You are correct, at least partially. I agree that a benevolent dictator helps a project (see Perl version 5 up until 5.002 and Linux as examples). However, it is not at all required. See Perl after 5.002; The X Window System (committee with many forks); GCC (formerly B.D. type, now committee with many forks); Apache; Gnome; Mozilla; and many, many other projects for details on ther ways it can work and work well.
Ok, so now let's take that back to the Sun/Solaris example and see how it applies. Sun is taking the role of dictator, but a) they cannot be deposed if the users dislike the direction the software is taking b) "small forks" (like ilinux; egcs; phoenix; etc) cannot be made, maintained and then merged back in as required c) priorities set by Sun on compatability with external software and/or standards cannot be re-set by users of the software without wholly maintaining thier own fork.
I'm not saying that what Sun is doing is bad. I'm not saying that Sun will fail to re-capture some of the market lost to Linux and BSD in the short term. I'm actually not taking a stand on either one, and just waiting to see what happens.
What I am saying is that Sun cannot walk halfway down the road to open source software development and reap half the benefits. In fact, I would agrue that they won't get within an order of magnitude of the boost to creative input, collaberation and niche-filling power of an open source project.
Should we pat Sun on the back for this move? No. Sun is responding to market pressures in the same way that Microsoft is. If they were responding in the way that Netscape did, I would be patting them on the back because they would have joined the community and put their eggs into the global basket that is open source collaberation.
That's the interesting part, really. As more an more companies ponder what becoming open source would mean, they all come to the point of thinking, "would I have to give up my valuable source code, and let others benefit from my years of work?"
I understand that sort of provincial attitude, but it's warped. It doesn't look at the flip-side. Why would Sun not want to benefit from Linux's or BSD's work on user-mode virtualization? Why would Sun not want Solaris x86 to gain access to the vast array of hardware that Linux supports? Why would Solaris customers want to live without a well-supported version of gcc with the improvements that Sun could ultimately make to it for Sparc?
It's hard to look at the move from proprietary to open source software positively because it breaks the business model of proprietary software. It makes a vendor think in terms of their value-added instead of the value behind secret software. Sun would have to let go of the safety-rail and become a Solaris services provider, and that's a scary thing.
The problem is that ultimately (I'm talking about years down the line) their choice is to do just that or watch either MS or Linux or both put them out of business. I do not envy them this choice, because even if they make what I think is the right call, I think they're probably too late.
There's not much that they're going to be able to do circa the release of the new SpamAssassin, which can personalize itself to a user's mail by training a bayesian filter based on the results of all of the other SA tests. I'm using the CVS version now, and have the following 3 behaviors implemented in prcomail based on the SA results:
1. If Razor2 and the Bayesian filter say it's spam, dump it. Razor2 is much more reliable than R1, and combined with the local "looks like spam", I'm willing to trust the result. 2. If SA scores it above a 20, dump it (amazingly the new version does this on a reasonable percentage of spam, and I've never seen a false positive or even anything that came close). 3. If SA scores it above 4, mark the subject and deliver to me.
This last rule allows me to quickly scan for friends, family and co-workers email addresses and then delete the lot of it (via a virtual mailbox in evolution). Fast, easy, painless.
I also auto-train the Bayesian filter and manually correct it on any false-negatives. It's gotten to the point that I haven't had to do that in a few weeks now.
Same can be said for many such books. Not really the issue. It was an American comic. Same goes for Sandman which was written by Neil Gaiman. I think he's from one of those silly little countries on the other side of the pond too;-)
First off, you've covered a lot of stuff that I like, but I was trying to stick with the stuff I actually read now. Watchment, V for Vendetta and Preacher are all over.
Preacher, I had mixed feelings about. It was a well-written story, but ultimately it was about Texas and more generally, the Southern U.S.
The gore was a bit of a turn-off, but I took it as written that it was part of the genre. The attitudes of the main characters were occasionally confusing to me, and dammit, I'm old fashioned. I want a story to redeem at least one of the flaws in at least one of the main characters. If anything, the point of this story was to keep pointing out more and more flaws until the story ended.
Like I said, it was well written. Just not sure that I liked it.
