I doubt we are the audience. First, we are more likely to consider memory and processor speed important. The limit of one gigabyte and a 1.4 GHz G4 PowerPC processor are both major disadvantages to a geek audience.
As for the OS, you are right, I didn't factor it in. Once you add $120 Canadian for an OEM copy of Windows XP you are around the same price as the Mini Mac. I do expect piracy of the operating system on systems like the whitebox AMD I compared to, but even including a legal copy of Windows, the PC at worst goes to price parity.
The size of the Mini Mac is very nice, but that isn't something that most people think about when buying a computer. The volume is probably smaller than many laptops, even if you sliced off the display and keyboard. I don't see that being a decidng factor to many people, though.
I think your points are good, but I still don't think the Mini Mac will be compelling to a large enough audience to materially affect Apple's success.
The Celeron processor is still faster than the 1.25 GHz G4 PowerPC. The integrated graphics is definitely a minus for both the whitebox AMD machine and the Dell machine. The Dell machine is comparable to the whitebox AMD machine in every category except memory, where it is better.
I consider price, peripherals and operating system to be the key points, since those are the things that will most directly affect the end user experience in the target market. Processor, video, memory, warranty, connectibility--who cares? The target market is people who want a new PC, have a few hundred dollars in their pocket, and like their ipod. The PC wins on two of the three most important points and gets a toss-up on the other. I'm sorry that I didn't make it clear in my original post that I consider some factors more important than others.
The AMD system I included is pre-built from a local whitebox vendor. You can go to their store and point to it on a sheet of paper (it's easy to find, because it's the cheapest one). They will stick the processor into the pre-assembled chassis and bring it out for you within fifteen minutes. You can also order it online and have it delivered anywhere in North America (although shipping would likely raise the cost above that of the Dell machine if you live outside of Toronto). It is a machine someone can buy on a whim.
The Mac Mini is aimed clearly at PC users looking to switch, but featurewise it is a disappointment.
It has a DVI video port (with VGA adapter included), and it comes with no monitor, mouse, or keyboard. Although Apple states on their website that "Mac Mini will take advantage of your two button USB mouse with scroll wheel", I think that decision is a mistake. Most PC users don't have USB peripherals, and will find the Mac Mini somewhat, err, nonfunctional upon purchase.
Hardware-wise, the specs are pretty low-end (what do you expect from a $500 Mac?). Comparing them to a local no-name store, here's what you get when comparing the cheapest options:
Price-the cheap PC is $450 Canadian, the cheap Mac is $500 US. The PC is significantly cheaper.
Processor-the Sempron 2200+ is slow, for a PC. It would quiver beside the 2.0+ GHz Powermac G5's, but it beats the crap out of the 1.25 (or even 1.4 GHz upgraded) Mac Mini. PC wins again.
Video-the Mac's discrete Radeon 9200 is an ok chip, the PC's integrated video isn't. Easy win for the Mac Mini.
Memory-they both come with 256 megs, so where's the contest? The Mac Mini is only upgradeable to 1 gigabyte. Slashdotters considering purchasing one may want to avoid it for this reason. Another victory for the PC.
Peripherals-the PC comes with a monitor, a mouse, a keyboard, and speakers. All cheap junk, but still a much better deal than the Mac, which doesn't come with any of that.
Optical drive-both come with CD burners, but the PC's doesn't read DVD's, and the Mac's does. Another point for the Mac Mini.
Cool factor-the Mac is much smaller and less of an eyesore. I don't know about volume, since neither manufacturer gives specs. This is a big plus for the Mac.
OS-Mac OS X versus no OS included. Victory for the Mac, although the PC will probably end up with Windows XP. Most users will prefer Windows since it is compatible with more programs and they are more likely to be familiar with it, although opinions can be very polarized about which is better. (=
Warranty-each offers a one year warranty. Victory for the Mac, since Apple service is much easier to get if you go outside of southern Ontario.
Connectability-the Mac offers a modem, IEEE1394 and DVI, as well as USB2 ports. The PC offers only VGA and USB2 for modern peripherals, although it has old style serial and parallel ports, and PS2 keyboard and mouse jacks that the Mac lacks. Both machines include 10/100 ethernet. I prefer the modern options, so the Mac gets a slight win.
Other specs such as hard drive and sound capabilities seem similar, so I won't use them for my conclusions.
To anyone who complains about the white box vs. brand name comparison, show me a white box Macintosh and I'll yield the point. But not much, since Dell offers a similar system (with more RAM plus Windows XP, but a Celeron processor and $499 Canadian price point).
Overall, the Mac just isn't a good deal compared to the PC. The price is significantly higher, and you still need to get a mouse and keyboard. Further, the limited memory expansion means that it isn't a good deal for hobbyists like the Slashdot crowd who would like to play with a Macintosh. I suspect sales will be decent, but that many customers will be unsatisfied when the performance is no better (or even worse) than their current PC, and they are forced to make a second trip to the store to shell out for a new mouse and keyboard. Although they may cash in on the current ipod craze, this machine is unlikely to win users for Apple in the longer term, or to stop Apple's market share from shrinking.
