I can't see what not watching TV has to do with ignorance - TV is about entertainment for the purpose of selling advertising or subscriptions. Period. You don't really get much 'news' in 20 second soundbites - I think 'infotainment' is the best description I've seen of TV news. It scares me that someone might think that they actually know what is 'going on in the society' by watching TV.
Newspapers and National Radio (in New Zealand) are at least slightly better, and websites/email lists can be great for news related to paricular issues .
Oh, and having aquired a TV recently after 4 years without one, I don't watch it much, and doubt I'll be watching any of the Olympics. I'll be too busy doing interesting/useful things to 'vegetate in front of the box'.
I can well believe that Linux is poised to overtake MacOS in pure numbers of users - it runs on a _lot_ more hardware for a start.
However, this in no way condemns MacOS - there is room for both for different sorts of tasks. I use Linux at work and MacOS at home, and those are the environements that are best for me at each place.
Linux is great because it's main focus is power and configurability. However these very things detract from how easy it is to use. Great for hackers, servers, and command line stuff.
MacOS has the best interface I've ever used (I never used a NeXT box), and that is because this is the primary goal of its designers. Great for all that GUI stuff - graphics, writing, and getting not too complex things done as simply as possible (and I really like using an IDE for small to medium programming projects). There is a group of users that MacOS is not going to lose in a hurry.
I think the OSes currently have different audiences for the most part, such as people with different preferences about command-line/GUI.
Of course, I have great hopes for MacOS X combining the best of both worlds (I already run Darwin for tasks I can't use MacOS for), but it will invariably make some compromises because I don't think it is yet possible to combine the best possible power _and_ usability.
It would be really neat if Eazel does well, but that is a bit further down the track, and I don't think they have the resources to do the sort of things being done with MacOS X.
I really like a bit of diversity and choice is OSes - this helps make them all better.
Write the code using your favourite development environment, then port it to CodeWarrior (it shouldn't be that hard as long as you stick to fairly standard C++).
This gives you the advantage of using your own environment, and still meeting the requirement.
Of course if it _is_ difficult to port, then you've found out something about writing multi-platform code. C++ is notorious with differences between platforms.
Where MacOS X might have a niche as a server is for an organisation that is too small to have a UNIX Admin - I suspect that it's going to have a hugely better learning curve than UNIX, and be easier to do simple administration.
I suspect that if you've got someone who can administrate UNIX well, then you'd be better going with a UNIX or Linux solution. MacOS X (which I assume this discussion is actually about, rather than MacOS X Server which has been around for a while) looks like it's going to be a great desktop and workstation environment, but doesn't AFAIK (I am not a server expert) offer much extra particularly aimed for serving (Web Objects is one such thing?).
> My only hesitation is supporting a monopoly, be it Microsoft or Apple. Apple killed off all of their hardware competition, which was a shame, as I think some of their "competitors" were actually making better "Mac" hardware than Apple was (and at a lower price).
Yes they did (in some ways, although I don't think the quality of many of the clones was as good). That was one of the reasons that Apple (arguably) needed to kill them. The argument as I understand it was that the clonemakers were taking the fruits of Apples R&D (motherboard designs, etc), putting faster CPUs in them (which they could do because they were working with smaller volumes, and were nimbler than Apple anyway), then undercutting Apple. This wasn't particularly sustainable - Apple doing a lot of the work, and others taking the profits.
... unless you are prepared to do a proper job of the user interface.
The real advantage of MacOS is the relatively clean user interface, and that most of the applications follow the rules.
Hopefully MacOS X is going to maintain that usability (and dare I hope even improve on it?).
Now look at the X world (I use Linux at work). Many apps have different appearances because they use different windowing toolkits and many different conventions for how to do things (how many ways of scrolling or cut and paste are there? - compare Netscape, xemacs, xman). It all goes well enough, but I don't find it as nice to use as MacOS.
I think you'll find that MacOS users generally are used to a consistant user interface and expect one. (There's been a bit of a fuss about some of Apple's own stuff not following the rules - many people use the Quicktime 3 player with Quicktime 4 because the Quicktime 4 player is a mess). Anything that is just 'ported' and doesn't follow the standards is likely to sink like a lead balloon if there is any sort of 'native' alternative.
Having said that, there are lots of non-gui things that I'd be really happy to be able to use - CORBA, Mercury, etc. that don't need a GUI. Being able to port such 'back-end' stuff will be a real boost.
If you are only going to use the command line, what features does MacOS X have that Darwin doesn't? (OK, there are quite a few like some drivers that can't be open sourced and ppp for the version I'm running (1.0.2)).
Darwin is the core of MacOS X , and is moving towards giving most of the command line capabilities of MacOS X. And it's open source.
