Re:Medical Applications a long way off
on
Frankenstein Time
·
· Score: 1
This is completely right, all we have done is to write down 90% of the genome, we have not even begun to be able to understand what the words mean. This is sort of like sitting through a latin mass and writting down what you hear. The text is nice, but we don't speak latin, we still have a long way untill we can understand a few words, let alone be able to construct new sentances.
In order to be able fix genetic diseases you have to be able to recognise basic gramatical/spelling mistakes. For those of you who don't speak foreign languages, this is not trivial. Selecting for certain traits is moderately easier since (with a lot of work, and only crudely) we can compare what we want and wat we don't, but since we don't understand punctuation yet, still a daunting task (right now we can only recognize paragrahs and chapters).
Actually going in and creating new structures in DNA is a long way off, for that we would not only have to understand the grammer of the language, but also have sence of style (in a writting sense) and an understanding of the subject matter. And trying for something as abstract as inteligence, or creativity, would require a master poet in this language. Remember, the supermice we have created in the lab only have a spike in a certain brain chemical which was already there, we just spiked it some. We still don't really understand why it works, and that was already in the code, but presumably smarter mice are not the much more survivable in the wild, as this gene has not come to the forefront in evolution.
Now there is one thing that does scare me. It is a tough thing to be able to create in a new language, it is another to destroy. By making the contents of the genome public we are giving DNA weapons makers a blueprint on which to build custom weapons. My prediction is that this area will be the first to "benifit" from this research.
This is in responce to the side question, yes there are a few of ways of roaming on the Macintosh. The least-usefull but go-anywhere version is to get youself a 10 Mb slice of server space at mac.com (Apple provides this for free), and mount that wherever you go (requires MacOS 9+). This will give you file roaming. If you have a Mac (OS 9) on the net, you can mount it remotely if you enable TCP/IP filesharing in that control pannel. If you have an AppleShareIP 6.3 server, this drive slice can also be avalible under FTP and SMB.
The next solution, Macintosh Manager, is nice in an office solution, and can be used with client computers running MacOS 8+ (possibly 7.5+, but I would have to look that up). You just need either a MacOS X Server box or a AppleShareIP 6.3 box with the Macintosh Manager software installed. With this software the client computer partially boots and then asks you to log in. It authenticates off the server, and then loads prefernces and other files off the server (can take some time). When you log out, it can then save the changed files back to the server. All of this is configurable by a remote admin control. This does require that client software be loaded on all the computers (on OS 9 it is a custome install option).
The third option, Netboot, requires a MacOS X Server, and it's clients must be "flavored" Macs, that is anything newer than the Blue&White G3's. If you hold down the "b" key, or choose to netboot in the Statup control panel, then the computer will search the ntwork for an avalible netboot server. If it finds one fo wich it is authorized, it boots partially off the server, and then asks the user for a name/pass. Once you have authenticated to the right server, it lets you boot all the way with your assigned disk image (stored on the server all the time), which can be read-only or read-write. In this configuration you can actually remove the HD from the client coputer completely (on the new iMacs this makes a really quiet computer). The disk images can be combined, so you cna have a system disk for each individual user, or by groups, and then have a common image for all the Apps, a very flexible solution. But it really needs to be on full duplex, switched 100baseT for good performance.
Apple is definaltely looking at taking all three of these aproaches an work them into solutions for MacOS X (Server version). Hints can already be seen by looking at the/Local/Network, etc directories that Apple has keep from the NeXT system.
Generally I do agree with you that the Stepwise group is going a littel overboard, but they do have a point with the portables. MacOS X Server has two main uses, one as a workgroup server, and one as a WebObjets development box.