Since this is a poll I don't feel bad about saying me too, but I have a couple minor changes.
It's not necessarily strips that matter, but overall cost. Monitoring is very $$$ intensive now.
Continuous monitoring would be VERY nice. If not continuous monitoring, something that was convenient and cheap enough that I could test every twenty minutes or so would be great. That pretty much rules out an invasive meter; I wouldn't poke myself that often under normal circumstances. Bonus points if I can attach it to my finger or arm or whatever and have it continue to function test automatically in the night time.
A few words on insulin pumps. I don't have one, but my younger brother does. As a student, I can't afford one. Meter-pump integration would be nice, but you would obviously need reliable and regular monitoring for that. For anyone developing such a thing, a big part would be to convince insurance companies to pay for it.
Two other features that other people have mentioned that I second (or third, or so on) are alarms for bad results (or results that are headed in a bad direction, if the monitoring is regular enough to detect a trend) and easy to read displays that have some lighting. I would rather have that than ultra-long battery life, particularly if the meter can take some type of rechargeable battery (such as NiMH AA or AAA cells).
I think there are no UXGA tablets yet, but the Toshiba M200 has an SXGA+ display. I believe all Windows XP tablets that have active pens are pressure sensitive, and most tablet PCs have active pens. If you've got applications that don't understand your current pen's pressure sensitivity, try downloading Penabled software from Wacom. They make most of the pens and sensors, and their driver is required for applications like photoshop to read the pressure.
I agree UXGA or higher resolution would be nice, but you also need to consider that tablets are usually meant to be carried around, and a large and heavy device would be less convenient. I'd love to see more resolution in a small form factor, but I'd probably prefer my M200 to a 15" UXGA tablet.
I recently got a tablet PC, and a Windows XP tablet is a very nice thing. There are three very huge problems with Linux support for tablet PC's, though.
First is the lack of hardware support. If you happen to have a TC1000 then this guy has the drivers for you. If you happen to have an M200 (like I do), then he doesn't have the drivers for you.
Second is the lack of handwriting recognition. That's essential for using a tablet in, you know, tablet mode. Without it, even choosing to visit www.slashdot.com is a chore, and you can forget about word processing or email in the comfortable tablet form factor while riding the bus.
Third is the lack of applications. There are a few well chosen applications that support handwriting as a first class input mechanism. When scratching and scribbling on things it is comforting to have circles and lines, and even my messy handwriting, be the same as I put them in.
It would be nice to have Linux working well on my tablet, but the tablet PC is a new hardware and software platform. Microsoft doesn't have a great and polished interface for it yet, only one that is good enough. Still, every little bit of that new platform that Microsoft and others provide for Windows XP on a tablet is a little bit that Linux doesn't have yet at all.
Check out Photoshop actions for simple scripting capabilities; they are simple macros that you can record and playback.
If you need more complicated scripting capabilities, you can use Visual Basic, Javascript, or Applescript to drive Photoshop. Look at the scripting guide that is installed with Photoshop for more details.
You could also continue to use the GIMP, of course, but don't let lack of scripting stop you from using Photoshop, if that is what you prefer. Photoshop definitely has scripting capabilities.
If you click on it and tell it to "Install..." it pops up the annoying dialog that it used to show pre-SP2, with one very nice addition: a "don't install controls from this publisher, ever" button.
Are you sure that Word loads dll's at boot time? IE does, because it's largely the same program as Windows Explorer. I can't think of any reason why Word would do so. On the other hand, I run Outlook most of the time, which likely does share many libraries with Word, so I'll grant that Word does have an unfair advantage with respect to load times in my particular case.
On the other hand, I don't accept your claim that OO.o is Good Enough(TM). It isn't good enough for me (see the previous post for reasons why it isn't good enough), and anecdotal evidence suggests that it isn't good enough for many other people who have to do spend significant amounts of time in front of a word processor.
Correct formating of tables for imported Word documents. Some people insist on using Word as a database (I know, doesn't make sense to me either). OO.o makes it even harder to deal successfully with these people.
Don't lose graphics in imported Word documents.
When you export Word documents, they need to have file sizes that are similar to what they would have if you saved them with Word. I can't email someone back a document that has had a huge increase in file size. Word is bad enough with file sizes, but OO.o is much, much worse.
Don't crash so much. That's just annoying.
A grammar checker would be nice. Word and Wordperfect have had this for over a decade.
Faster load times would be great. Word loads in about one second on my computer; there is no excuse for OO.o taking more than ten seconds.
This is just a minor nit, but still... I use a text editor to edit text documents. OO.o shouldn't claim that its formatted word processor document is a text document.