I'm 32 now, and am working on a research project with people (I think) about 5-10 years younger than myself. I've noticed a few differences as I've got older, and these differences show up between me and other people at work:
- I down tools and go home after my 37.5 hour week, and I've noticed that my colleagues often work evenings or weekends (actually, so does my boss, and he's older than me). If I happen to pass through on an evening or weekend, I suggest to people that they go home;-), but when I was their age, I was doing all-nighters.
- I seem to have lots more non-programming things on the go. This is a good thing.
- I've lost the burning desire to learn about everything that I used to have - I tend to focus more on a few things that really interest me. In particular, learning new languages just for their own sake is not as much fun anymore, and I (with help) only minimally maintain my Linux box, rather than learning all the ins and outs of Linux like I might have done 10 years ago.
- On the plus side, I've often encountered problems before, and I think I sometimes have a broader way of looking at some things (My Slashdot reading colleagues might want to write in and disagree).
I think that as I get older, I'm getting better at coding than I used to be, not because I think faster, but partly because I have more experience, and mainly because I have been finally bitten enough by my sloppy coding and commenting that I'm learning to find ways of doing things right the first time, even if it is not as much fun as hacking. (Actually I think my colleagues are ahead of me on this). However, my motivation has gone down - I do much less coding in my own time nowadays.
Part of these differences will be because when I went to university, Computer Science wasn't worth doing as a major (no Honours course), so I did Mathematics instead. That means I'm a largely self-taught programmer, with lots of maths - this tends to give me some poor coding habits, but a usefully different perspective on problems than those around me who did Computer/Information Science.
I think I've got a lot of useful years yet, but I'm finding a shift to being interested in design rather than the nuts and bolts coding. I still like coding for the really hard problems though.
A site with some information about Ada
on
Why Not Ada?
·
· Score: 2
OK, this is a site with a great deal of Ada advocacy:
http://www.adahome.com/
I've not used Ada (it's on my list of things to try), but it looks like it's got a lot going for it.
I personally like protecting developers from themselves to some extent, because it results in more robust software (depending on the choice of language, no mysterious overflows or pointer problems). This is why at my job we use Java rather than C++... it produces much slower code, but is much faster to develop.
> The fact that they were cold-blooded mass murderers sort of trumps their "victim" status, I'm afraid. NOTHING excuses their crime.
Yes, you are right that nothing _excuses_ their crime, but victim and criminal aren't mutually exclusive. I see too many people throwing away lessons that can be learnt from this sort of thing by painting one side as pure good and another as pure bad.
They were cold-blooded mass murderers, and what they did was wrong, but some of the treatment they received was also wrong, and neither wrong justifies the other. Look at the whole system of what went on, and try to find ways of dealing with the underlying causes.
> Human nature really has not changed, and likely will not.
But it's part of human nature (some humans anyway) to try and find ways changing human nature for the better. Of course we'll never fully succeed, but it's the trying to create a better society, and whatever gains we make along the way, that is important.
I think that one possible use for open source benchmarks are benchmanrks which _can_ be tweeked for individual processors.
To explain: define a set of tasks (this could include some of the same set of tasks as some of the current synthetic benchmarks), but define then in the algorithm that must be used rather than the implementation. Then write a C/whatever standard that implements that algorithm as well as possible to use as a base. _Then_ allow the proponents of particular platforms to modify a version of the code (possibly using #ifdefs or whatever to keep it in one code base) as long as they use the same algorithm.
One possible test (I'm only using it as an example, not suggesting it) would be to calculate a certain portion of the Mandelbrot Set down to a depth of 10000 and put the results in an array of a certain structure, where it must be done using brute force with a presicion of a least 40 binary significant digits (i.e. 64-bit longs or doubles)... edge following not allowed. Part of doing the whole benchmark is doing the test n times, where the position of the result array keeps moving. With that, we'd start with some base code that does the job fairly well, than people can add #ifdef PPC_G3, #ifdef AMD_K6_2 and write pieces of code (using assembler if they like) to speed it up for their favourite architecture. A little bit of competition could be fun:-).
The current distributed.net RC5-64 could be considered an example of such a benchmark - using processor tweeks are good as long as you solve the problem.
Open source can be used to prevent cheating, in that it can be seen that everyone is following the correct algorithm (or strict review by trusted organization as in the case of RC5-64). It also means that people can look over the tweeks for other platforms and see of any of them are applicable.
The rationale for this approach: (1) change the rules so that what is currently 'cheating' becomes part of the process - it becomes very difficult to cheat.
(2) A lot of 'real world' applications like Photoshop and Quake are presumably using these sorts of tweeks for their inner loops anyway, so this is mirrored by allowing the same tweaks in the tests.
