I'm thinking the same thing with the Dell phone. Personally, my ancient Handspring PDA died a few months ago, so I got a new phone with PDA features. It's not as nice as the PDA (needs expensive cable to sync with Outlook, less memory, fewer features, less software available), but it does the job ok. To my surprise, T9 is a much faster text input method than the stylus on my Visor.
The one thing I wish it did besides being a better PDA is music playing. I'd like a phone that took a CF or SD card full of MP3 or WMA files and was able to play them, but it seems that's the realm of $$$ PDA phones that I can't afford. Since Dell is so notably good at driving prices down, I have hopes that I may be able to get a music playing Dell phone.
Branden Robinson led the Debian X strike force that patched and packaged XFree86 for Debian. During that time, he had a graphic on the top of the X strike force page telling users to "Have a nice cup of shut the fuck up!" When I first installed Debian and wondered if my video card would be supported any time soon, that was the first place I looked, and I was mildly offended. At the time I continued using Debian, and just built my own copy of X for a while, but I would say that his attitude is one of the significant reasons I no longer use Debian.
Maybe he's changed since then, and maybe the attitude problem was more one of poor communication than of obnoxiousness. I don't know him personally, so I'm not the most qualified judge, but I do not consider his election a good thing for Debian. Leaders should ideally be good at communicating, and less good at ignoring and insulting people, and what I've seen of him reflects those negative traits more than the positive one.
Wow that sounds nice. I could search all my loads of PDFs of scientific papers, with results updating live as I type. Yup, if I couldn't already do that using MSN desktop search with Adobe's Acrobat IFilter plugin already, I'd be really excited.
Seems like typical Apple nonsense to me. Copy something that someone else did first, and then claim it's the other way around. I can see why Steve Jobs is going on the offensive, though. If Apple didn't have such great marketing they would be on the short track to irrelevance.
I'm currently in grad school in computer engineering. I know that it's not a financially prudent move, but I enjoy it, so there. I claim that I was one of the best students in my class when I got my bachelor's degree. I also claim that the research group I am in now is highly competent. I turned down some good job offers (+70k USD) after undergrad, and I see that similar and better opportunities are available to people in my group upon graduation.
Among my friends from undergrad, I note that the ones who are very intelligent and communicate reasonably well got good jobs (similar to what I was offered). People who were not so good typically got jobs that are not so rewarding, financially or intellectually. The only exceptional cases I know of are one friend who has quite unpolished english (he speaks fluently, but his speech does not reflect his high intelligence), and one other friend who decided that she wanted a boring job so that she could spend more time partying. Both of them got jobs that I would not consider enjoyable or well paying. Other than that, my smart friends got good jobs.
From my current research group, the sample space is much smaller, but the people I know who have graduated from the group are all gainfully employed in jobs that seem interesting.
My overall impression is that the job market now is smaller than it was several years ago, but that the shrinking has occured mainly at the expense of less qualified people. Intelligent people with any motivation still seem able to come up with jobs in the computing field that pay well and offer some level of interest.
Re:Is anyone else curious what SSA trees are?
on
GCC 4.0.0 Released
·
· Score: 4, Informative
As other people have said, SSA is static single assignment. It means that each variable in the program is assigned in only one place. SSA is for optimization, and is usually done in intermediate forms generated by the compiler, rather than in programs written by a human in common computer languages such as C, C++, Perl or assembly languages.
Trying to recall my knowledge of optimizing compilers:
SSA makes optimization easier, since it is obvious where a variable was assigned (since it was assigned in only one location) and what value it contains (since there is only one value being assigned to it). The complexity moves to register allocation, where there can be many more variables to allocate because of SSA. Register allocation is Hard, but doing an ok job is quite possible. Most optimizations are impossible unless you can prove various properties about the variables involved, which is often much easier with variables in SSA form.
Well, there was the whole stagnation on 2, leading to the egcs fork and eventual reconciliation with the FSF branch. So it's not really surprising that development is happening a whole lot faster now.
The most popular formats that embed font information are pdf and ps. You are SOL if you use a proprietary font that can not be redistributed, but most (not all, though) proprietary fonts permit redistribution in that form. I do check the licenses of proprietary fonts that I get. I have some that do not permit commercial use, but all of the font I have permit me to redistribute them embedded into a document, and permit others to display, print, and redistribute that document.