It is this second instance where they have a good point. If I, as a WebObjects developer were to need a new laptop for on-site demonstrations of prototypes (WebObjects is great for mocking up a rapid prototype that works), then I would have no options from Apple. And it looks like this is how things are going to be untill MacOS X and WebObjects 5.0 (for Java) come out, and then that solution will orphan almost all of the WebObjects 4.x work wich is done in WebScript or Objective-C. That is unless I want to buy a NT laptop, wich will become a second-class WebObjects citizen when MacOS X spins up... not a great position to put a solid group of developers....
except the user only sees the file as it is presented to him/her in the finder (or standard open dialouge). So the file always looks the same to the user. The only people who are subjected to this schiofrenia (sp?) are the developers who work accross different programming models. As a Perl coder who has to deal with the file system differennces in Mac/Unix/Win I will tell you that things can get a little funny at times, and only having a swictch between two metacharicters is trivial. And as a MacOS X user, it is implimeed so well that I sometimes forget that it is there at all... it is just transparent. This seems to be bet best weld of the idosycratocies of the two systems.
umm... beyond the extreme speculatatory nature of the post, there are a few problems with this post:
1) CP/M and DOS were only two of the competing OS's on the intel hardware at that point, basicly who-ever IBM chose to liscence to attach to their mainframes was goign to win. DOS did not win because it was technicaly supperior to its compeditors, but because IBM chose it.
2) Your point on SCSI is more than a littel misleading. In the days of SCSI 1 there was no real standard. Every SCSI product was engineered for the server that it was made for. Apple came out with products that worked with it's computers (duhh), and not everyone elses hardware worked. Apple was not differnt in this, they were just the only one in the comsumer space using SCSI, so took the heat. When SCSI 1 finally came together (with an industry consortion includig Apple), Apple was right there (in fact their old devices were complient to the new standard).
and 3) There were more cloners than you think, just in the US: Motorola (StarMax), PowerComputing (PowerTower, PowerBase), UMAX, Mactell (sub-licencee of UMAX), IBM (never brought out a product), etc... And there were a number of licencees in Asia.... the redux on that who saga is that no-one was making money, and it was killing Apple, so they made the right buisness decision, and stopped. Remember, IBM has been loosing money on their personal PC buisness for a long time, and been using it as a loss-leader for their Buisnes Services and Big-Iron divisions. Apple has never had these divsions to fall back on (although we will see how far iServices goes).
According to a factoid that ran on CNN today, 50% of American have access to a computer, as of 1999. I think this one shoots down the rest f the numbers....
Maybe because Apple does not (and does not want to) compete in the Enterprise Class Server Market. Applt.com gets a enourmous load on it all the time, desktop machines are simply not up to the task. I run a coupple of MacOS X Server boxes, and they are wonderfull for the workgroup serving that they do, and the small web sites that they host, but would die under the loads above 1000 hits a minute. Considering that Apple.com is probably getting 10,000 hits a second, this is not even an option.
The simple answer to this is that if you are going to do a lot of work with regular expresions, or text manipulation at all, do the work in perl. If you are going to be sending email to people, or sending events/scripts to programs running on the server (or other servers using AppleScriptIP - used in image shops), then use AppleScript.
Or, if you have to do both, use the MacPerl AppleScript extensions to make a perl script that generates OSAX evnets, and have the best of both worlds! This can be a littel slow, but if you are doin this sort of thing, you are probably not concerned with speed too much, doing things like remote image manipulation, etc...
The combination of the two methods is really a great feature in MacPerl!
Yes, this is very true! In fact the laptop computers that are regularly carried into space are far more powerfull than the shuttles main computers. And many of the shutle research missions could not be accomplished without them. But they are used for payload missions only! Nothing mission critical!
For the actuall running of the shuttle, the main computers, working in concert with the big computers a mission control, do the job every time. They are predicatable! No one dies because an obscuer error jump up at the wrong moment.
There are a lot of OS's that are true real-time OS's, but they are ones that you never directly interact with. A lot of the phone switches run them, your car's engine may run on another, they are present in a lot of medical equipment, they run most of the heavy industrial equipment, are the driving force in most aircraft avionics, and they are staring to make their way into things like digital televisions.
Each RTOS is fairly stupid, they do one things really, really well, and basicaly nothing else. Because the goals in mind are so limited they can be so small that one person can understand everything about them, and predict how they will react in any circumstance (devide by zero is the most common example). They usually have little to no user interface, or have one that is contained in a program that runs on top of them, that way if something goes wrong, it can be restarted without interfering with the main job that the OS is built to do, whatever that may be.