This idea has several downsides: (1) it can only provide synthetic benchmarks, and on fairly small examples (so optimizing if for particular archectures doesn't require huge resources)
(2) it only tests the speed that can be got using assembler... how good the compilers are doesn't really get factored in.
(3) It requires each platform to have some advocates good enough and willing to put time into optimizing code so every platform gets a fair go.
(4) because the tests are so small, it needs a moderately large number of individual benchmarks - for instance RC5-64 on its own is useless since it doesn't test memory speed, and PowerPC and x86 architectures have the huge advanatage of having rotate instructions.
(5) rather than give a single number (which is what people tend to want), resulting benchmarks would give a set of results for various aspects of the chip - the would make the results of more interest to technically oriented people.
I'd be willing to put a little work into PowerPC G3 and possibly G4(Altivec) optimization in such a project.
A more extreme version of this idea is to allow algorithm optimization too... like do the Mandelbrot example (allowing edge following etc.) as fast as you can as long as the precision of the results is up to standard. I think that this would require too much time on part of the optimization writers though.
I don't have a link to this info handy, but my recollection is that bubble memory was _way_ too slow - and hard drives just kept getting better...
It seems to be a bit of a trend in this industry that whatever works early on gets a lot of resources put into incrementally improving it and making it cheap, such that competing technologies have to be _hugely_ better to have any chance of taking over.
That is (IMO) partly why: - we still use hard drives, - CPU's still use CMOS rather than one of the faster switching methods, - the x86 architecture is still dominant, - the UNIX model is the base of nearly all operating systems.
There may be potentially 'better' technologies than these out there, but there has been so much engineering and optimisation gone into these technologies that it is really hard for anything to compete.
The case of the Exponential PowerPC is an example of that - it used ECL rather than CMOS to get substantially higher clock speeds, but before it had really got up to speed, the incremental improvements in CMOS had passed made it look less attractive, and Exponential was dead.
I expect someone to reply to this and say how much better CMOS (or whatever) is better than anything else... but at least some of that will be due to the massive research that has gone into making the current technology work well.
OK, I'm not going to give a full technical description here, but:
USB, PS/2, Serial Port, ADB (Apple Desktop Bus) are designed for low speed peripherals such as keyboards, modems or printers. USB does or will rule the roost here.
SCSI, IDE, Firewire are designed for higher speed devices such as hard disks. IDE is very limited (only really any good for internal hard disks), SCSI is (or was) superior to IDE in that it can support up to 7 periperals, longer cables and devices other than hard disks, but has cable termination and multiple version issues. Firewire is being touted as a replacement to SCSI, and will eventually be faster, but SCSI is still the choice for high perfomance if you can put up with its limitations.
Ethernet is designed for connecting computers (and printers) to each other in networks.
There is some overlap - in particular, Intel are pushing USB 2 as something to use instead of firewire. However, USB 2 is still vaporware, will be slower than Firewire, and has a lot of penalties stemming from its design as a relatively low speed protocol. USB 1 is also used for low speed storage like Zip disks.
So the comparison between USB and SCSI doesn't make an awful lot of sense - they are really designed for different things (unless you are talking about something like a Zip disk, in whch case SCSI will be higher performance if available).
Oh, and every computer I have ever bought has had SCSI built in and hasn't been any hassle, so what is buit in depends on what you buy;-)
As for your keyboard, I doubt that Word can keep up with a USB keyboard... you'll just have to type slower:-)
The comments about not allowing 1 as a prime so as not to mess up unique factorization have already been made (and pretty much all the number theory of primes would get messed up with it), but there is also a neat(ish) way of not allowing 1 as a prime number (this is a paraphrase of the definition that I remember being taught):
"A prime number is a positive integer having _two_ unique positive integer factors, 1 and itself"
1 is therefore let off the hook, because 1 only has one factor - itself, not the two required by this definition.
I've looked through the comments, and seen a couple of what I regard as misconceptions:
1) that UNIX will be killed of by Windows or BeOS or...
I think that if something is to replace UNIX, it will have to be something _significantly_ better than UNIX. I don't know of anything else currently out there that can claim this (although I will admit to not having looked at Plan9 and Hurd).
2) that UNIX is immortal because it is somehow 'above' operating system paridyms (sp)?
This is clearly wrong, is the whole UNIX model is based on a hierachy of files, where a file is a stream of bytes/characters. As a consequence of this, a most of the tools that make UNIX so powerful are text stream based.
I see room for a new sort of operating system that works with a much higer level of abstraction throughout, including type management and garbage collection as part of the OS (relegating streams of text to where you really just want to store text information), high level messages between components allowing easy distribution, different views of the file system (or maybe not even so strongly file-system based), and maybe taking some component ideas from Apple's failed OpenDoc to end monolithic applications.