Perhaps you could learn a bit by reading some more documents on the architecture of the K8. There are definite advantages to its architecture, but reduced cache coherence traffic is not one of them. I note that other people have already explained the northbridge situation in more detail than I am familiar with.
So what you are saying is that AMD CPUs have more overhead due to cache coherency traffic on the point-to-point CPU links, whereas Intel CPUs don't generate cache cache coherency traffic except on invalid misses, since they can snoop the shared memory bus?
And perhaps you could clear up for me what the northbridge for a newer AMD CPU does. I thought the main function of the northbridge was the memory controller, which is included on die on newer AMD CPUs.
This is totally offtopic, but that post reminds me of a conversation I had with a professor a while ago. We were commenting about how little the frosh understood computers, and he made the comment that many of them saw a keyboard as a device to display characters on a monitor.
This perception (which I agree is common) makes it difficult to teach the students UNIX or programming, because the students don't see the keyboard as something that can give commands to the computer. You would be further ahead thinking of a keyboard as an input device that is typically used to enter text. Without limiting other uses that I'm not mentioning, that text can be displayed on a monitor, interpreted as instructions for the computer, or it can be the source code of a computer program.
I believe it does not fix the issues I've previously mentioned that prevent me from using OO.o. I will check it out when it is released, but last I checked on the bugs list, my issues had been reported many months previously and were still unassigned. I actually find MS Office 2003 to be an excellent suite. The new help features make it quite easy for me to find features in Word that I wouldn't know about in OO.o, so I do have some reasons for sticking with MS Office, even if the OO.o people manage to fix the basic functionality issues that are keeping me away. I would probably be much more inclined to write documents on my Linux systems if I had a useable word processor that would run on them, though.
I'm an advocate of intellectual property rights, which puts me somewhat in the minority here. Still, this goes too far.
Notice where the lawyer points out that Red Hat does not permit unauthorized linking to their website? Since when does using the Red Hat name along with a link to the Red Hat corporate website cause confusion about who you are? Eliminating every possibility of confusion and building brand identity is fine, but this is just stupid.
Come on, Red Hat. Just because you fancy yourselves competing with Sun, Microsoft and IBM doesn't mean you have to behave more obnoxiously than they do.
10 characters is too short; 10 character alphanumeric passwords can be brute forced, even if they are completely random. If they include dictionary words, then the attack will take mere seconds. At only 10 characters you need punctuation as well.
A further problem with passwords sent over the wire is that they are vulnerable to timing attacks. By checking the timing of packets containing password keystrokes, it is possible to reduce the amount of randomness even more. I don't know of any software actually doing this, but it may become a problem in the future.
A better approach is to use a password protected RSA key. The key will not be as vulnerable to theft, and you avoid having to send the password over the wire. Using ssh-agent along with an RSA key protected by a long passphrase is probably the best solution, unless you move between computers a lot.
There is still the problem of revealing passwords and passphrases to compromised machines. I don't know how to deal with that. I suspect that a combination of passphrases and secure hardware (trusted computing style) will be able to ensure that there is no single point of failure.
Anyway, I'm not sure that what you've got now is insufficient; it's probably fine. However, the best available practises are quite a bit stronger than what you are doing now.
I'd expect the clock on the PPC core to be 2.3 GHz, especially as it has been indicated that the core is a relatively simple in-order design. That would be half of the 4.6 GHz clock of the fastest units on the cell processor, in the same way that the older 3.4 GHz (for example) Pentium 4 processors run at half of the 6.8 GHz clock of the fastest units on the chip.
Now compared to a 1.25 GHz Macintosh G4, the 2.3 GHz cell chip is probably faster. It is certainly slower on a per-clock basis, with no out of order execution, and a narrow issue design. On the other hand, it is running at almost twice the clock speed, so it likely will be faster regardless.
As for Apple using it, since the cell is certainly slower for most tasks than a PPC970, it won't be used in the high end desktops, or even the iMacs. It seems rather silly to put an FP monster chip in a low-end notebook, but it is possible. The Powerbooks seem the most likely place for Apple to use a cell processor. It would depend on two things. First, power consumptions of a huge 8-core chip. The cell processor will probably use much less power than a Pentium 4, but exactly how much is unknown right now. The second thing is software. That would be a tough sell for Apple, since unlike Altivec there is no promise that it would eventually be available in the whole Macintosh line. You'd basically be asking technical application authors to spend a whole lot of effort to optimize for the Powerbook line only.