The dialog box that asks if you are sure you want to export to a non-native file format because you might lose information should tell you what information you might lose. When I import a document, add a few sentences, then save it, I should not be seeing this nonsensical warning. In fairness, Word has this problem as well for some older formats, although not for Word 97 or later formats.
My most annoying point to me(since this one means I can't even use OO.o for documents that I distribute in pdf form only): support for using custom styles for section numbering.
Fix the last one of those and I will use OO.o again. Fix most of them and I will give it another try for regular use. Right now, though, OO.o is as useful to me as Wordperfect for the Atari ST is.
I've noticed how poorly OO.o deals with MS Word files. It does a miserable job of importing stuff with embedded graphics, and when exporting the files back to Word format, it quadruples the file size.
It's more an issue of OO.o's poor handling of MS Word file formats than it is of MS Word file formats themselves.
I consider this article to be firm proof that alternate browsers are a form of security through obscurity. Not that that is a bad thing if it works, and in this case it is clear that IE is being targeted more than its alternatives.
But make sure that your alternate browser it is a recent version of Firefox or Mozilla. They have responded very quickly to security issues, and are being proactive about security, much more so than the the people behind Konqueror or Opera. Also, keep your alternate browser patched just as vigilantly as you would Internet Explorer. As the popularity increases you will see more attacks against Mozilla based browsers.
I don't know what the answer to security is. I hope it isn't educating users, because that just plain doesn't work for most people. The problem is that right now there doesn't seem to be any other way.
Konqueror and Opera were affected by different vulnerabilities than Mozilla, but he demonstrated several in each browser. IE was the only browser not affected, indicating that it was the only browser getting basic testing for security.
The Mozilla team quickly corrected the issues he found with Mozilla, and started using his tool to perform basic security testing. Those bug fixes are in place for the Firefox 1.0 release, which is what I'm using on Linux.
Remember the recent Slashdot article wherein a security researcher demonstrated that Konqueror, Mozilla and Opera were bug-ridden security time-bombs waiting to happen? The Mozilla team took it to heart and quickly fixed the bugs he found, and started using his tool to check for bugs.
Anyone know if the Konqueror team did the same? I've Switched to Firefox because of this issue, but I really prefer Konqueror, and if I knew that they were being proactive with security I would switch back.
I think these are the two products we're talking about when Dell says that Red Hat Linux is too expensive. Comparing prices...
RHEL ES has two versions, priced at $350 and $800, depending on the support level. W2K3 SBS (Small Business Server) is available at different prices from different vendors, but is typically around $500. All prices in US dollars. The prices are quite similar. If you need support for more than installation and basic configuration, Windows 2003 is actually cheaper.
If small businesses find Windows easier to setup and maintain, then it could be worthwhile. I'm not able to personally confirm this one way or the other, but various people I know who have configured both Linux and Windows 2003 as servers claim that Windows is easier to configure and tune for performance.
Perhaps Dell simply means that for the market they are selling into and the price they are charging, there is a better product available from Microsoft. It's hard to see how Red Hat could compete on price; they really aren't charging a huge amount. For businesses that can't afford a full time server administrator and don't have any Linux expertise, it is quite possible that Windows just plain is a better choice.
Going off topic, Red Hat's website has the Ghandi quote that Slashdot loves: "First they ignore you, then they laugh at you, then they fight you, then you win." A year ago, Microsoft was fighting Red Hat. Now they are laughing at Red Hat. Linux still has a chance, but this battle definitely isn't going the way that Red Hat planned.
I hate to say it, but with IBM preferring Novell and SLES, I think Red Hat has lost.
I guess I let my irritation with OO.o show too much. I've had to deal way too much with documents in the past few months, and after some technical difficulties with LaTeX, I found that of OO.o Writer, KWord, AbiWord, ThinkFree Office and Microsoft Word, Microsoft Word was by far the best.
I'm fortunate enough to fall under a volume licensing plan for MS Office, so I got it for about CDN $90. If you can get one of the heavily discounted versions it's a worthwhile purchase, but otherwise it's too expensive for individuals who don't use it a lot.
I do have a friend who still uses Linux exclusively, and he uses OO.o to prepare presentations. He exports to pdf and uses Adobe Reader to view them. Reader has a full screen mode that is suitable for giving presentations. Go to the Window->Full Screen View menu item, or press Control-L. I doubt it handles animations or transition effects from OO.o, but if you can do without those it may be suitable.
I hate to sound troll-ish, but the solution to both your problem and the original poster's is to use Microsoft Office.
Microsoft makes stand-alone viewers for Powerpoint (97-2003) and Word (97-2000). Microsoft also makes a browser plugin for IE that understands WordML and will correctly display Word 2003 documents that are saved in XML format. The old viewer will show Word XP and non-XML Word 2003 documents, however some features may be unavailable, so the document may not look exactly right. You can fit any, or all, of the viewers on your 16 megabyte flash drive.
If alternate formats are an option, you can also convince a host of programs to read RTF. It's not as versatile as Word native format, but chances are any computer with an OS less than ten years old can read it.