However, I don't think that there is anything out there that comes even close to this, and there won't be for a while, and even if there is, it will take years to catch on.
UNIX will only be overtaken when the problems of the high level of complexity required by what people want to do need a higher level of abstraction than the basic model of UNIX can provide, and even then, UNIX will take a very long time to actually die - just go the same way as DOS that is still widely used but generally regarded as inferior by most people.
I have a dual boot Mac with MacOS/Linux at home. I have CodeWarrior for MacOS, and I definitely prefer MacOS and CodeWarrior, hardly using Linux.
I have a dual boot Linux/NT box at work, and spend nearly all my time in Linux there. I couldn't get used to Windows no matter how much I used it.
It's good to have Linux around (and have ports so fast), but don't write MacOS off so quickly.
For things like the debugger problem, you have enabled the debugger in the Project menu, and turned optimisation off? Sounds like with a little bit of help getting started, your experience of CodeWarrior could be much more pleasant. And with experience, you learn not to write much code that going to cause segfaults/crashes. I _very_ seldom crash my Mac with a program I am working on. I use buggy NetScape to do that instead:(
Convenience to end users doesn't seem to have a huge impact on what succeeds, at least until marketing has reduced the better product to a niche (like VHS did when it was the format that the movies all came out in).
For SCSI and EIDE, the usability is no different whatsoever to consumers, and if you want to go deeper than that, SCSI is much easier than IDE to add another device to (at least it is on the harware I've used).
Similarly, I can't really see that having one type of connector with vastly different speed devices is any better than having two, particularly when we probably end up with something like the connectors on the PC I use - the plugs for the mouse and the keyboard are the same, but if I put them in the wrong way around they won't work.
USB and FireWire (IEEE-1394) aren't competing standards - at least they weren't designed to be. USB is designed for low bandwidth devices, FireWire for high bandwidth, with not a lot of overlap. USB is not offically a standard at all, in that the definition is still under the control of Intel. I'm not saying that USB is bad - but it is optimised for low bandwidth, right down to having cheap cables.
Then Intel gets this idea that they can have the whole pie to themselves (talk about a bob each way - they are part of the IEEE-1394 consortium too), so they bring on this USB 2.0 vaporware - the _specification_ due in a maybe a few months specifies something that maybe will (under ideal conditions) have a similar speed to that available with FireWire _now_.
There is an excellent article on this at:
http://www.MacKiDo.com/Hardware/USB2.html
Read that article if you think that USB 2 will be as good as FireWire.
If USB 2.0 succeeds, it will be on Intel marketing power, not technical merit.
OTOH, that hasn't stopped inferior techology before. I'm writing this on a G3 macintosh with EIDE drives, when EIDE ousted SCSI _not_ because it was in any real way better, but because the fact that it is the 'standard' in PCs made the drives a lot cheaper. I hope we don't also get the 'nearly good enough' technology of USB 2.0 in a similar way.
In terms of pure clock speed, they probably are giving us the best that they can, but clock speed is not the whole story.
The clock speed is the most visible part of the specification of a chip whereas other things like cache size/speed and bus speed are not so obvious to the less technically inclined.
Clock speed being the obvious identifier for marketing, what has happened? There is this race to get the highest number possible before the 'MHz'. Now increasing the clock speed _does_ make the chip faster (all other things being equal), but there's bound to be tradeoffs, and are they doing this at the expense of being able to increase the cache size/speed that might actually make the chip perform better?
For an example of this (admittedly for a different architecture), I was looking at buying a Macintosh or clone a couple of years ago (about the time the clones got killed), and compiled a big list of all the PowerPC models, their CPU, clock speed, bus speed (where I could find it), and cache size/speed, and MacBench scores. I was surprised to find that the cache size/speed was the biggest determinant of MacBench CPU (non floating-point) speed.
Then the G3 machine came out, where the G3 is essentially a 603e with better caching, and a 233 MHz G3 performed comparably with a 350 MHz 604e (which is in many ways a superior chip) simply because the G3 has a fast (for those days) backside L2 cache.
And then of course (still in the PowerPC line) there's things like Altivec, which don't increase the clock speed at all, but in many cases can significantly increase performance.
Oh, and please note that this post is staying out of the minefield of comparing x86 and PowerPC performance/MHz... I'm just using PowerPC as the example I know the best.
So while Intel and AMD might be offering you the fastest clock speeds that they can, they might not be selling the 'fastest' chips possible, because it's high MHz and not performance that markets the best.
>What about all the people who need something low power? It will drive >prices down for other chips, but who needs a 1ghz cpu for home?
If you want something low power, and don't mind switching architecture, you can always try a PowerPC G3/G4. Runs nice and cool. It'll be a while before they reach 1 GHz though.