How well you can use 8 dsps really depends on your code. I'd guess in most cases the answer is no, you can't use the vector units to make up for the lost performance of the main core. If you got effective use of VMX, then you might be able to, because easily vectorizable calculations should be possible to port to the dsps more often than most code.
With the low performance PPC cpu, I doubt Apple will want these things. Apple has too much interest in the general purpose computer market to care much about something like the cell processor, that is built for a niche market. For Apple, the cell processor will not only be expensive, but also slow. That goes double, since it is harder for Apple to get developers to suddenly switch from Altivec to the cell architecture than it is for Sony to do the same.
So somewhere along promoting "grid computing" as the next big thing, Sun and IBM (and presumably HP, although I haven't heard from them yet) decided that "use your organizations numerous spare computing resources" meant "pay us to use our computing resources". Somehow, I'm not surprised by this.
Grid computing always seemed like a stupid idea to me. Computing resources are cheap; it's managing them and setting them up that is expensive. Dedicated machines are cheaper to manage and easier to set up than grids of under-utilized computers. Now IBM or Sun gets the cost savings of a dedicated grid of cheap hardware, and you get to pay for to do the tricky stuff, every time you want to run it.
At least this business model has the profit step really obvious. I pity the people who get to work with the IBM or Sun (or even internal) grids, though, and I pity even more the people who get laid off so that their salaries can be redirected to funding grid rentals from IBM or sun.
Sure you could ban by MAC address. It would be easy! You'd just have to convince users to honestly and accurately enter a 12 digit hexadecimal number that they don't know in the first place. Assuming they got all that right, and the number was wrong, banning them would be a trivial check against your blacklist.
Obviously, Jupiter Research Senior Analyst Joe Wilcox didn't bother reading the license that Microsoft posted, along with the specifications for their Office XML file formats. The older formats are not open, but the new ones certainly are.
The license (two licenses, actually; one for the specification, and another for all MS patents that cover it) may not be GPL compatible, but it sure looks compatible with other open source licenses, including so-called viral licenses.
The catch with the GPL is the additional restrictions part. Microsoft adds two restrictions, both of which are more-or-less reasonable. The first is the obnoxious BSD-like advertising clause; that's irritating, but not a showstopper except for the GPL. The second restriction is that implementations must be conformant to be distributable. That is, reading and writing done by implementations based on the spec must read or write valid Office XML files. Since the format is a well-designed XML format, this is trivially easy to do. The requirement is to prevent forking of Office XML formats, which is obviously a concern of Microsoft's. Again, it would be nice for developers if this restriction was removed, but it would be detrimental to both Microsoft and Microsoft's customers.
Apple sells all of their LCD displays at 96 dpi. They claim it is the optimum resolution. Clearly bogus, since I hold my laptop a lot closer to my face than my desktop, but whatever. I find 96 dpi uncomfortably large, and really like my 145 dpi pc laptop. Other people have noted the small fonts, though.
One thing you can do with Windows is adjust the resolution. Windows XP calls 96 dpi standard resolution and 120 dpi high resolution. If you change the DPI setting under display properties->advanced->general->DPI setting, Windows will scale text and graphics to fit your new resolution. It looks pretty ugly, though, since most icons are small raster graphics that look really bad when scaled. Windows Longhorn is supposed to ship with vector graphics to fix that... I just leave it at the default setting, since I like small fonts, even if it means 1 cm on my screen doesn't correspond exactly to 1 cm on paper.
According to the judge in the anti-Rambus case, Rambus did disclose their patents, and their intent to charge for them. He went so far as to say that he would have charged the manufacturers for conspiracy to put Rambus out of business in order to obtain their IP, except that he believed that that was outside the jurisdiction of the case he was trying. That finding is likely making it much easier for Rambus to make good on their patent claims.
It's a tough act for Rambus to carry out; on the one hand, they have to deal with a small group of manufacturers who have (reportedly) been trying to defraud them and put them out of business, on the other hand, they have to rely on that same small group of manufacturers for all of their future revenue, so aggravating them too much is probably also a bad idea.