OpenOffice.org is something that you suffer through when forced to use Linux to read or prepare editable documents for distribution. It's definitely not something you should ever force on anyone else.
P.S. When I used Linux exclusively, I distributed KOffice presentations by exporting to HTML format. Unfortunately KOffice does (or at least it did; it may be fixed by now) that by converting the presentations into graphics files, so it is resolution dependent. If you know what display resolution the other computer has, that will work. A better solution is probably to distribute presentations as PDF format, and use Adobe Reader as the small viewer program. I believe OpenOffice.org supports saving to PDF directly, and if not you can get a PDF printer driver.
While I completely agree with you that real pro photographers don't use Linux, you haven't done anything to solve this guy's problem.
To me it sounds like he's an amateur photographer who is just starting to experiment with digital. It's quite possible that the mediocre quality of The GIMP would suffice for him. Perhaps all he knows about colour management is "use Velvia". Further, it seems that his scanner is working with Linux.
It actually sounds like his hardware is the problem. He's got a cheap scanner with a slide adapter, and it gives him blurry results. If the results are blurry, you should try software first. Check that you are scanning at realistic resolution. If the scanner resolution is too high, drop the scanner resolution or downsample. Using that scanner with slides, that won't be a problem. If the scanner resolution is ok and the results are a bit soft, an unsharp mask should fix them. You can do that with The GIMP just as well as you can with Photoshop.
More likely he needs to invest in some decent hardware in order to make the setup work to his satisfaction. Windows or a Mac might be necessary, but solve the first problem first. If he isn't a pro, he might be able to use Linux for this.
My initial thought is that Sun would want two things that would make the GPL unsuitable. First, they would want the ability to distribute binaries containing both community contributions and proprietary bits that they may not be unwilling or unable to distribute. Second, they would want the anti-submarine warfare patent protection stuff, similar to what IBM put in the CPL.
Actually looking at the license, I see that it is based on the Mozilla license (MPL), which addresses the two issues I noted. Sun's changes remove the part about being covered by a future version of the license, and remove some notice requirements and clarify a few things that are unclear or poorly stated in the original MPL.
The license may well be GPL 3 compatible, since Stallman has made noises about wanting to clear up the patent protection stuff. You'd really need to get a lawyer's opinon on that, though, after the GPL 3 has been released.
The same DRAM manufacturers produced both SDR (and later DDR) DRAM, and Rambus DRAM. Rambus never had fabrication facilities for producing memory, so it relied on the DRAM manufacturers to license its intellectual property, then make and sell RDRAM.
The price fixing that Rambus alleged was a conspiracy to charge much more for Rambus memory than a competitive market would. Since the same group of companies controlled the supply of both types of memory, they could do this. Rambus claimed that Rambus memory should be about ten percent (IIRC) more expensive than SDR DRAM due to higher manufacturing costs and Rambus IP licensing. The DRAM manufacturers sold it for twice the price of SDR DRAM. If Rambus is correct about the costs, then their claim that the high price of Rambus memory was due to price fixing is plausible.
Based on the article, I believe this lawsuit is related to the allegations that Rambus made earlier. Rambus claimed that DRAM manufacturers were conspiring to price Rambus memory out of the market so they could drive the company out of business and get its technology cheaply.
Rambus said that Rambus memory should have been priced only slightly higher than SDR SDRAM, based on manufacturing and licensing costs, and that they were always clear with JEDEC on what patents they held and what licensing terms they wanted. Rambus said two things in particular: first, that the high cost of Rambus memory was due to a conspiracy between the DRAM manufacturers, and second, that the claim of submarine patents was a smear campaign by the DRAM manufacturers.
It seems from this guilty plea that there is enough evidence to prove at least the first of those claims in court. I personally don't have any special insight into who is right, but I do wonder if we were all wrong about Rambus.
Thank you for picking up on something intelligent rather than just adding noise. You're right, I wasn't very clear. Let me clarify.
You don't patent software, so even though the term "software patent" is in common use, it is really quite meaningless. Instead you patent an algorithm. I am in favour of algorithms being patentable, and more specifically, I am in favour of algorithms that can be implemented in software being patentable, with software implementations covered by such patents.
Each particular implementation is also covered under copyright law. I'm not saying anything about that in this context. Copyright is a different issue that can be treated separately.
I think this clarifies what you are saying, but if what I write is still difficult to understand, I will attempt to make it clear.
I know this is against conventional instinct here, but the majority of opinions on software patents I have seen presented on this site are so simplistic and obviously wrong that they don't deserve to be called thought.
Software patents are even more important than patents in other fields, due to the ease with which software techniques can be duplicated. Patents are absolutely necessary to protect small companies from having their ideas taken without any credit or compensation to the original source.
Much of the criticism against patents that has been leveled on this website is also driven by ignorance. People do not realize how specific patents are. I have seen posts on many patent articles here that read the first one or two claims and assume that a huge range of existing work is covered, without checking out the remaining claims that make it clear that one very small thing that is original is the actual target of the patent.