To make a usefull comparison, geralized OS's (such as Win9x, WinNT, Linux, Digital UNIX, MacOS, etc..) are like the CPU's they run on; they are big, complex, and can do almost anything, but are great at nothing. In comparison RTOS's are like math co-prosessors, DSP's, or graphics chips; they do one thing really well, much better than their generalizd brethern, but are incapable of doing other things.
9.0.4 and DVD Player 2.2 have solved all the issues I have seen, on iMacs and G4's.. you might want to make sure you have both... (neither is yet in the software update... *sigh*)
Well, sounds like you have a solution... as your needs are really small, and you are comming from analouge, probably the best solution is to use whatever came with your digitising card (sounds like videoWave III ?). If you want more flexibility to cut and paste, and work without paying attention to media types, I like QuickTime Pro, at under $30 it is well worth it for either Mac or PC... and you get a version of the Sorenson compresser which will probably work better than MPEG-1 for most things.
For transitions, etc... sounds like you are right in the iMovie range, but since you seem to be on the Wintel side already, and as such it woudl cost you more to switch platforms than to buy even mid-rnge editing packages... I woudl simply wait for Microsoft to finnish up with their forthcomming responce to iMovie (the name of the project escapes me..), it will probably fill the needs you have, and I seem to remember that it was intended to e bundled with WindowsME (this was before it was called that, so things may have changed)...
The real problem here is that the posters have never defined what they are doing, and what they want. Are you looking ot do ollywood post-production work? Are you doing a little simple video editing (splicing, soundtrack, and transitions), or are you goign to be creating content on the compouter to interact/overlay on DV video? are you going back to the DV cammera with the work, to a DV storrage array over fiberchannel, making a video-Cd, or goign to be publishing on the web? Who are your customers, and what quality do you expect?
For most people the answer is going to be the simple home movie, publishing back to the cammera, or mabey to the web for short videos. In both cases my recomendation is an iMac DV or Special Edition, mabey replacing the Hd with a larger one. This gives you a quality computer with enough horsepower to work on video, built-in OS supported FireWire (1384), and a great consumer level editing package (iMovie.. i have given 2 hour courses on its use, and it is simple to learn, and very powerful for the usual stuff, I highly recomend it for most uses). I would recomend having someone demo iMovie for you once, as it is a great piece of comsumer level software! the size of te Hd is a consideration, but not as big a one as you might imagine, as most of the time you are not woring on more than 20 minutes of video, and you just toss stuff back onto the cammera when you are done, if you are going to be doing hours of video at a time, get a profesional system...
On the next level of stuff.. TV broadcast quality work, I would recomend a G4, a Cannon XL-1, and either Apple's Final Cut Pro (my personal recomendation), or Adobe Premere, and a copy of Media Cleaner Pro, oh and a copy of Apple's QuickTime Pro (the best value tool you will find out there!). If you really feel that portablility is important (say to cut together a news clip while you are russing back to the station in the van), the new FireWire powerbook can serve as a nice little mobile station. Incidentally, this seems to be the combination that ABC has chosen to send it's teams into the field with.
And the final level that I am going to be talking about, the high end content creation level: Here I would go with a a G4 or a SGI (depending on what your company is better at supporting) decked out with a Arora or Avid card (top end is $10,000+ a card), a Gig-and-a-half of memor (remebering of course to get 2-2-2 memory), a fibrecannel Gigabit interface, Dual Channel SCSI-160 (one for scratch, 10K RPM of course), and maybe one of those nice quad-processor digital co-processor cards (4 G3's on a card... not real multi-processing, but even better fo DSP stuff...).
Hope this helps someone out there, and if anyone is in the Madison, Wi area, I am more than willing to do short demo's of the lower end products mentioned (I am not a salesman...). I do have a pro-mac bais, and in this case that is really where the professionals are going, so the bais is justified.
A browser: try OmniWeb 4 from http://www.omnigroup.com
As to Applications, what do you want? If you can't find it in a Cocco, Carbon, or BSD version then just fire up te Classic environment, and go to town on the vast array of traditional MacOS Apps...