>Anyone else see some corrolaries between RAM >prices and gas prices? I just saw gas below 60 >cents/litre today for the first time in months. >I think both are ripping the consumer off >because we've no choice.
No choice? Of course you've got a choice... (1) Don't buy RAM (2) Walk, cycle, or use public transport.
The choices might not be very palatable to some people ((2) might require relocating/changing jobs), but don't confuse that with not having a choice.
I'm waiting for prices to go to something sensible before getting a RAM upgrade. That's my choice.
A few inaccuracies with the post I am replying to (Most of these are just nit-picks from someone familiar with the instruction set):
* It is SIMD, in that each altivec operation does the same thing to each part of the altivec register.
* Altivec has 32 Altivec (128 bit) registers, not 64, in addition to the 64-bit floating point and 32-bit integer registers.
* Altivec registers can be treated as: - 16 8-bit, 8 16-bit or 4-32-bit integers - 4 32-bit floats - 128 bits for bitwise operations There is no support for 64-bit integers, or for 16-bit or 64-bit floats - only 32-bit floats are supported.
* There is no explicit support for fixed point operations, although there is provision for converting between floating and fixed point.
* Altivec instructions take 1,2 _or_3_ registers as source, and one as destination.
It is also rather fun to program (even if only using an emulator at the moment):-)
Slashdot Mirror
I can't see what not watching TV has to do with ignorance - TV is about entertainment for the purpose of selling advertising or subscriptions. Period. You don't really get much 'news' in 20 second soundbites - I think 'infotainment' is the best description I've seen of TV news. It scares me that someone might think that they actually know what is 'going on in the society' by watching TV.
Newspapers and National Radio (in New Zealand) are at least slightly better, and websites/email lists can be great for news related to paricular issues .
Oh, and having aquired a TV recently after 4 years without one, I don't watch it much, and doubt I'll be watching any of the Olympics. I'll be too busy doing interesting/useful things to 'vegetate in front of the box'.
Roy Ward.
... I've definitely seen this layout somewhere before ;-) :
- ------------------------
The requested URL could not be retrieved
-----------------------------------------------
While trying to retrieve the URL: http://hobbes.resnet.tamu.edu/
The following error was encountered:
Unable to determine IP address from host name for hobbes.resnet.tamu.edu
The dnsserver returned:
Name Error: The domain name does not exist.
This means that:
The cache was not able to resolve the hostname presented in the URL.
Check if the address is correct.
I can well believe that Linux is poised to overtake MacOS in pure numbers of users - it runs on a _lot_ more hardware for a start.
However, this in no way condemns MacOS - there is room for both for different sorts of tasks. I use Linux at work and MacOS at home, and those are the environements that are best for me at each place.
Linux is great because it's main focus is power and configurability. However these very things detract from how easy it is to use. Great for hackers, servers, and command line stuff.
MacOS has the best interface I've ever used (I never used a NeXT box), and that is because this is the primary goal of its designers. Great for all that GUI stuff - graphics, writing, and getting not too complex things done as simply as possible (and I really like using an IDE for small to medium programming projects). There is a group of users that MacOS is not going to lose in a hurry.
I think the OSes currently have different audiences for the most part, such as people with different preferences about command-line/GUI.
Of course, I have great hopes for MacOS X combining the best of both worlds (I already run Darwin for tasks I can't use MacOS for), but it will invariably make some compromises because I don't think it is yet possible to combine the best possible power _and_ usability.
It would be really neat if Eazel does well, but that is a bit further down the track, and I don't think they have the resources to do the sort of things being done with MacOS X.
I really like a bit of diversity and choice is OSes - this helps make them all better.
Write the code using your favourite development environment, then port it to CodeWarrior (it shouldn't be that hard as long as you stick to fairly standard C++).
This gives you the advantage of using your own environment, and still meeting the requirement.
Of course if it _is_ difficult to port, then you've found out something about writing multi-platform code. C++ is notorious with differences between platforms.
Roy Ward.
Where MacOS X might have a niche as a server is for an organisation that is too small to have a UNIX Admin - I suspect that it's going to have a hugely better learning curve than UNIX, and be easier to do simple administration.
I suspect that if you've got someone who can administrate UNIX well, then you'd be better going with a UNIX or Linux solution. MacOS X (which I assume this discussion is actually about, rather than MacOS X Server which has been around for a while) looks like it's going to be a great desktop and workstation environment, but doesn't AFAIK (I am not a server expert) offer much extra particularly aimed for serving (Web Objects is one such thing?).
> My only hesitation is supporting a monopoly, be it Microsoft or Apple. Apple killed off all of their hardware competition, which was a shame, as I think some of their "competitors" were actually making better "Mac" hardware than Apple was (and at a lower price).