I'm going to go out on a limb here: You cannot approach being open source.
This can be compared to the statement that you cannot approach being pregnant. You simply are or you are not. To quote Yoda, "There is no try, only do."
How's that? Well, quite simply, unless your source code is available for modification and you can run the software wherever you need it, and you can contribute your changes to whatever maintainer you please (note that I just excluded, for example, qmail), then you cannot build an open source community around the software. You can certainly have a strong and dedicated community (Windows even has one), but you cannot build a community that has the same benefits. You cannot have the people who need to scratch the itch scratch it at will for themselves and anyone else with the same itch.
Ultimately Sun will decide who can collaborate on their own versions. They will maintain a centralized set of priorities, and contributors outside of Sun will be viewed as submitters of bug-reports that compile, not co-workers.
This is not open source. This is a company, faced with extinction via open source, trying to hide in the tall grass. Hint: it's not going to work. I say this being a big fan of what Sun did for the industry in their day. It doesn't matter. The more they say, "look, we have the benefits of open source too," the more people will begin to ask, "why not just go with truly open source software?"
Linux, BSD (amusingly, Sun's old source code base), and many other smaller-niche free systems are rapidly eclipsing the proprietary operating systems. You look at MS and see very little movement, but that's because they're so large and move in different circles for now. When you look at Sun or HP you begin to see the devastation that these upstarts are creating in the industry. Why? Because collaboration with your peers is powerful. Collaboration between customer and vendor is almost always not.
The reason that comics in the US do not have the wide appeal of manga in Japan is obvious in just listening to a comic fan attempt to describe manga. It usally goes something like, "they're like comics, but they're about anything. Everyone reads them."
There's a perception in the US that comics (the name itself is skewed, which is why so many have tried to use terms like "graphic literature", etc.) can only contain "kids" stories and their only fan-base are otaku.
This is not true, of course, but it's correct for a majority of comics (getting less so over the past 10 years) and a majority of readers (still quite true). Until both of those change, the stereotypes will remain. There *should* be more comics about everyday life that appeal to everyone.
If you've been out of the loop and are interested in catching some of the more interesting comics out there, here are some of the ones that I've found interesting (note: not all of these are traditional comics, but some are and the rest are certainly not mainstream literature by any measure):
Superheros:
* Astro City -- What its like to live with supers * TOP10 -- In a world where everyone from the meter-maids to the homeless are super-heros, what are the police like? Odd premise, great book. * Rising Stars -- In a midwest town, a meteor strikes causing all in-utero babies in the area (113 of them) to develop unusual abilities when they're born. The 24-issue series follows their lives and deaths and the politics that surrounds them.
Fantasy / Alternate History:
* League of Extraordinary Gentlemen -- An amazingly cool look at an alternate history where all of the late 1800s and early 1900s fantasy, adventure and science fiction books are true. Everyone from Sherlock Holmes to Captn. Nemo to Dr Jeckle are in the story, and it works well. * Girl Genius -- A fun story about a world that can only do high-tech through magical individuals known as sparks. * Lucifer -- The title character is the angel, cast out of heaven and formerly ruler of hell. This is a spin-off of the classic late-80s/early-90s series Sandman.
These are the books that I read now. Fantasy and super-heros are well established genres for comics, and they're done well in many cases. It's just too bad that there aren't more genres being allowed in. Real science fiction makes an attempt every now and then, and sometimes it works, but often it does not. The slice-of-life stories that really made manga are almost non-existant. In fact, the closest US comics came to that, AFAICT is Archie.
By proven, I meant that it may never be given such a weight of evidence that it's incorporated into the body of theories that we currently maintain to be, "the way it happened".
Also, it is never possible to prove that an event that I just observed actually occured, and as you get further and further from that level of certainty things get fuzzier and fuzzier. However, there are many potential ways to prove the big theories to a degree of certainty that is as acceptable as claiming that any given even just occured.
They just require higher tech than we have. Obviously time-travel solves this problem (even if it's only capable of observation, which is the most plausible type of time-travel). Other solutions vary per theory. This theory is an easy one. If we conclusively determine what DNA corrosponds to the reflex, and can get DNA samples from the fossils that we think it came from, then we can see if that DNA sequence exists. If it does, and does not exist in that creature's ancestors, you've proven your theory. Woefully, that assumes all sorts of difficult things.