Of course, it's also possible that the judge was Just Plain Wrong, and Rambus was just trying to get submarine patents in place while they were a member of JEDEC. I don't have the expertise to make that judgement.
Microsoft really wants to be a good guy. The thing that remains to be seen is if they are any good at it. This is the latest move. Earlier moves include lots of customer communication initiatives, encouraging employee blogs, and settling open legal issues so that Microsoft is not seen as happy to be in court.
Microsoft is losing customers, particularly European and American state governments, because they don't like Microsoft. Microsoft really does have the best office suite in a technical sense. OO.o is generally less intuitive, and has less features (particularly in spreadsheets, but even the word processor lacks much advanced functionality). Costs are hard to judge, but most studies suggest that using a free office suite instead of MS Office won't pay off over the time periods that corporations and governments make long range financial plans. Switching to OO.o is about politics, not technical or financial superiority.
It's also difficult to switch right now, partially because of proprietary lock-in to the file format. That's one of the things that makes switching so expensive (although probably not the major one, with OO.o import filters being somewhat decent). Customers want to be free to switch. They also want to be free to generate documents from sources other than MS Office and import them natively, and they want to be able to process documents using their own custom tools. Open file formats help all of those things, and so customers are happy.
Microsoft really wants to make customers happy. Opening file formats helps, so Microsoft is doing that. There are risks; if customers continue to hate Microsoft, and Microsoft makes it easier to switch away from them, the obvious result is losing customers. The upside is that they may make customers happy, convincing more to stay. Being a nice guy is directly connected to making customers happy.
From an open-source community view, opening file formats is good. It makes interoperability easier. By itself, though, it's not enough to make customers happy, or to make Microsoft a friend to the OSS community. More moves are necessary, and what they are and when (if ever) the will come is still a big question.
Just a question here, what would Microsoft have to do for you to consider them to be a friendly corporation, rather than an evil and menacing corporate giant? I kind of like them already, but I know I'm unusual in that regard.
Your vendor instead thinks it would be better to do:
gcc -o f f.c
Where f.c looks like:
#include "f1.c" #include "f2.c" #include "f3.c"
Am I right, or am I completely off track?
If I'm right, you'd probably still want to include header files because you want everything to remain modular. According to software engineering type people, that makes maintenance easier. Another problem is symbol scoping. C keeps symbols local to the module they appear in, so you want to make sure you have naming conventions, namespaces, or some other protection against naming clashes. I'm dubious about the benefits, but I work on projects that take significant amounts of time to compile. Not hours, but enough time that if you wait for all the objects to compile you are wasting a lot of time. In general, I'd claim that the larger the project, the worse an idea it is.
I wouldn't mind seeing InDesign or GoLive in there, since both can legitimately contain lots of text. I'm a bit surprised to see Photoshop and Illustrator in there, but the support seems pretty useless since Photoshop support is version 4 only (current version: 8) and Illustrator support stops at version 9 (current version: 11).
The inclusion of lots of old DOS formats seems pretty gimicky (as opposed to useful), although it seems they just took whatever file formats the people they licensed the technology from support. Perhaps there is a legitimate reason for a more specialized search tool to support all those obsolete file types.
If you really edit movies regularly, you should invest in some more memory. It will save you a lot of time. My media creation requirements are limited to still photographs, and I still find less than one gigabyte to be a huge and time-consuming pain. I've run Photoshop and Visual Studio together with as little as 20 megabytes of memory (no, really, I did), and these programs degrade gracefully in performance, but they really do much better with more memory.
The iLife apps are nice, no doubt. They are an advantage that Apple has over a PC. Still, the base model Mini Mac doesn't include a DVD burner. The whitebox PC doesn't even include a DVD reader. Combine that with the low memory, and media creation options are limited. The iWork stuff isn't bundled; it is to compete with Microsoft Office, and will cost you a pretty penny.
There are two big reasons to stick with Windows. One is that people are more likely to be familiar with it. Since this machine is aimed at Windows users, this will be a factor to people buying it. The second reason is for all that stupid junk you can download and run. I'm sure you don't do that; I don't either. Most of the target audience for this machine do download all sorts of crap and just expect it to work (or at least to install).
