direction, a vector graphics program. The Bezier features are very nice, and enterprising artists can do really amazing things with them. Misery has some very good examples of what pseudo-3d work one can do with Bezier curves in a raster graphics program. For the GIMP to be able to compete with, say, Illustrator (which I think *is* possible without architectural changes -- programs like SuperPaint have combined vector and raster elements before), it needs objects, fitting to paths, and better numerical control over elements. You need to be able to say "set a constraint that this object is three inches to the left of this object" and be able to say that object A is two inches big and 300x200 pixels down from the top of the canvas.
It's possible to "stroke" a path -- make a brush draw along a path, but not possible to make text follow a path. This would be a really nice feature, and crucial for a vector graphics program.
Finally, the paths are the closest things to objects the GIMP has -- they are no more than outlines, however, and cannot contain color information, and do not have dependencies. For example, I cannot set a path to contain a gradient (and reshaping the path thus produce a mutated object with the same gradient). The GIMP also has no concept of dependencies -- I can't say "stroke along this path" or "make a drop shadow of the selection contained by this path", update the path, and then have the generated graphics automatically update, though this would be a *tremendous* boon to a graphic artist. The closest I've seen to this is using Illustrator in conjunction with Photoshop with Illustrator's few raster features (letting one, say, create an glass object that distorts objects behind it-- and has an updated distorted area if I reshape the glass object).
Anyway, those are my wish list for the GIMP. I'd like to see an end to statements like "I made this with Paint Shop Pro and the GIMP" or "I made this with Photoshop and the GIMP" and just see "I made this with the GIMP".
I like the fact that the OSS community has banded together enough to make something of the scale of Mozilla
Frankly, I really, really doubt if Mozilla would exist today if Netscape hadn't been such a strong, driving force behind it. They supplied many (if not most) of the engineers to write the code.
The largest "volunteerish" Open Source project I can think of is probably XFree86 or maybe Linux.
Succcess of such a large system requires extreme modularity, so that someone can drop in and write a patch. It's not that easy to just drop into the Moz source.
I get the feel that, while people *could* volunteer to write stuff for Moz, Moz was more of a Netscape project that happened to be very open to the public.