What exactly do you mean by "Sufficient documentation to make our own codecs"... do you popose Apple give you a Phd in recominitoric math? Modern Codecs are very complicated algorithms that take a lot of tallent to write. This is not the sort of thing that you just bang out in your spare time whever you feel like it. We are talking about very complicated processes that few people can comprehend. This is why people make money off of codecs, and why it is unlikely that anyone is goign to be making a open-source codec that beat anything by Sorenson (remembe this is not Appl's product, but one that is liscenced), or QualCom, or any of the other producers out there... You are asking people to give their life's work for free, so that you can watch movies...
No, it also does not include the Cocca, MacOS Classic, and Carbon API's or the Quarz display environment that Aqua is built on. These are Apple's family jewels, where they are going to make their money. I know that many on/. hate the idea of a company actually making money on their own work, but this is the real world...
Yep, you got it right, except there is one more API, the "Clasic" API, where MacOS 7, 8, and 9 (and previos ones too.. more or less), can run in an environment that is MacOS 9 (you install it just like you would normally, it has a system enabler to change a few of the behaviors).
The other minor point is that it is fairly trivial to port command line things, and once the port of X-Windows is done (third party efforts), then much of the rest can start to come over too...
ummm.. KNI is actually two 64 bit functioning units operating in parallel. Intell had to do this so that they kept compatability with MMX instruction ergisters... so calling it 128 bit is like calling a dual processor PentiumPro machine a 64bit architeture....
Umm... I think you mistunderstand things a little...the G3 PowerBooks may not be the Mhz kings anymore, but they are still the fastest/among the fastest portable computers out there, and with the rumors of a impending speed bump/upgrade to the PowerBook line, Apple might jut be moving back into that position, now that Motorolla has been talking about 500-600 Mhz G3's.. one has to wonder..
<p>And then when you consider that Transmeda has not been making claims that they would be competing in the speed areana, but instead makeing reasonably fast processors at low power settings, thi does not sound like a product where Apple would be interested in for their Pro lines, and since they are now beginnning to see major cost saving s from the Unified Motherboard Architcture (UMA), suddenly this sounds like a Really Bad Idea(tm)...</p>
Re:Why not just use the Crusoe as a G4?
on
Darwin on Crusoe?
·
· Score: 1
RISC = Rationalized Instruction Set Computer, meaning the instructions have been each been simplified so that they all are they same size and all take the same amount of time to run. RISC computers typically have more instructions that a comparable CISC computer. The reason for this can be seen in an example:
<p>In the example of adding two numbers together a CISc computer may have one instruction to add two numbers together, whether thay are integers, floats, etc.. where as a RISC computer must have a single instruction for each eventuallity. The idea is to take complexity out of the chip and put it into the compiler (which already had to know what type of numbers were involved in the add anyways...).</p> <p>Transmedia has decided to take this idea one step further, and ime will tell if they are sucsessfull in it. In any event, you can bet that this general idea will become a staple of processor/computer design in the next ten years</p>
I agree with this in principal, but how many consumers are going to follow htis reasoning? And since Apple focuses a lot of it's marketing at general consumers, these rips (one of them done a mere 4 hours after Apple unveils Aqua.. and a used to sell a comercial product) definately hurt Apple, even if it is just mindshare...
The Newton did not have graffiti, it had/(has?) full handwriting recognition, you simply wrote like normal, and the PDA translated it into ASCII... The first version made a lot of mistakes, but the final one (on the 2100), was really very impresive. I have used it a few times, and even in normal mode (it could also learn your individual take on hadwriting), it worked perfectly with my script writing, and almost perfectly (fewer mistakes to correct then I have typing this reply) with my cursive.
And since Apple did all this work, they have all the patents on how it was done... since they are rumord to be in negotiations for co-branding with Palm right now, maybe we will see the Neton technology back out in the field again (I hope!).
We are currently waiting for Apple to release their new Driver SDK, which will probably be very different than anything else out there (that is what Apple has been hinting at for a while, dynamic loading/unloading are almost a shoe in.. and for everyting...)
Sure, if you want to port Quartz (the imaging system... think of it as X11 on lots of Seroids) to your LinuxPPC, box.. Or wait for Apple to work out a way of getting the ClosedSource third party codecs to open up so that you can compile it for your hardware/setup... If you really want this, talk to Sorenson, or any of the other Codec patent holders, and tell them that you want this, then Apple will have the ability ot do this!