There are a number of nice operating systems out there, including OS X, Solaris and IRIX. The market seems to have chosen Windows and Linux, though. Running something else is a liability, not an asset, regardless of how nice the operating system is.
Slashdot Mirror
The one thing I wish it did besides being a better PDA is music playing. I'd like a phone that took a CF or SD card full of MP3 or WMA files and was able to play them, but it seems that's the realm of $$$ PDA phones that I can't afford. Since Dell is so notably good at driving prices down, I have hopes that I may be able to get a music playing Dell phone.
Maybe he's changed since then, and maybe the attitude problem was more one of poor communication than of obnoxiousness. I don't know him personally, so I'm not the most qualified judge, but I do not consider his election a good thing for Debian. Leaders should ideally be good at communicating, and less good at ignoring and insulting people, and what I've seen of him reflects those negative traits more than the positive one.
Seems like typical Apple nonsense to me. Copy something that someone else did first, and then claim it's the other way around. I can see why Steve Jobs is going on the offensive, though. If Apple didn't have such great marketing they would be on the short track to irrelevance.
Among my friends from undergrad, I note that the ones who are very intelligent and communicate reasonably well got good jobs (similar to what I was offered). People who were not so good typically got jobs that are not so rewarding, financially or intellectually. The only exceptional cases I know of are one friend who has quite unpolished english (he speaks fluently, but his speech does not reflect his high intelligence), and one other friend who decided that she wanted a boring job so that she could spend more time partying. Both of them got jobs that I would not consider enjoyable or well paying. Other than that, my smart friends got good jobs.
From my current research group, the sample space is much smaller, but the people I know who have graduated from the group are all gainfully employed in jobs that seem interesting.
My overall impression is that the job market now is smaller than it was several years ago, but that the shrinking has occured mainly at the expense of less qualified people. Intelligent people with any motivation still seem able to come up with jobs in the computing field that pay well and offer some level of interest.
Trying to recall my knowledge of optimizing compilers:
SSA makes optimization easier, since it is obvious where a variable was assigned (since it was assigned in only one location) and what value it contains (since there is only one value being assigned to it). The complexity moves to register allocation, where there can be many more variables to allocate because of SSA. Register allocation is Hard, but doing an ok job is quite possible. Most optimizations are impossible unless you can prove various properties about the variables involved, which is often much easier with variables in SSA form.
Well, there was the whole stagnation on 2, leading to the egcs fork and eventual reconciliation with the FSF branch. So it's not really surprising that development is happening a whole lot faster now.
The most popular formats that embed font information are pdf and ps. You are SOL if you use a proprietary font that can not be redistributed, but most (not all, though) proprietary fonts permit redistribution in that form. I do check the licenses of proprietary fonts that I get. I have some that do not permit commercial use, but all of the font I have permit me to redistribute them embedded into a document, and permit others to display, print, and redistribute that document.
Perhaps you could learn a bit by reading some more documents on the architecture of the K8. There are definite advantages to its architecture, but reduced cache coherence traffic is not one of them. I note that other people have already explained the northbridge situation in more detail than I am familiar with.
So what you are saying is that AMD CPUs have more overhead due to cache coherency traffic on the point-to-point CPU links, whereas Intel CPUs don't generate cache cache coherency traffic except on invalid misses, since they can snoop the shared memory bus? And perhaps you could clear up for me what the northbridge for a newer AMD CPU does. I thought the main function of the northbridge was the memory controller, which is included on die on newer AMD CPUs.
This is totally offtopic, but that post reminds me of a conversation I had with a professor a while ago. We were commenting about how little the frosh understood computers, and he made the comment that many of them saw a keyboard as a device to display characters on a monitor. This perception (which I agree is common) makes it difficult to teach the students UNIX or programming, because the students don't see the keyboard as something that can give commands to the computer. You would be further ahead thinking of a keyboard as an input device that is typically used to enter text. Without limiting other uses that I'm not mentioning, that text can be displayed on a monitor, interpreted as instructions for the computer, or it can be the source code of a computer program.
MSN's desktop search tool will search PDF files if you install Adobe's Acrobat IFilter plugin. I've found it valuable several times.