I get that. The point is, how many people will want to buy a system that already has a floppy drive for the reasons that I stated. If it's too large a number, Dell could be in trouble.
The article relates a new theory, nothing more. It's a promising theory, and one which can be disproven easily. If the test fails to disproove the theory, then it can be taken more seriously as an explanation. Still, it may never be PROVEN, per se.
Actually, I think Ximian has a preview release from Red Carpet (not sure) of 2.0, but from their Web site: "the new Ximian Desktop 2.0 [is] scheduled for this spring"
Working well so far. The basic libs and such compiled cleanly. My laptop battery died last night, and as you can imagine, the download times are awful given the Slashdotting that the servers are taking.
You're right to avoid 8.0. When 8.1 comes out, *if* it's based on Gnome 2.2, I may swap back, but by then Ximain will probably have come out with their 2.2 distribution, and I'll just install that over whatever RH ships. RH does an ok job with Gnome, but Ximian really puts the pollish on it.
Like the idea of 2.2, and want to know if it works?
Don't want to have to compile a practically infinite number of packages from scratch by downloading each... and... every... one?
Don't want to trash your system by overwriting tons of libraries?
Don't trust your local package management system (or the packagers for that matter) to back out once you upgrade?
Check out Garnome. It's easy to download and install, and best of all, it installs under your home directory by default! They have a good walk-through on the main page. I've found this to be very nice, and a far smoother testing mechansim than anything else out there!
You just download a small BSD-style ports tree, and then kick off a compile (follow the README). It downloads, compiles and installs everything else!
One of the first things to do when planning a change to a product is to go out and find out what your customers requirements are.
Dell will find, should they do this, that customers use floppies to fit three requirements:
1: Small data distributions from vendors (e.g. drivers, license keys, etc)
2: Legacy media access (e.g. old documents from school)
3: Storage for small pieces of data which must be prortable (e.g. resume, encryption keys, browser passwords, etc).
The first item is easily taken care of. Vendors have been moving toward CD for this for years, and this announcement will only spur them on.
The third is where USB-keychains really shine over floppies.
The second is something that Dell is either going to not care about or wish they'd done a whole lot more thinking on. The question is: how many potential Dell customers have old floppies that they think they might need to access. How many of those customers will buy a non-Dell system because they fear that they will need access to that old media?
As I've pointed out elsewhere, that only happens if you're using astoundingly bad filtering software.
SpamAssassin's scores for each of its many dozens of tests are assigned by a genetic algorithm that uses a database of many thousands of examples of spam and good mail to tune the scoring. Such features of commercial mail accounts are one of the very large set of features that are automatically accounted for without anyone having to lift a finger.
Not at all, and you're thinking about far too simple a model.
With SpamAssassin, I deal with spam in 3 ways:
1. Mail that gets a score of 20 or more is sent to/dev/null. Mail would have to be carefully crafted to achieve a 20 unless it truly is spam. Such effort is not to be rewarded:-) Keep in mind that mail about laws on sexuality or other simple examples cited in this article would never get NEAR a 20.
2. Mail that triggers both the Bayesian and Razor2 tests is sent to/dev/null. This is a very nice way to identify that a) there's a consensus that this very message is spam and b) my local mail patterns indicate that this is spam.
3. Anything else with a score of 4 or more is marked in the subject line and I have a virtual mailbox in my mail client that I use to glance at the from addresses. If something looks plausible, I check it out.
As of the development version of SpamAssassin that I'm using (about a week old out of CVS), I get a false positive rate of about 1:100-200 messages and during testing over the last couple of months, I copied the messages that would have gone to/dev/null to a mailbox that I scanned carefully. None of the messages that I would have thrown out were non-spam.
I get a LOT of mail form lists, spammers, friends, random people on the net, machines spewing status, etc. I feel that I'm a reasonably good QAer for this sort of thing, and the new SpamAssassin will rock your world (and the spammers')!