Slashdot Mirror
This is completely right, all we have done is to write down 90% of the genome, we have not even begun to be able to understand what the words mean. This is sort of like sitting through a latin mass and writting down what you hear. The text is nice, but we don't speak latin, we still have a long way untill we can understand a few words, let alone be able to construct new sentances.
In order to be able fix genetic diseases you have to be able to recognise basic gramatical/spelling mistakes. For those of you who don't speak foreign languages, this is not trivial. Selecting for certain traits is moderately easier since (with a lot of work, and only crudely) we can compare what we want and wat we don't, but since we don't understand punctuation yet, still a daunting task (right now we can only recognize paragrahs and chapters).
Actually going in and creating new structures in DNA is a long way off, for that we would not only have to understand the grammer of the language, but also have sence of style (in a writting sense) and an understanding of the subject matter. And trying for something as abstract as inteligence, or creativity, would require a master poet in this language. Remember, the supermice we have created in the lab only have a spike in a certain brain chemical which was already there, we just spiked it some. We still don't really understand why it works, and that was already in the code, but presumably smarter mice are not the much more survivable in the wild, as this gene has not come to the forefront in evolution.
Now there is one thing that does scare me. It is a tough thing to be able to create in a new language, it is another to destroy. By making the contents of the genome public we are giving DNA weapons makers a blueprint on which to build custom weapons. My prediction is that this area will be the first to "benifit" from this research.
This is in responce to the side question, yes there are a few of ways of roaming on the Macintosh. The least-usefull but go-anywhere version is to get youself a 10 Mb slice of server space at mac.com (Apple provides this for free), and mount that wherever you go (requires MacOS 9+). This will give you file roaming. If you have a Mac (OS 9) on the net, you can mount it remotely if you enable TCP/IP filesharing in that control pannel. If you have an AppleShareIP 6.3 server, this drive slice can also be avalible under FTP and SMB.
/Local /Network, etc directories that Apple has keep from the NeXT system.
The next solution, Macintosh Manager, is nice in an office solution, and can be used with client computers running MacOS 8+ (possibly 7.5+, but I would have to look that up). You just need either a MacOS X Server box or a AppleShareIP 6.3 box with the Macintosh Manager software installed. With this software the client computer partially boots and then asks you to log in. It authenticates off the server, and then loads prefernces and other files off the server (can take some time). When you log out, it can then save the changed files back to the server. All of this is configurable by a remote admin control. This does require that client software be loaded on all the computers (on OS 9 it is a custome install option).
The third option, Netboot, requires a MacOS X Server, and it's clients must be "flavored" Macs, that is anything newer than the Blue&White G3's. If you hold down the "b" key, or choose to netboot in the Statup control panel, then the computer will search the ntwork for an avalible netboot server. If it finds one fo wich it is authorized, it boots partially off the server, and then asks the user for a name/pass. Once you have authenticated to the right server, it lets you boot all the way with your assigned disk image (stored on the server all the time), which can be read-only or read-write. In this configuration you can actually remove the HD from the client coputer completely (on the new iMacs this makes a really quiet computer). The disk images can be combined, so you cna have a system disk for each individual user, or by groups, and then have a common image for all the Apps, a very flexible solution. But it really needs to be on full duplex, switched 100baseT for good performance.
Apple is definaltely looking at taking all three of these aproaches an work them into solutions for MacOS X (Server version). Hints can already be seen by looking at the
Generally I do agree with you that the Stepwise group is going a littel overboard, but they do have a point with the portables. MacOS X Server has two main uses, one as a workgroup server, and one as a WebObjets development box.