I won't defend the existing patent system too much, since it is flawed with respect to software. Since software patents are easy to implement, the costs required to develop them can be recuperated much faster than other areas, so a shorter duration of protection would probably be better. Some patents have been applied overly broadly, or granted when they were not merited.
But the abuses do not stop the patent system from being useful for software. The problems are things that can be worked out, not fundamental flaws with the idea of patenting software algorithms.
No one, and I mean no one, would use it unless it had two features. I'm just naming these two because I haven't seen any web application that has them.
First, when you click the close button, it has to pop up a dialog box and ask, "Document has been changed. Would you like to save your changes now?" The possible responses must include the standard choices of yes, no, or cancel.
Second, when your browser crashes, it has to attempt to save the file and automatically recover it when you start up again. If it can't save it, it must have an autosave feature so that the maximum possible amount of data loss is bounded.
Web applications for word processing may be possible some day. As you point out, most of the stuff you need is there already. You do need to add some extra hooks in the browser, though. I wouldn't call those things trivial, either. HTTP is supposed to be stateless. Cookies violate that, but they are too limited to be used for word processing.
Word processing is very stateful, and to enable a web application to provide word processing will require fundamental changes to the way that web browsers operate. Perhaps the changes will be easy to implement, but they are very significant, and should not be made too lightly. The implications for security and privacy need to be examined thoroughly, at the very least.
Perhaps you are right and Google is under the impression that whatever makes money is good. A lot of people think DRM is evil, but Google does it for their book searches, and a lot of people think censorship is evil, but Google does it to please the Chinese government.
What this really says to me is that Google is looking for places to invest the money from their IPO. Presumably that means they can't think of places to spend it themselves. That would make the chance of Google having some big, revolutionary plan much less likely, since such a plan would probably consume as many resources as they could throw at it to improve the chances of success. That makes me kind of sad. Google had a chance to change the world, but it seems all they really changed is the world of Internet searches.
Slashdot Mirror
As for the OS, you are right, I didn't factor it in. Once you add $120 Canadian for an OEM copy of Windows XP you are around the same price as the Mini Mac. I do expect piracy of the operating system on systems like the whitebox AMD I compared to, but even including a legal copy of Windows, the PC at worst goes to price parity.
The size of the Mini Mac is very nice, but that isn't something that most people think about when buying a computer. The volume is probably smaller than many laptops, even if you sliced off the display and keyboard. I don't see that being a decidng factor to many people, though.
I think your points are good, but I still don't think the Mini Mac will be compelling to a large enough audience to materially affect Apple's success.
The Celeron processor is still faster than the 1.25 GHz G4 PowerPC. The integrated graphics is definitely a minus for both the whitebox AMD machine and the Dell machine. The Dell machine is comparable to the whitebox AMD machine in every category except memory, where it is better.
I consider price, peripherals and operating system to be the key points, since those are the things that will most directly affect the end user experience in the target market. Processor, video, memory, warranty, connectibility--who cares? The target market is people who want a new PC, have a few hundred dollars in their pocket, and like their ipod. The PC wins on two of the three most important points and gets a toss-up on the other. I'm sorry that I didn't make it clear in my original post that I consider some factors more important than others.
The AMD system I included is pre-built from a local whitebox vendor. You can go to their store and point to it on a sheet of paper (it's easy to find, because it's the cheapest one). They will stick the processor into the pre-assembled chassis and bring it out for you within fifteen minutes. You can also order it online and have it delivered anywhere in North America (although shipping would likely raise the cost above that of the Dell machine if you live outside of Toronto). It is a machine someone can buy on a whim.
It has a DVI video port (with VGA adapter included), and it comes with no monitor, mouse, or keyboard. Although Apple states on their website that "Mac Mini will take advantage of your two button USB mouse with scroll wheel", I think that decision is a mistake. Most PC users don't have USB peripherals, and will find the Mac Mini somewhat, err, nonfunctional upon purchase.
Hardware-wise, the specs are pretty low-end (what do you expect from a $500 Mac?). Comparing them to a local no-name store, here's what you get when comparing the cheapest options:
Price-the cheap PC is $450 Canadian, the cheap Mac is $500 US. The PC is significantly cheaper.
Processor-the Sempron 2200+ is slow, for a PC. It would quiver beside the 2.0+ GHz Powermac G5's, but it beats the crap out of the 1.25 (or even 1.4 GHz upgraded) Mac Mini. PC wins again.
Video-the Mac's discrete Radeon 9200 is an ok chip, the PC's integrated video isn't. Easy win for the Mac Mini.
Memory-they both come with 256 megs, so where's the contest? The Mac Mini is only upgradeable to 1 gigabyte. Slashdotters considering purchasing one may want to avoid it for this reason. Another victory for the PC.