Yes they did (in some ways, although I don't think the quality of many of the clones was as good). That was one of the reasons that Apple (arguably) needed to kill them. The argument as I understand it was that the clonemakers were taking the fruits of Apples R&D (motherboard designs, etc), putting faster CPUs in them (which they could do because they were working with smaller volumes, and were nimbler than Apple anyway), then undercutting Apple. This wasn't particularly sustainable - Apple doing a lot of the work, and others taking the profits.
... unless you are prepared to do a proper job of the user interface.
The real advantage of MacOS is the relatively clean user interface, and that most of the applications follow the rules.
Hopefully MacOS X is going to maintain that usability (and dare I hope even improve on it?).
Now look at the X world (I use Linux at work). Many apps have different appearances because they use different windowing toolkits and many different conventions for how to do things (how many ways of scrolling or cut and paste are there? - compare Netscape, xemacs, xman). It all goes well enough, but I don't find it as nice to use as MacOS.
I think you'll find that MacOS users generally are used to a consistant user interface and expect one. (There's been a bit of a fuss about some of Apple's own stuff not following the rules - many people use the Quicktime 3 player with Quicktime 4 because the Quicktime 4 player is a mess). Anything that is just 'ported' and doesn't follow the standards is likely to sink like a lead balloon if there is any sort of 'native' alternative.
Having said that, there are lots of non-gui things that I'd be really happy to be able to use - CORBA, Mercury, etc. that don't need a GUI. Being able to port such 'back-end' stuff will be a real boost.
Download Darwin. Install. Don't bother with X.
If you are only going to use the command line, what features does MacOS X have that Darwin doesn't? (OK, there are quite a few like some drivers that can't be open sourced and ppp for the version I'm running (1.0.2)).
Darwin is the core of MacOS X , and is moving towards giving most of the command line capabilities of MacOS X. And it's open source.
... will crashing a few of these into Mars make much of a difference? ;-)
(sorry, couldn't resist).
I'm 32 now, and am working on a research project with people (I think) about 5-10 years younger than myself. I've noticed a few differences as I've got older, and these differences show up between me and other people at work:
;-), but when I was their age, I was doing all-nighters.
- I down tools and go home after my 37.5 hour week, and I've noticed that my colleagues often work evenings or weekends (actually, so does my boss, and he's older than me). If I happen to pass through on an evening or weekend, I suggest to people that they go home
- I seem to have lots more non-programming things on the go. This is a good thing.
- I've lost the burning desire to learn about everything that I used to have - I tend to focus more on a few things that really interest me. In particular, learning new languages just for their own sake is not as much fun anymore, and I (with help) only minimally maintain my Linux box, rather than learning all the ins and outs of Linux like I might have done 10 years ago.
- On the plus side, I've often encountered problems before, and I think I sometimes have a broader way of looking at some things (My Slashdot reading colleagues might want to write in and disagree).
I think that as I get older, I'm getting better at coding than I used to be, not because I think faster, but partly because I have more experience, and mainly because I have been finally bitten enough by my sloppy coding and commenting that I'm learning to find ways of doing things right the first time, even if it is not as much fun as hacking. (Actually I think my colleagues are ahead of me on this). However, my motivation has gone down - I do much less coding in my own time nowadays.
Part of these differences will be because when I went to university, Computer Science wasn't worth doing as a major (no Honours course), so I did Mathematics instead. That means I'm a largely self-taught programmer, with lots of maths - this tends to give me some poor coding habits, but a usefully different perspective on problems than those around me who did Computer/Information Science.
I think I've got a lot of useful years yet, but I'm finding a shift to being interested in design rather than the nuts and bolts coding. I still like coding for the really hard problems though.
OK, this is a site with a great deal of Ada advocacy:
... it produces much slower code, but is much faster to develop.
http://www.adahome.com/
I've not used Ada (it's on my list of things to try), but it looks like it's got a lot going for it.
I personally like protecting developers from themselves to some extent, because it results in more robust software (depending on the choice of language, no mysterious overflows or pointer problems). This is why at my job we use Java rather than C++
Roy Ward.
> The fact that they were cold-blooded mass murderers sort of trumps their "victim" status, I'm afraid. NOTHING excuses their crime.
Yes, you are right that nothing _excuses_ their crime, but victim and criminal aren't mutually exclusive. I see too many people throwing away lessons that can be learnt from this sort of thing by painting one side as pure good and another as pure bad.
They were cold-blooded mass murderers, and what they did was wrong, but some of the treatment they received was also wrong, and neither wrong justifies the other. Look at the whole system of what went on, and try to find ways of dealing with the underlying causes.
> Human nature really has not changed, and likely will not.
But it's part of human nature (some humans anyway) to try and find ways changing human nature for the better. Of course we'll never fully succeed, but it's the trying to create a better society, and whatever gains we make along the way, that is important.