It is this second instance where they have a good point. If I, as a WebObjects developer were to need a new laptop for on-site demonstrations of prototypes (WebObjects is great for mocking up a rapid prototype that works), then I would have no options from Apple. And it looks like this is how things are going to be untill MacOS X and WebObjects 5.0 (for Java) come out, and then that solution will orphan almost all of the WebObjects 4.x work wich is done in WebScript or Objective-C. That is unless I want to buy a NT laptop, wich will become a second-class WebObjects citizen when MacOS X spins up... not a great position to put a solid group of developers....
except the user only sees the file as it is presented to him/her in the finder (or standard open dialouge). So the file always looks the same to the user. The only people who are subjected to this schiofrenia (sp?) are the developers who work accross different programming models. As a Perl coder who has to deal with the file system differennces in Mac/Unix/Win I will tell you that things can get a little funny at times, and only having a swictch between two metacharicters is trivial. And as a MacOS X user, it is implimeed so well that I sometimes forget that it is there at all... it is just transparent. This seems to be bet best weld of the idosycratocies of the two systems.
umm... beyond the extreme speculatatory nature of the post, there are a few problems with this post:
1) CP/M and DOS were only two of the competing OS's on the intel hardware at that point, basicly who-ever IBM chose to liscence to attach to their mainframes was goign to win. DOS did not win because it was technicaly supperior to its compeditors, but because IBM chose it.
2) Your point on SCSI is more than a littel misleading. In the days of SCSI 1 there was no real standard. Every SCSI product was engineered for the server that it was made for. Apple came out with products that worked with it's computers (duhh), and not everyone elses hardware worked. Apple was not differnt in this, they were just the only one in the comsumer space using SCSI, so took the heat. When SCSI 1 finally came together (with an industry consortion includig Apple), Apple was right there (in fact their old devices were complient to the new standard).
and 3) There were more cloners than you think, just in the US: Motorola (StarMax), PowerComputing (PowerTower, PowerBase), UMAX, Mactell (sub-licencee of UMAX), IBM (never brought out a product), etc... And there were a number of licencees in Asia.... the redux on that who saga is that no-one was making money, and it was killing Apple, so they made the right buisness decision, and stopped. Remember, IBM has been loosing money on their personal PC buisness for a long time, and been using it as a loss-leader for their Buisnes Services and Big-Iron divisions. Apple has never had these divsions to fall back on (although we will see how far iServices goes).
According to a factoid that ran on CNN today, 50% of American have access to a computer, as of 1999. I think this one shoots down the rest f the numbers....
Maybe because Apple does not (and does not want to) compete in the Enterprise Class Server Market. Applt.com gets a enourmous load on it all the time, desktop machines are simply not up to the task. I run a coupple of MacOS X Server boxes, and they are wonderfull for the workgroup serving that they do, and the small web sites that they host, but would die under the loads above 1000 hits a minute. Considering that Apple.com is probably getting 10,000 hits a second, this is not even an option.
The simple answer to this is that if you are going to do a lot of work with regular expresions, or text manipulation at all, do the work in perl. If you are going to be sending email to people, or sending events/scripts to programs running on the server (or other servers using AppleScriptIP - used in image shops), then use AppleScript.
Or, if you have to do both, use the MacPerl AppleScript extensions to make a perl script that generates OSAX evnets, and have the best of both worlds! This can be a littel slow, but if you are doin this sort of thing, you are probably not concerned with speed too much, doing things like remote image manipulation, etc...
The combination of the two methods is really a great feature in MacPerl!
Yes, this is very true! In fact the laptop computers that are regularly carried into space are far more powerfull than the shuttles main computers. And many of the shutle research missions could not be accomplished without them. But they are used for payload missions only! Nothing mission critical!
For the actuall running of the shuttle, the main computers, working in concert with the big computers a mission control, do the job every time. They are predicatable! No one dies because an obscuer error jump up at the wrong moment.
There are a lot of OS's that are true real-time OS's, but they are ones that you never directly interact with. A lot of the phone switches run them, your car's engine may run on another, they are present in a lot of medical equipment, they run most of the heavy industrial equipment, are the driving force in most aircraft avionics, and they are staring to make their way into things like digital televisions.
Each RTOS is fairly stupid, they do one things really, really well, and basicaly nothing else. Because the goals in mind are so limited they can be so small that one person can understand everything about them, and predict how they will react in any circumstance (devide by zero is the most common example). They usually have little to no user interface, or have one that is contained in a program that runs on top of them, that way if something goes wrong, it can be restarted without interfering with the main job that the OS is built to do, whatever that may be.