I'm not sure if I agree with the idea or not, but here's what I think the reasoning is:
File-sharing in general is at the center of a huge right-to-share/fair use vs. copyright/profit debate right now. Thus, any/. story concerning file sharing is likely to show up in this section regardless of its specific relation to rights or lack thereof.
I must be missing something fundamental to XRender. It's mostly focused on anti-aliased fonts and alpha blending both of which it will be used for by Gnome when XRender adoption becomes a bit more universal (Sun (Solaris 9) and XFree86 (version 4) are, AFAIK, the only X displays to offer this extension currently). XRender's big advantage, in fact, is that it doesn't require you to interact with X in a radically different way.
Suggesting that not using such a sparsely available and relatively new technology to do something that it's not intended for is "misprogram[ming] X" seems to be way out of line to me.
OpenGL is not the right way to go, as it's designed for an entirely different purpose.
The correct solution is to re-think the way X works (specifically with regards to XImage data, Colormaps, Fonts and drawing primatives), and design it to use modern graphics cards correctly. That's not going to happen tomorrow, but it is happening. KDE/Qt and Gnome/GTK+ will take advantage of such features at the become available.
I disagree, though you may have been attempting humor.
vi and Emacs are products of the era in which they were built, and both were considered excellent UIs for the day. Given today's advances in UI design, I find it interesting that those two are still the best ways to edit text. There *should* be better, but there simply isn't. There are some basic reasons for this that are interesting and fly in the face of a lot of UI design:
[note: I'm talking about Emacs as a text-editor, which is a fair slice of its functionality to discuss, though certainly not all of it]
Both editors were designed in an era when the mouse wasn't an option. This means that even given recent modifications to add mouse support in vim and Lucid Emacs (GNU Emacs had primative mouse support first, but Lucid modernized it, and much of their work was merged back into GNU Emacs, though not all, hence Lucid's transition to the even-today-maintained XEmacs) in the last 10 or so years, they still remain useful when your fingers never leave the keyboard.
The other thing that both of these do that puts them head-and-shoulders above many related tools is interaction with external programs. In vi or Emacs, I can send a portion of a document through any external program I want and read back the result into the original document. Here's an example in vi:
!Gsort -n +2
This sorts the lines from the current position to the bottom of the file based on the numeric value of the third column. To add this feature into the editor would be fine (Emacs can do this actually), but if an external tool already knows how to do it, why not take advantage of that fact?
I have to disagree. First, a good HIG does not constrain you to do the wrong thing for your application, but rather gives you a sense of how best to fit into a particular paradigm. From there, you can do what you need to. A classic example of where this is important is the current trend in "media players" like Quicktime from Apple, WinAmp, XMMS, Microsoft Media Player, etc. They all try to look as snazzy as possible, at the cost of good user interface design. I can't count the number of times I've been at a party where some geek with good intentions put up an XMMS or WinAmp-based juke-box for people to play music with and all of the non-techies would give up after a few minutes because they couldn't decipher the UI.
If those applications had followed the UI guidelines for the platforms they run on, they could still have had all of the features, all of the great flexibility, but they didn't need to have pseudo-round volume sliders, non-standard title-bars that do application-local window management, context-sensitive menus that don't have commonly-performed operations, out-of-the-box unique font selection, out-of-the-box unique color selection, etc, etc.
That kind of awful behavior is what makes a desktop unusable (and certainly if more apps go the "branded UI" route, dekstops will become totally unusable).
But you will certainly get 100 different ways to do something and 100 different widget sets.
You may also get common user interface standards, but that has little or nor bearing on the above.
Also, pointing to Apple as an example good HIG is probably not the moral high-ground it once was, given recent software from that company (e.g. the latest Quicktime) pointedly ignores such things....
Incorrect. They are combining the two documents into one, with sections that conflict annotated in XML such that either document can be viewed as a stand-alone. Then they are going to work on the differences.
Read the announcement on the mailing list archives.
Slashdot Mirror
You are correct, at least partially. I agree that a benevolent dictator helps a project (see Perl version 5 up until 5.002 and Linux as examples). However, it is not at all required. See Perl after 5.002; The X Window System (committee with many forks); GCC (formerly B.D. type, now committee with many forks); Apache; Gnome; Mozilla; and many, many other projects for details on ther ways it can work and work well.