Peripherals-the PC comes with a monitor, a mouse, a keyboard, and speakers. All cheap junk, but still a much better deal than the Mac, which doesn't come with any of that.
Optical drive-both come with CD burners, but the PC's doesn't read DVD's, and the Mac's does. Another point for the Mac Mini.
Cool factor-the Mac is much smaller and less of an eyesore. I don't know about volume, since neither manufacturer gives specs. This is a big plus for the Mac.
OS-Mac OS X versus no OS included. Victory for the Mac, although the PC will probably end up with Windows XP. Most users will prefer Windows since it is compatible with more programs and they are more likely to be familiar with it, although opinions can be very polarized about which is better. (=
Warranty-each offers a one year warranty. Victory for the Mac, since Apple service is much easier to get if you go outside of southern Ontario.
Connectability-the Mac offers a modem, IEEE1394 and DVI, as well as USB2 ports. The PC offers only VGA and USB2 for modern peripherals, although it has old style serial and parallel ports, and PS2 keyboard and mouse jacks that the Mac lacks. Both machines include 10/100 ethernet. I prefer the modern options, so the Mac gets a slight win.
Other specs such as hard drive and sound capabilities seem similar, so I won't use them for my conclusions.
To anyone who complains about the white box vs. brand name comparison, show me a white box Macintosh and I'll yield the point. But not much, since Dell offers a similar system (with more RAM plus Windows XP, but a Celeron processor and $499 Canadian price point).
Overall, the Mac just isn't a good deal compared to the PC. The price is significantly higher, and you still need to get a mouse and keyboard. Further, the limited memory expansion means that it isn't a good deal for hobbyists like the Slashdot crowd who would like to play with a Macintosh. I suspect sales will be decent, but that many customers will be unsatisfied when the performance is no better (or even worse) than their current PC, and they are forced to make a second trip to the store to shell out for a new mouse and keyboard. Although they may cash in on the current ipod craze, this machine is unlikely to win users for Apple in the longer term, or to stop Apple's market share from shrinking.
It's not necessarily strips that matter, but overall cost. Monitoring is very $$$ intensive now.
Continuous monitoring would be VERY nice. If not continuous monitoring, something that was convenient and cheap enough that I could test every twenty minutes or so would be great. That pretty much rules out an invasive meter; I wouldn't poke myself that often under normal circumstances. Bonus points if I can attach it to my finger or arm or whatever and have it continue to function test automatically in the night time.
A few words on insulin pumps. I don't have one, but my younger brother does. As a student, I can't afford one. Meter-pump integration would be nice, but you would obviously need reliable and regular monitoring for that. For anyone developing such a thing, a big part would be to convince insurance companies to pay for it.
Two other features that other people have mentioned that I second (or third, or so on) are alarms for bad results (or results that are headed in a bad direction, if the monitoring is regular enough to detect a trend) and easy to read displays that have some lighting. I would rather have that than ultra-long battery life, particularly if the meter can take some type of rechargeable battery (such as NiMH AA or AAA cells).
I agree UXGA or higher resolution would be nice, but you also need to consider that tablets are usually meant to be carried around, and a large and heavy device would be less convenient. I'd love to see more resolution in a small form factor, but I'd probably prefer my M200 to a 15" UXGA tablet.
First is the lack of hardware support. If you happen to have a TC1000 then this guy has the drivers for you. If you happen to have an M200 (like I do), then he doesn't have the drivers for you.
Second is the lack of handwriting recognition. That's essential for using a tablet in, you know, tablet mode. Without it, even choosing to visit www.slashdot.com is a chore, and you can forget about word processing or email in the comfortable tablet form factor while riding the bus.
Third is the lack of applications. There are a few well chosen applications that support handwriting as a first class input mechanism. When scratching and scribbling on things it is comforting to have circles and lines, and even my messy handwriting, be the same as I put them in.
It would be nice to have Linux working well on my tablet, but the tablet PC is a new hardware and software platform. Microsoft doesn't have a great and polished interface for it yet, only one that is good enough. Still, every little bit of that new platform that Microsoft and others provide for Windows XP on a tablet is a little bit that Linux doesn't have yet at all.
If you need more complicated scripting capabilities, you can use Visual Basic, Javascript, or Applescript to drive Photoshop. Look at the scripting guide that is installed with Photoshop for more details.
You could also continue to use the GIMP, of course, but don't let lack of scripting stop you from using Photoshop, if that is what you prefer. Photoshop definitely has scripting capabilities.
If you click on it and tell it to "Install..." it pops up the annoying dialog that it used to show pre-SP2, with one very nice addition: a "don't install controls from this publisher, ever" button.
On the other hand, I don't accept your claim that OO.o is Good Enough(TM). It isn't good enough for me (see the previous post for reasons why it isn't good enough), and anecdotal evidence suggests that it isn't good enough for many other people who have to do spend significant amounts of time in front of a word processor.
Don't lose graphics in imported Word documents.