I believe it does not fix the issues I've previously mentioned that prevent me from using OO.o. I will check it out when it is released, but last I checked on the bugs list, my issues had been reported many months previously and were still unassigned. I actually find MS Office 2003 to be an excellent suite. The new help features make it quite easy for me to find features in Word that I wouldn't know about in OO.o, so I do have some reasons for sticking with MS Office, even if the OO.o people manage to fix the basic functionality issues that are keeping me away. I would probably be much more inclined to write documents on my Linux systems if I had a useable word processor that would run on them, though.
Notice where the lawyer points out that Red Hat does not permit unauthorized linking to their website? Since when does using the Red Hat name along with a link to the Red Hat corporate website cause confusion about who you are? Eliminating every possibility of confusion and building brand identity is fine, but this is just stupid.
Come on, Red Hat. Just because you fancy yourselves competing with Sun, Microsoft and IBM doesn't mean you have to behave more obnoxiously than they do.
A further problem with passwords sent over the wire is that they are vulnerable to timing attacks. By checking the timing of packets containing password keystrokes, it is possible to reduce the amount of randomness even more. I don't know of any software actually doing this, but it may become a problem in the future.
A better approach is to use a password protected RSA key. The key will not be as vulnerable to theft, and you avoid having to send the password over the wire. Using ssh-agent along with an RSA key protected by a long passphrase is probably the best solution, unless you move between computers a lot.
There is still the problem of revealing passwords and passphrases to compromised machines. I don't know how to deal with that. I suspect that a combination of passphrases and secure hardware (trusted computing style) will be able to ensure that there is no single point of failure.
Anyway, I'm not sure that what you've got now is insufficient; it's probably fine. However, the best available practises are quite a bit stronger than what you are doing now.
Now compared to a 1.25 GHz Macintosh G4, the 2.3 GHz cell chip is probably faster. It is certainly slower on a per-clock basis, with no out of order execution, and a narrow issue design. On the other hand, it is running at almost twice the clock speed, so it likely will be faster regardless.
As for Apple using it, since the cell is certainly slower for most tasks than a PPC970, it won't be used in the high end desktops, or even the iMacs. It seems rather silly to put an FP monster chip in a low-end notebook, but it is possible. The Powerbooks seem the most likely place for Apple to use a cell processor. It would depend on two things. First, power consumptions of a huge 8-core chip. The cell processor will probably use much less power than a Pentium 4, but exactly how much is unknown right now. The second thing is software. That would be a tough sell for Apple, since unlike Altivec there is no promise that it would eventually be available in the whole Macintosh line. You'd basically be asking technical application authors to spend a whole lot of effort to optimize for the Powerbook line only.
With the low performance PPC cpu, I doubt Apple will want these things. Apple has too much interest in the general purpose computer market to care much about something like the cell processor, that is built for a niche market. For Apple, the cell processor will not only be expensive, but also slow. That goes double, since it is harder for Apple to get developers to suddenly switch from Altivec to the cell architecture than it is for Sony to do the same.
Grid computing always seemed like a stupid idea to me. Computing resources are cheap; it's managing them and setting them up that is expensive. Dedicated machines are cheaper to manage and easier to set up than grids of under-utilized computers. Now IBM or Sun gets the cost savings of a dedicated grid of cheap hardware, and you get to pay for to do the tricky stuff, every time you want to run it.
At least this business model has the profit step really obvious. I pity the people who get to work with the IBM or Sun (or even internal) grids, though, and I pity even more the people who get laid off so that their salaries can be redirected to funding grid rentals from IBM or sun.
It's the perfect solution!
The license (two licenses, actually; one for the specification, and another for all MS patents that cover it) may not be GPL compatible, but it sure looks compatible with other open source licenses, including so-called viral licenses.
The catch with the GPL is the additional restrictions part. Microsoft adds two restrictions, both of which are more-or-less reasonable. The first is the obnoxious BSD-like advertising clause; that's irritating, but not a showstopper except for the GPL. The second restriction is that implementations must be conformant to be distributable. That is, reading and writing done by implementations based on the spec must read or write valid Office XML files. Since the format is a well-designed XML format, this is trivially easy to do. The requirement is to prevent forking of Office XML formats, which is obviously a concern of Microsoft's. Again, it would be nice for developers if this restriction was removed, but it would be detrimental to both Microsoft and Microsoft's customers.