Ok, so now let's take that back to the Sun/Solaris example and see how it applies. Sun is taking the role of dictator, but a) they cannot be deposed if the users dislike the direction the software is taking b) "small forks" (like ilinux; egcs; phoenix; etc) cannot be made, maintained and then merged back in as required c) priorities set by Sun on compatability with external software and/or standards cannot be re-set by users of the software without wholly maintaining thier own fork.
I'm not saying that what Sun is doing is bad. I'm not saying that Sun will fail to re-capture some of the market lost to Linux and BSD in the short term. I'm actually not taking a stand on either one, and just waiting to see what happens.
What I am saying is that Sun cannot walk halfway down the road to open source software development and reap half the benefits. In fact, I would agrue that they won't get within an order of magnitude of the boost to creative input, collaberation and niche-filling power of an open source project.
Should we pat Sun on the back for this move? No. Sun is responding to market pressures in the same way that Microsoft is. If they were responding in the way that Netscape did, I would be patting them on the back because they would have joined the community and put their eggs into the global basket that is open source collaberation.
That's the interesting part, really. As more an more companies ponder what becoming open source would mean, they all come to the point of thinking, "would I have to give up my valuable source code, and let others benefit from my years of work?"
I understand that sort of provincial attitude, but it's warped. It doesn't look at the flip-side. Why would Sun not want to benefit from Linux's or BSD's work on user-mode virtualization? Why would Sun not want Solaris x86 to gain access to the vast array of hardware that Linux supports? Why would Solaris customers want to live without a well-supported version of gcc with the improvements that Sun could ultimately make to it for Sparc?
It's hard to look at the move from proprietary to open source software positively because it breaks the business model of proprietary software. It makes a vendor think in terms of their value-added instead of the value behind secret software. Sun would have to let go of the safety-rail and become a Solaris services provider, and that's a scary thing.
The problem is that ultimately (I'm talking about years down the line) their choice is to do just that or watch either MS or Linux or both put them out of business. I do not envy them this choice, because even if they make what I think is the right call, I think they're probably too late.
There's not much that they're going to be able to do circa the release of the new SpamAssassin, which can personalize itself to a user's mail by training a bayesian filter based on the results of all of the other SA tests. I'm using the CVS version now, and have the following 3 behaviors implemented in prcomail based on the SA results:
1. If Razor2 and the Bayesian filter say it's spam, dump it. Razor2 is much more reliable than R1, and combined with the local "looks like spam", I'm willing to trust the result.
2. If SA scores it above a 20, dump it (amazingly the new version does this on a reasonable percentage of spam, and I've never seen a false positive or even anything that came close).
3. If SA scores it above 4, mark the subject and deliver to me.
This last rule allows me to quickly scan for friends, family and co-workers email addresses and then delete the lot of it (via a virtual mailbox in evolution). Fast, easy, painless.
I also auto-train the Bayesian filter and manually correct it on any false-negatives. It's gotten to the point that I haven't had to do that in a few weeks now.
Same can be said for many such books. Not really the issue. It was an American comic. Same goes for Sandman which was written by Neil Gaiman. I think he's from one of those silly little countries on the other side of the pond too ;-)
First off, you've covered a lot of stuff that I like, but I was trying to stick with the stuff I actually read now. Watchment, V for Vendetta and Preacher are all over.
Preacher, I had mixed feelings about. It was a well-written story, but ultimately it was about Texas and more generally, the Southern U.S.
The gore was a bit of a turn-off, but I took it as written that it was part of the genre. The attitudes of the main characters were occasionally confusing to me, and dammit, I'm old fashioned. I want a story to redeem at least one of the flaws in at least one of the main characters. If anything, the point of this story was to keep pointing out more and more flaws until the story ended.
Like I said, it was well written. Just not sure that I liked it.
I'm going to go out on a limb here: You cannot approach being open source.
This can be compared to the statement that you cannot approach being pregnant. You simply are or you are not. To quote Yoda, "There is no try, only do."