When you export Word documents, they need to have file sizes that are similar to what they would have if you saved them with Word. I can't email someone back a document that has had a huge increase in file size. Word is bad enough with file sizes, but OO.o is much, much worse.
Don't crash so much. That's just annoying.
A grammar checker would be nice. Word and Wordperfect have had this for over a decade.
Faster load times would be great. Word loads in about one second on my computer; there is no excuse for OO.o taking more than ten seconds.
This is just a minor nit, but still... I use a text editor to edit text documents. OO.o shouldn't claim that its formatted word processor document is a text document.
The dialog box that asks if you are sure you want to export to a non-native file format because you might lose information should tell you what information you might lose. When I import a document, add a few sentences, then save it, I should not be seeing this nonsensical warning. In fairness, Word has this problem as well for some older formats, although not for Word 97 or later formats.
My most annoying point to me(since this one means I can't even use OO.o for documents that I distribute in pdf form only): support for using custom styles for section numbering.
Fix the last one of those and I will use OO.o again. Fix most of them and I will give it another try for regular use. Right now, though, OO.o is as useful to me as Wordperfect for the Atari ST is.
It's more an issue of OO.o's poor handling of MS Word file formats than it is of MS Word file formats themselves.
But make sure that your alternate browser it is a recent version of Firefox or Mozilla. They have responded very quickly to security issues, and are being proactive about security, much more so than the the people behind Konqueror or Opera. Also, keep your alternate browser patched just as vigilantly as you would Internet Explorer. As the popularity increases you will see more attacks against Mozilla based browsers.
I don't know what the answer to security is. I hope it isn't educating users, because that just plain doesn't work for most people. The problem is that right now there doesn't seem to be any other way.
The Mozilla team quickly corrected the issues he found with Mozilla, and started using his tool to perform basic security testing. Those bug fixes are in place for the Firefox 1.0 release, which is what I'm using on Linux.
Anyone know if the Konqueror team did the same? I've Switched to Firefox because of this issue, but I really prefer Konqueror, and if I knew that they were being proactive with security I would switch back.
RHEL ES has two versions, priced at $350 and $800, depending on the support level. W2K3 SBS (Small Business Server) is available at different prices from different vendors, but is typically around $500. All prices in US dollars. The prices are quite similar. If you need support for more than installation and basic configuration, Windows 2003 is actually cheaper.
If small businesses find Windows easier to setup and maintain, then it could be worthwhile. I'm not able to personally confirm this one way or the other, but various people I know who have configured both Linux and Windows 2003 as servers claim that Windows is easier to configure and tune for performance.
Perhaps Dell simply means that for the market they are selling into and the price they are charging, there is a better product available from Microsoft. It's hard to see how Red Hat could compete on price; they really aren't charging a huge amount. For businesses that can't afford a full time server administrator and don't have any Linux expertise, it is quite possible that Windows just plain is a better choice.
Going off topic, Red Hat's website has the Ghandi quote that Slashdot loves: "First they ignore you, then they laugh at you, then they fight you, then you win." A year ago, Microsoft was fighting Red Hat. Now they are laughing at Red Hat. Linux still has a chance, but this battle definitely isn't going the way that Red Hat planned.
I hate to say it, but with IBM preferring Novell and SLES, I think Red Hat has lost.
I'm fortunate enough to fall under a volume licensing plan for MS Office, so I got it for about CDN $90. If you can get one of the heavily discounted versions it's a worthwhile purchase, but otherwise it's too expensive for individuals who don't use it a lot.
I do have a friend who still uses Linux exclusively, and he uses OO.o to prepare presentations. He exports to pdf and uses Adobe Reader to view them. Reader has a full screen mode that is suitable for giving presentations. Go to the Window->Full Screen View menu item, or press Control-L. I doubt it handles animations or transition effects from OO.o, but if you can do without those it may be suitable.
Microsoft makes stand-alone viewers for Powerpoint (97-2003) and Word (97-2000). Microsoft also makes a browser plugin for IE that understands WordML and will correctly display Word 2003 documents that are saved in XML format. The old viewer will show Word XP and non-XML Word 2003 documents, however some features may be unavailable, so the document may not look exactly right. You can fit any, or all, of the viewers on your 16 megabyte flash drive.
If alternate formats are an option, you can also convince a host of programs to read RTF. It's not as versatile as Word native format, but chances are any computer with an OS less than ten years old can read it.
OpenOffice.org is something that you suffer through when forced to use Linux to read or prepare editable documents for distribution. It's definitely not something you should ever force on anyone else.
P.S. When I used Linux exclusively, I distributed KOffice presentations by exporting to HTML format. Unfortunately KOffice does (or at least it did; it may be fixed by now) that by converting the presentations into graphics files, so it is resolution dependent. If you know what display resolution the other computer has, that will work. A better solution is probably to distribute presentations as PDF format, and use Adobe Reader as the small viewer program. I believe OpenOffice.org supports saving to PDF directly, and if not you can get a PDF printer driver.