To make a usefull comparison, geralized OS's (such as Win9x, WinNT, Linux, Digital UNIX, MacOS, etc..) are like the CPU's they run on; they are big, complex, and can do almost anything, but are great at nothing. In comparison RTOS's are like math co-prosessors, DSP's, or graphics chips; they do one thing really well, much better than their generalizd brethern, but are incapable of doing other things.
9.0.4 and DVD Player 2.2 have solved all the issues I have seen, on iMacs and G4's.. you might want to make sure you have both... (neither is yet in the software update... *sigh*)
Well, sounds like you have a solution... as your needs are really small, and you are comming from analouge, probably the best solution is to use whatever came with your digitising card (sounds like videoWave III ?). If you want more flexibility to cut and paste, and work without paying attention to media types, I like QuickTime Pro, at under $30 it is well worth it for either Mac or PC... and you get a version of the Sorenson compresser which will probably work better than MPEG-1 for most things.
For transitions, etc... sounds like you are right in the iMovie range, but since you seem to be on the Wintel side already, and as such it woudl cost you more to switch platforms than to buy even mid-rnge editing packages... I woudl simply wait for Microsoft to finnish up with their forthcomming responce to iMovie (the name of the project escapes me..), it will probably fill the needs you have, and I seem to remember that it was intended to e bundled with WindowsME (this was before it was called that, so things may have changed)...
The real problem here is that the posters have never defined what they are doing, and what they want. Are you looking ot do ollywood post-production work? Are you doing a little simple video editing (splicing, soundtrack, and transitions), or are you goign to be creating content on the compouter to interact/overlay on DV video? are you going back to the DV cammera with the work, to a DV storrage array over fiberchannel, making a video-Cd, or goign to be publishing on the web? Who are your customers, and what quality do you expect?
For most people the answer is going to be the simple home movie, publishing back to the cammera, or mabey to the web for short videos. In both cases my recomendation is an iMac DV or Special Edition, mabey replacing the Hd with a larger one. This gives you a quality computer with enough horsepower to work on video, built-in OS supported FireWire (1384), and a great consumer level editing package (iMovie.. i have given 2 hour courses on its use, and it is simple to learn, and very powerful for the usual stuff, I highly recomend it for most uses). I would recomend having someone demo iMovie for you once, as it is a great piece of comsumer level software! the size of te Hd is a consideration, but not as big a one as you might imagine, as most of the time you are not woring on more than 20 minutes of video, and you just toss stuff back onto the cammera when you are done, if you are going to be doing hours of video at a time, get a profesional system...
On the next level of stuff.. TV broadcast quality work, I would recomend a G4, a Cannon XL-1, and either Apple's Final Cut Pro (my personal recomendation), or Adobe Premere, and a copy of Media Cleaner Pro, oh and a copy of Apple's QuickTime Pro (the best value tool you will find out there!). If you really feel that portablility is important (say to cut together a news clip while you are russing back to the station in the van), the new FireWire powerbook can serve as a nice little mobile station. Incidentally, this seems to be the combination that ABC has chosen to send it's teams into the field with.
And the final level that I am going to be talking about, the high end content creation level: Here I would go with a a G4 or a SGI (depending on what your company is better at supporting) decked out with a Arora or Avid card (top end is $10,000+ a card), a Gig-and-a-half of memor (remebering of course to get 2-2-2 memory), a fibrecannel Gigabit interface, Dual Channel SCSI-160 (one for scratch, 10K RPM of course), and maybe one of those nice quad-processor digital co-processor cards (4 G3's on a card... not real multi-processing, but even better fo DSP stuff...).
Hope this helps someone out there, and if anyone is in the Madison, Wi area, I am more than willing to do short demo's of the lower end products mentioned (I am not a salesman...). I do have a pro-mac bais, and in this case that is really where the professionals are going, so the bais is justified.
A browser: try OmniWeb 4 from http://www.omnigroup.com
As to Applications, what do you want? If you can't find it in a Cocco, Carbon, or BSD version then just fire up te Classic environment, and go to town on the vast array of traditional MacOS Apps...