How's that? Well, quite simply, unless your source code is available for modification and you can run the software wherever you need it, and you can contribute your changes to whatever maintainer you please (note that I just excluded, for example, qmail), then you cannot build an open source community around the software. You can certainly have a strong and dedicated community (Windows even has one), but you cannot build a community that has the same benefits. You cannot have the people who need to scratch the itch scratch it at will for themselves and anyone else with the same itch.
Ultimately Sun will decide who can collaborate on their own versions. They will maintain a centralized set of priorities, and contributors outside of Sun will be viewed as submitters of bug-reports that compile, not co-workers.
This is not open source. This is a company, faced with extinction via open source, trying to hide in the tall grass. Hint: it's not going to work. I say this being a big fan of what Sun did for the industry in their day. It doesn't matter. The more they say, "look, we have the benefits of open source too," the more people will begin to ask, "why not just go with truly open source software?"
Linux, BSD (amusingly, Sun's old source code base), and many other smaller-niche free systems are rapidly eclipsing the proprietary operating systems. You look at MS and see very little movement, but that's because they're so large and move in different circles for now. When you look at Sun or HP you begin to see the devastation that these upstarts are creating in the industry. Why? Because collaboration with your peers is powerful. Collaboration between customer and vendor is almost always not.
The reason that comics in the US do not have the wide appeal of manga in Japan is obvious in just listening to a comic fan attempt to describe manga. It usally goes something like, "they're like comics, but they're about anything. Everyone reads them."
There's a perception in the US that comics (the name itself is skewed, which is why so many have tried to use terms like "graphic literature", etc.) can only contain "kids" stories and their only fan-base are otaku.
This is not true, of course, but it's correct for a majority of comics (getting less so over the past 10 years) and a majority of readers (still quite true). Until both of those change, the stereotypes will remain. There *should* be more comics about everyday life that appeal to everyone.
If you've been out of the loop and are interested in catching some of the more interesting comics out there, here are some of the ones that I've found interesting (note: not all of these are traditional comics, but some are and the rest are certainly not mainstream literature by any measure):
Superheros:
* Astro City -- What its like to live with supers
* TOP10 -- In a world where everyone from the meter-maids to the homeless are super-heros, what are the police like? Odd premise, great book.
* Rising Stars -- In a midwest town, a meteor strikes causing all in-utero babies in the area (113 of them) to develop unusual abilities when they're born. The 24-issue series follows their lives and deaths and the politics that surrounds them.
Fantasy / Alternate History:
* League of Extraordinary Gentlemen -- An amazingly cool look at an alternate history where all of the late 1800s and early 1900s fantasy, adventure and science fiction books are true. Everyone from Sherlock Holmes to Captn. Nemo to Dr Jeckle are in the story, and it works well.
* Girl Genius -- A fun story about a world that can only do high-tech through magical individuals known as sparks.
* Lucifer -- The title character is the angel, cast out of heaven and formerly ruler of hell. This is a spin-off of the classic late-80s/early-90s series Sandman.
These are the books that I read now. Fantasy and super-heros are well established genres for comics, and they're done well in many cases. It's just too bad that there aren't more genres being allowed in. Real science fiction makes an attempt every now and then, and sometimes it works, but often it does not. The slice-of-life stories that really made manga are almost non-existant. In fact, the closest US comics came to that, AFAICT is Archie.
By proven, I meant that it may never be given such a weight of evidence that it's incorporated into the body of theories that we currently maintain to be, "the way it happened".
Also, it is never possible to prove that an event that I just observed actually occured, and as you get further and further from that level of certainty things get fuzzier and fuzzier. However, there are many potential ways to prove the big theories to a degree of certainty that is as acceptable as claiming that any given even just occured.
They just require higher tech than we have. Obviously time-travel solves this problem (even if it's only capable of observation, which is the most plausible type of time-travel). Other solutions vary per theory. This theory is an easy one. If we conclusively determine what DNA corrosponds to the reflex, and can get DNA samples from the fossils that we think it came from, then we can see if that DNA sequence exists. If it does, and does not exist in that creature's ancestors, you've proven your theory. Woefully, that assumes all sorts of difficult things.
I get that. The point is, how many people will want to buy a system that already has a floppy drive for the reasons that I stated. If it's too large a number, Dell could be in trouble.