One thing you can do with Windows is adjust the resolution. Windows XP calls 96 dpi standard resolution and 120 dpi high resolution. If you change the DPI setting under display properties->advanced->general->DPI setting, Windows will scale text and graphics to fit your new resolution. It looks pretty ugly, though, since most icons are small raster graphics that look really bad when scaled. Windows Longhorn is supposed to ship with vector graphics to fix that... I just leave it at the default setting, since I like small fonts, even if it means 1 cm on my screen doesn't correspond exactly to 1 cm on paper.
It's a tough act for Rambus to carry out; on the one hand, they have to deal with a small group of manufacturers who have (reportedly) been trying to defraud them and put them out of business, on the other hand, they have to rely on that same small group of manufacturers for all of their future revenue, so aggravating them too much is probably also a bad idea.
Of course, it's also possible that the judge was Just Plain Wrong, and Rambus was just trying to get submarine patents in place while they were a member of JEDEC. I don't have the expertise to make that judgement.
Microsoft is losing customers, particularly European and American state governments, because they don't like Microsoft. Microsoft really does have the best office suite in a technical sense. OO.o is generally less intuitive, and has less features (particularly in spreadsheets, but even the word processor lacks much advanced functionality). Costs are hard to judge, but most studies suggest that using a free office suite instead of MS Office won't pay off over the time periods that corporations and governments make long range financial plans. Switching to OO.o is about politics, not technical or financial superiority.
It's also difficult to switch right now, partially because of proprietary lock-in to the file format. That's one of the things that makes switching so expensive (although probably not the major one, with OO.o import filters being somewhat decent). Customers want to be free to switch. They also want to be free to generate documents from sources other than MS Office and import them natively, and they want to be able to process documents using their own custom tools. Open file formats help all of those things, and so customers are happy.
Microsoft really wants to make customers happy. Opening file formats helps, so Microsoft is doing that. There are risks; if customers continue to hate Microsoft, and Microsoft makes it easier to switch away from them, the obvious result is losing customers. The upside is that they may make customers happy, convincing more to stay. Being a nice guy is directly connected to making customers happy.
From an open-source community view, opening file formats is good. It makes interoperability easier. By itself, though, it's not enough to make customers happy, or to make Microsoft a friend to the OSS community. More moves are necessary, and what they are and when (if ever) the will come is still a big question.
Just a question here, what would Microsoft have to do for you to consider them to be a friendly corporation, rather than an evil and menacing corporate giant? I kind of like them already, but I know I'm unusual in that regard.
gcc -c f1.c gcc -c f2.c gcc -c f3.c gcc -o f f1.o f2.o f3.o
Your vendor instead thinks it would be better to do:
gcc -o f f.c
Where f.c looks like:
Am I right, or am I completely off track?
If I'm right, you'd probably still want to include header files because you want everything to remain modular. According to software engineering type people, that makes maintenance easier. Another problem is symbol scoping. C keeps symbols local to the module they appear in, so you want to make sure you have naming conventions, namespaces, or some other protection against naming clashes. I'm dubious about the benefits, but I work on projects that take significant amounts of time to compile. Not hours, but enough time that if you wait for all the objects to compile you are wasting a lot of time. In general, I'd claim that the larger the project, the worse an idea it is.
The inclusion of lots of old DOS formats seems pretty gimicky (as opposed to useful), although it seems they just took whatever file formats the people they licensed the technology from support. Perhaps there is a legitimate reason for a more specialized search tool to support all those obsolete file types.
The iLife apps are nice, no doubt. They are an advantage that Apple has over a PC. Still, the base model Mini Mac doesn't include a DVD burner. The whitebox PC doesn't even include a DVD reader. Combine that with the low memory, and media creation options are limited. The iWork stuff isn't bundled; it is to compete with Microsoft Office, and will cost you a pretty penny.
There are two big reasons to stick with Windows. One is that people are more likely to be familiar with it. Since this machine is aimed at Windows users, this will be a factor to people buying it. The second reason is for all that stupid junk you can download and run. I'm sure you don't do that; I don't either. Most of the target audience for this machine do download all sorts of crap and just expect it to work (or at least to install).
There are a number of nice operating systems out there, including OS X, Solaris and IRIX. The market seems to have chosen Windows and Linux, though. Running something else is a liability, not an asset, regardless of how nice the operating system is.