What exactly do you mean by "Sufficient documentation to make our own codecs"... do you popose Apple give you a Phd in recominitoric math? Modern Codecs are very complicated algorithms that take a lot of tallent to write. This is not the sort of thing that you just bang out in your spare time whever you feel like it. We are talking about very complicated processes that few people can comprehend. This is why people make money off of codecs, and why it is unlikely that anyone is goign to be making a open-source codec that beat anything by Sorenson (remembe this is not Appl's product, but one that is liscenced), or QualCom, or any of the other producers out there... You are asking people to give their life's work for free, so that you can watch movies...
No, it also does not include the Cocca, MacOS Classic, and Carbon API's or the Quarz display environment that Aqua is built on. These are Apple's family jewels, where they are going to make their money. I know that many on /. hate the idea of a company actually making money on their own work, but this is the real world...
Yep, you got it right, except there is one more API, the "Clasic" API, where MacOS 7, 8, and 9 (and previos ones too.. more or less), can run in an environment that is MacOS 9 (you install it just like you would normally, it has a system enabler to change a few of the behaviors).
The other minor point is that it is fairly trivial to port command line things, and once the port of X-Windows is done (third party efforts), then much of the rest can start to come over too...
ummm.. KNI is actually two 64 bit functioning units operating in parallel. Intell had to do this so that they kept compatability with MMX instruction ergisters... so calling it 128 bit is like calling a dual processor PentiumPro machine a 64bit architeture....
Umm... I think you mistunderstand things a little...the G3 PowerBooks may not be the Mhz kings anymore, but they are still the fastest/among the fastest portable computers out there, and with the rumors of a impending speed bump/upgrade to the PowerBook line, Apple might jut be moving back into that position, now that Motorolla has been talking about 500-600 Mhz G3's.. one has to wonder..
<p>And then when you consider that Transmeda has not been making claims that they would be competing in the speed areana, but instead makeing reasonably fast processors at low power settings, thi does not sound like a product where Apple would be interested in for their Pro lines, and since they are now beginnning to see major cost saving s from the Unified Motherboard Architcture (UMA), suddenly this sounds like a Really Bad Idea(tm)...</p>
RISC = Rationalized Instruction Set Computer, meaning the instructions have been each been simplified so that they all are they same size and all take the same amount of time to run. RISC computers typically have more instructions that a comparable CISC computer. The reason for this can be seen in an example:
<p>In the example of adding two numbers together a CISc computer may have one instruction to add two numbers together, whether thay are integers, floats, etc.. where as a RISC computer must have a single instruction for each eventuallity. The idea is to take complexity out of the chip and put it into the compiler (which already had to know what type of numbers were involved in the add anyways...).</p>
<p>Transmedia has decided to take this idea one step further, and ime will tell if they are sucsessfull in it. In any event, you can bet that this general idea will become a staple of processor/computer design in the next ten years</p>
I agree with this in principal, but how many consumers are going to follow htis reasoning? And since Apple focuses a lot of it's marketing at general consumers, these rips (one of them done a mere 4 hours after Apple unveils Aqua.. and a used to sell a comercial product) definately hurt Apple, even if it is just mindshare...
The Newton did not have graffiti, it had/(has?) full handwriting recognition, you simply wrote like normal, and the PDA translated it into ASCII... The first version made a lot of mistakes, but the final one (on the 2100), was really very impresive. I have used it a few times, and even in normal mode (it could also learn your individual take on hadwriting), it worked perfectly with my script writing, and almost perfectly (fewer mistakes to correct then I have typing this reply) with my cursive.
And since Apple did all this work, they have all the patents on how it was done... since they are rumord to be in negotiations for co-branding with Palm right now, maybe we will see the Neton technology back out in the field again (I hope!).
We are currently waiting for Apple to release their new Driver SDK, which will probably be very different than anything else out there (that is what Apple has been hinting at for a while, dynamic loading/unloading are almost a shoe in.. and for everyting...)
Sure, if you want to port Quartz (the imaging system... think of it as X11 on lots of Seroids) to your LinuxPPC, box.. Or wait for Apple to work out a way of getting the ClosedSource third party codecs to open up so that you can compile it for your hardware/setup... If you really want this, talk to Sorenson, or any of the other Codec patent holders, and tell them that you want this, then Apple will have the ability ot do this!
Actually, if you get a Voodoo 1, there is still that Glide Driver floating arround, and the copy of Quake... :)