Brings new meaning to, "Dude, you got Delled!"
The article relates a new theory, nothing more. It's a promising theory, and one which can be disproven easily. If the test fails to disproove the theory, then it can be taken more seriously as an explanation. Still, it may never be PROVEN, per se.
Actually, I think Ximian has a preview release from Red Carpet (not sure) of 2.0, but from their Web site: "the new Ximian Desktop 2.0 [is] scheduled for this spring"
Working well so far. The basic libs and such compiled cleanly. My laptop battery died last night, and as you can imagine, the download times are awful given the Slashdotting that the servers are taking.
You're right to avoid 8.0. When 8.1 comes out, *if* it's based on Gnome 2.2, I may swap back, but by then Ximain will probably have come out with their 2.2 distribution, and I'll just install that over whatever RH ships. RH does an ok job with Gnome, but Ximian really puts the pollish on it.
I've used 0.20 with no problems under RH8.0.
I'm now compiling 0.21 under 7.3... we shall see...
Like the idea of 2.2, and want to know if it works?
Don't want to have to compile a practically infinite number of packages from scratch by downloading each... and... every... one?
Don't want to trash your system by overwriting tons of libraries?
Don't trust your local package management system (or the packagers for that matter) to back out once you upgrade?
Check out Garnome. It's easy to download and install, and best of all, it installs under your home directory by default! They have a good walk-through on the main page. I've found this to be very nice, and a far smoother testing mechansim than anything else out there!
You just download a small BSD-style ports tree, and then kick off a compile (follow the README). It downloads, compiles and installs everything else!
One of the first things to do when planning a change to a product is to go out and find out what your customers requirements are.
Dell will find, should they do this, that customers use floppies to fit three requirements:
1: Small data distributions from vendors (e.g. drivers, license keys, etc)
2: Legacy media access (e.g. old documents from school)
3: Storage for small pieces of data which must be prortable (e.g. resume, encryption keys, browser passwords, etc).
The first item is easily taken care of. Vendors have been moving toward CD for this for years, and this announcement will only spur them on.
The third is where USB-keychains really shine over floppies.
The second is something that Dell is either going to not care about or wish they'd done a whole lot more thinking on. The question is: how many potential Dell customers have old floppies that they think they might need to access. How many of those customers will buy a non-Dell system because they fear that they will need access to that old media?
We shall see...
"Rick Berman's First Post-NEMESIS Interview"
... "Post-Nemesis"? ... ... nevermind.
Ha! I knew I recognized the writing style on few of those FPs! Well, Mr. Berman, I just want you to know I modded you down. So there.
What's that? "First"
As I've pointed out elsewhere, that only happens if you're using astoundingly bad filtering software.
SpamAssassin's scores for each of its many dozens of tests are assigned by a genetic algorithm that uses a database of many thousands of examples of spam and good mail to tune the scoring. Such features of commercial mail accounts are one of the very large set of features that are automatically accounted for without anyone having to lift a finger.
Not at all, and you're thinking about far too simple a model.
/dev/null. Mail would have to be carefully crafted to achieve a 20 unless it truly is spam. Such effort is not to be rewarded :-) Keep in mind that mail about laws on sexuality or other simple examples cited in this article would never get NEAR a 20.
/dev/null. This is a very nice way to identify that a) there's a consensus that this very message is spam and b) my local mail patterns indicate that this is spam.
/dev/null to a mailbox that I scanned carefully. None of the messages that I would have thrown out were non-spam.
With SpamAssassin, I deal with spam in 3 ways:
1. Mail that gets a score of 20 or more is sent to
2. Mail that triggers both the Bayesian and Razor2 tests is sent to
3. Anything else with a score of 4 or more is marked in the subject line and I have a virtual mailbox in my mail client that I use to glance at the from addresses. If something looks plausible, I check it out.
As of the development version of SpamAssassin that I'm using (about a week old out of CVS), I get a false positive rate of about 1:100-200 messages and during testing over the last couple of months, I copied the messages that would have gone to
I get a LOT of mail form lists, spammers, friends, random people on the net, machines spewing status, etc. I feel that I'm a reasonably good QAer for this sort of thing, and the new SpamAssassin will rock your world (and the spammers')!