I think that one possible use for open source benchmarks are benchmanrks which _can_ be tweeked for individual processors.
... edge following not allowed. Part of doing the whole benchmark is doing the test n times, where the position of the result array keeps moving. With that, we'd start with some base code that does the job fairly well, than people can add #ifdef PPC_G3, #ifdef AMD_K6_2 and write pieces of code (using assembler if they like) to speed it up for their favourite architecture. A little bit of competition could be fun :-).
... how good the compilers are doesn't really get factored in.
... like do the Mandelbrot example (allowing edge following etc.) as fast as you can as long as the precision of the results is up to standard. I think that this would require too much time on part of the optimization writers though.
To explain: define a set of tasks (this could include some of the same set of tasks as some of the current synthetic benchmarks), but define then in the algorithm that must be used rather than the implementation. Then write a C/whatever standard that implements that algorithm as well as possible to use as a base. _Then_ allow the proponents of particular platforms to modify a version of the code (possibly using #ifdefs or whatever to keep it in one code base) as long as they use the same algorithm.
One possible test (I'm only using it as an example, not suggesting it) would be to calculate a certain portion of the Mandelbrot Set down to a depth of 10000 and put the results in an array of a certain structure, where it must be done using brute force with a presicion of a least 40 binary significant digits (i.e. 64-bit longs or doubles)
The current distributed.net RC5-64 could be considered an example of such a benchmark - using processor tweeks are good as long as you solve the problem.
Open source can be used to prevent cheating, in that it can be seen that everyone is following the correct algorithm (or strict review by trusted organization as in the case of RC5-64). It also means that people can look over the tweeks for other platforms and see of any of them are applicable.
The rationale for this approach:
(1) change the rules so that what is currently 'cheating' becomes part of the process - it becomes very difficult to cheat.
(2) A lot of 'real world' applications like Photoshop and Quake are presumably using these sorts of tweeks for their inner loops anyway, so this is mirrored by allowing the same tweaks in the tests.
This idea has several downsides:
(1) it can only provide synthetic benchmarks, and on fairly small examples (so optimizing if for particular archectures doesn't require huge resources)
(2) it only tests the speed that can be got using assembler
(3) It requires each platform to have some advocates good enough and willing to put time into optimizing code so every platform gets a fair go.
(4) because the tests are so small, it needs a moderately large number of individual benchmarks - for instance RC5-64 on its own is useless since it doesn't test memory speed, and PowerPC and x86 architectures have the huge advanatage of having rotate instructions.
(5) rather than give a single number (which is what people tend to want), resulting benchmarks would give a set of results for various aspects of the chip - the would make the results of more interest to technically oriented people.
I'd be willing to put a little work into PowerPC G3 and possibly G4(Altivec) optimization in such a project.
A more extreme version of this idea is to allow algorithm optimization too
I don't have a link to this info handy, but my recollection is that bubble memory was _way_ too slow - and hard drives just kept getting better ...
... but at least some of that will be due to the massive research that has gone into making the current technology work well.
It seems to be a bit of a trend in this industry that whatever works early on gets a lot of resources put into incrementally improving it and making it cheap, such that competing technologies have to be _hugely_ better to have any chance of taking over.
That is (IMO) partly why:
- we still use hard drives,
- CPU's still use CMOS rather than one of the faster switching methods,
- the x86 architecture is still dominant,
- the UNIX model is the base of nearly all operating systems.
There may be potentially 'better' technologies than these out there, but there has been so much engineering and optimisation gone into these technologies that it is really hard for anything to compete.
The case of the Exponential PowerPC is an example of that - it used ECL rather than CMOS to get substantially higher clock speeds, but before it had really got up to speed, the incremental improvements in CMOS had passed made it look less attractive, and Exponential was dead.
I expect someone to reply to this and say how much better CMOS (or whatever) is better than anything else
OK, I'm not going to give a full technical description here, but:
;-)
... you'll just have to type slower :-)
USB, PS/2, Serial Port, ADB (Apple Desktop Bus) are designed for low speed peripherals such as keyboards, modems or printers. USB does or will rule the roost here.
SCSI, IDE, Firewire are designed for higher speed devices such as hard disks. IDE is very limited (only really any good for internal hard disks), SCSI is (or was) superior to IDE in that it can support up to 7 periperals, longer cables and devices other than hard disks, but has cable termination and multiple version issues. Firewire is being touted as a replacement to SCSI, and will eventually be faster, but SCSI is still the choice for high perfomance if you can put up with its limitations.
Ethernet is designed for connecting computers (and printers) to each other in networks.
There is some overlap - in particular, Intel are pushing USB 2 as something to use instead of firewire. However, USB 2 is still vaporware, will be slower than Firewire, and has a lot of penalties stemming from its design as a relatively low speed protocol. USB 1 is also used for low speed storage like Zip disks.