To me it sounds like he's an amateur photographer who is just starting to experiment with digital. It's quite possible that the mediocre quality of The GIMP would suffice for him. Perhaps all he knows about colour management is "use Velvia". Further, it seems that his scanner is working with Linux.
It actually sounds like his hardware is the problem. He's got a cheap scanner with a slide adapter, and it gives him blurry results. If the results are blurry, you should try software first. Check that you are scanning at realistic resolution. If the scanner resolution is too high, drop the scanner resolution or downsample. Using that scanner with slides, that won't be a problem. If the scanner resolution is ok and the results are a bit soft, an unsharp mask should fix them. You can do that with The GIMP just as well as you can with Photoshop.
More likely he needs to invest in some decent hardware in order to make the setup work to his satisfaction. Windows or a Mac might be necessary, but solve the first problem first. If he isn't a pro, he might be able to use Linux for this.
Actually looking at the license, I see that it is based on the Mozilla license (MPL), which addresses the two issues I noted. Sun's changes remove the part about being covered by a future version of the license, and remove some notice requirements and clarify a few things that are unclear or poorly stated in the original MPL.
The license may well be GPL 3 compatible, since Stallman has made noises about wanting to clear up the patent protection stuff. You'd really need to get a lawyer's opinon on that, though, after the GPL 3 has been released.
The price fixing that Rambus alleged was a conspiracy to charge much more for Rambus memory than a competitive market would. Since the same group of companies controlled the supply of both types of memory, they could do this. Rambus claimed that Rambus memory should be about ten percent (IIRC) more expensive than SDR DRAM due to higher manufacturing costs and Rambus IP licensing. The DRAM manufacturers sold it for twice the price of SDR DRAM. If Rambus is correct about the costs, then their claim that the high price of Rambus memory was due to price fixing is plausible.
Rambus said that Rambus memory should have been priced only slightly higher than SDR SDRAM, based on manufacturing and licensing costs, and that they were always clear with JEDEC on what patents they held and what licensing terms they wanted. Rambus said two things in particular: first, that the high cost of Rambus memory was due to a conspiracy between the DRAM manufacturers, and second, that the claim of submarine patents was a smear campaign by the DRAM manufacturers.
It seems from this guilty plea that there is enough evidence to prove at least the first of those claims in court. I personally don't have any special insight into who is right, but I do wonder if we were all wrong about Rambus.
You don't patent software, so even though the term "software patent" is in common use, it is really quite meaningless. Instead you patent an algorithm. I am in favour of algorithms being patentable, and more specifically, I am in favour of algorithms that can be implemented in software being patentable, with software implementations covered by such patents.
Each particular implementation is also covered under copyright law. I'm not saying anything about that in this context. Copyright is a different issue that can be treated separately.
I think this clarifies what you are saying, but if what I write is still difficult to understand, I will attempt to make it clear.
Software patents are even more important than patents in other fields, due to the ease with which software techniques can be duplicated. Patents are absolutely necessary to protect small companies from having their ideas taken without any credit or compensation to the original source.
Much of the criticism against patents that has been leveled on this website is also driven by ignorance. People do not realize how specific patents are. I have seen posts on many patent articles here that read the first one or two claims and assume that a huge range of existing work is covered, without checking out the remaining claims that make it clear that one very small thing that is original is the actual target of the patent.
I won't defend the existing patent system too much, since it is flawed with respect to software. Since software patents are easy to implement, the costs required to develop them can be recuperated much faster than other areas, so a shorter duration of protection would probably be better. Some patents have been applied overly broadly, or granted when they were not merited.
But the abuses do not stop the patent system from being useful for software. The problems are things that can be worked out, not fundamental flaws with the idea of patenting software algorithms.
First, when you click the close button, it has to pop up a dialog box and ask, "Document has been changed. Would you like to save your changes now?" The possible responses must include the standard choices of yes, no, or cancel.
Second, when your browser crashes, it has to attempt to save the file and automatically recover it when you start up again. If it can't save it, it must have an autosave feature so that the maximum possible amount of data loss is bounded.
Web applications for word processing may be possible some day. As you point out, most of the stuff you need is there already. You do need to add some extra hooks in the browser, though. I wouldn't call those things trivial, either. HTTP is supposed to be stateless. Cookies violate that, but they are too limited to be used for word processing.
Word processing is very stateful, and to enable a web application to provide word processing will require fundamental changes to the way that web browsers operate. Perhaps the changes will be easy to implement, but they are very significant, and should not be made too lightly. The implications for security and privacy need to be examined thoroughly, at the very least.
What this really says to me is that Google is looking for places to invest the money from their IPO. Presumably that means they can't think of places to spend it themselves. That would make the chance of Google having some big, revolutionary plan much less likely, since such a plan would probably consume as many resources as they could throw at it to improve the chances of success. That makes me kind of sad. Google had a chance to change the world, but it seems all they really changed is the world of Internet searches.