I'm not sure if I agree with the idea or not, but here's what I think the reasoning is:
/. story concerning file sharing is likely to show up in this section regardless of its specific relation to rights or lack thereof.
File-sharing in general is at the center of a huge right-to-share/fair use vs. copyright/profit debate right now. Thus, any
Hope that helps.
I must be missing something fundamental to XRender. It's mostly focused on anti-aliased fonts and alpha blending both of which it will be used for by Gnome when XRender adoption becomes a bit more universal (Sun (Solaris 9) and XFree86 (version 4) are, AFAIK, the only X displays to offer this extension currently). XRender's big advantage, in fact, is that it doesn't require you to interact with X in a radically different way.
Suggesting that not using such a sparsely available and relatively new technology to do something that it's not intended for is "misprogram[ming] X" seems to be way out of line to me.
OpenGL is not the right way to go, as it's designed for an entirely different purpose.
The correct solution is to re-think the way X works (specifically with regards to XImage data, Colormaps, Fonts and drawing primatives), and design it to use modern graphics cards correctly. That's not going to happen tomorrow, but it is happening. KDE/Qt and Gnome/GTK+ will take advantage of such features at the become available.
vi and Emacs are products of the era in which they were built, and both were considered excellent UIs for the day. Given today's advances in UI design, I find it interesting that those two are still the best ways to edit text. There *should* be better, but there simply isn't. There are some basic reasons for this that are interesting and fly in the face of a lot of UI design:
[note: I'm talking about Emacs as a text-editor, which is a fair slice of its functionality to discuss, though certainly not all of it]
Both editors were designed in an era when the mouse wasn't an option. This means that even given recent modifications to add mouse support in vim and Lucid Emacs (GNU Emacs had primative mouse support first, but Lucid modernized it, and much of their work was merged back into GNU Emacs, though not all, hence Lucid's transition to the even-today-maintained XEmacs) in the last 10 or so years, they still remain useful when your fingers never leave the keyboard.
The other thing that both of these do that puts them head-and-shoulders above many related tools is interaction with external programs. In vi or Emacs, I can send a portion of a document through any external program I want and read back the result into the original document. Here's an example in vi:This sorts the lines from the current position to the bottom of the file based on the numeric value of the third column. To add this feature into the editor would be fine (Emacs can do this actually), but if an external tool already knows how to do it, why not take advantage of that fact?
I have to disagree. First, a good HIG does not constrain you to do the wrong thing for your application, but rather gives you a sense of how best to fit into a particular paradigm. From there, you can do what you need to. A classic example of where this is important is the current trend in "media players" like Quicktime from Apple, WinAmp, XMMS, Microsoft Media Player, etc. They all try to look as snazzy as possible, at the cost of good user interface design. I can't count the number of times I've been at a party where some geek with good intentions put up an XMMS or WinAmp-based juke-box for people to play music with and all of the non-techies would give up after a few minutes because they couldn't decipher the UI.
If those applications had followed the UI guidelines for the platforms they run on, they could still have had all of the features, all of the great flexibility, but they didn't need to have pseudo-round volume sliders, non-standard title-bars that do application-local window management, context-sensitive menus that don't have commonly-performed operations, out-of-the-box unique font selection, out-of-the-box unique color selection, etc, etc.
That kind of awful behavior is what makes a desktop unusable (and certainly if more apps go the "branded UI" route, dekstops will become totally unusable).
But you will certainly get 100 different ways to do something and 100 different widget sets.
You may also get common user interface standards, but that has little or nor bearing on the above.
Also, pointing to Apple as an example good HIG is probably not the moral high-ground it once was, given recent software from that company (e.g. the latest Quicktime) pointedly ignores such things....
Check out the GNOME HIG and the effort linked to by this article to combine with KDE's interface guidelines. Certainly those are good places to start.
Which apple breaks left, right and sideways in the most recent Quicktime players. Guess no one is immune to shinny-over-substance. :-/
Incorrect. They are combining the two documents into one, with sections that conflict annotated in XML such that either document can be viewed as a stand-alone. Then they are going to work on the differences.
Read the announcement on the mailing list archives.