So the comparison between USB and SCSI doesn't make an awful lot of sense - they are really designed for different things (unless you are talking about something like a Zip disk, in whch case SCSI will be higher performance if available).
Oh, and every computer I have ever bought has had SCSI built in and hasn't been any hassle, so what is buit in depends on what you buy
As for your keyboard, I doubt that Word can keep up with a USB keyboard
The comments about not allowing 1 as a prime so as not to mess up unique factorization have already been made (and pretty much all the number theory of primes would get messed up with it), but there is also a neat(ish) way of not allowing 1 as a prime number (this is a paraphrase of the definition that I remember being taught):
"A prime number is a positive integer having _two_ unique positive integer factors, 1 and itself"
1 is therefore let off the hook, because 1 only has one factor - itself, not the two required by this definition.
I've looked through the comments, and seen a couple of what I regard as misconceptions:
...
1) that UNIX will be killed of by Windows or BeOS or
I think that if something is to replace UNIX, it will have to be something _significantly_ better than UNIX. I don't know of anything else currently out there that can claim this (although I will admit to not having looked at Plan9 and Hurd).
2) that UNIX is immortal because it is somehow 'above' operating system paridyms (sp)?
This is clearly wrong, is the whole UNIX model is based on a hierachy of files, where a file is a stream of bytes/characters. As a consequence of this, a most of the tools that make UNIX so powerful are text stream based.
I see room for a new sort of operating system that works with a much higer level of abstraction throughout, including type management and garbage collection as part of the OS (relegating streams of text to where you really just want to store text information), high level messages between components allowing easy distribution, different views of the file system (or maybe not even so strongly file-system based), and maybe taking some component ideas from Apple's failed OpenDoc to end monolithic applications.
However, I don't think that there is anything out there that comes even close to this, and there won't be for a while, and even if there is, it will take years to catch on.
UNIX will only be overtaken when the problems of the high level of complexity required by what people want to do need a higher level of abstraction than the basic model of UNIX can provide, and even then, UNIX will take a very long time to actually die - just go the same way as DOS that is still widely used but generally regarded as inferior by most people.
Y2K already in New Zealand ... and lights, internet etc. going without even a flicker at midnight here (Invercargill).
Happy New Year!
This so called 'paperless society' and 'paperless office' is what we occasianally talk about at work - while standing around waiting at the printer.
Roy Ward.
That depends on what you are used to.
:(
I have a dual boot Mac with MacOS/Linux at home. I have CodeWarrior for MacOS, and I definitely prefer MacOS and CodeWarrior, hardly using Linux.
I have a dual boot Linux/NT box at work, and spend nearly all my time in Linux there. I couldn't get used to Windows no matter how much I used it.
It's good to have Linux around (and have ports so fast), but don't write MacOS off so quickly.
For things like the debugger problem, you have enabled the debugger in the Project menu, and turned optimisation off? Sounds like with a little bit of help getting started, your experience of CodeWarrior could be much more pleasant. And with experience, you learn not to write much code that going to cause segfaults/crashes. I _very_ seldom crash my Mac with a program I am working on. I use buggy NetScape to do that instead
Roy Ward.
Marketing, yes :-(
I'm not sure what convenience has to do with it.
Convenience to end users doesn't seem to have a huge impact on what succeeds, at least until marketing has reduced the better product to a niche (like VHS did when it was the format that the movies all came out in).
For SCSI and EIDE, the usability is no different whatsoever to consumers, and if you want to go deeper than that, SCSI is much easier than IDE to add another device to (at least it is on the harware I've used).
Similarly, I can't really see that having one type of connector with vastly different speed devices is any better than having two, particularly when we probably end up with something like the connectors on the PC I use - the plugs for the mouse and the keyboard are the same, but if I put them in the wrong way around they won't work.
Roy Ward.
USB and FireWire (IEEE-1394) aren't competing standards - at least they weren't designed to be. USB is designed for low bandwidth devices, FireWire for high bandwidth, with not a lot of overlap. USB is not offically a standard at all, in that the definition is still under the control of Intel. I'm not saying that USB is bad - but it is optimised for low bandwidth, right down to having cheap cables.
Then Intel gets this idea that they can have the whole pie to themselves (talk about a bob each way - they are part of the IEEE-1394 consortium too), so they bring on this USB 2.0 vaporware - the _specification_ due in a maybe a few months specifies something that maybe will (under ideal conditions) have a similar speed to that available with FireWire _now_.
There is an excellent article on this at:
http://www.MacKiDo.com/Hardware/USB2.html
Read that article if you think that USB 2 will be as good as FireWire.
If USB 2.0 succeeds, it will be on Intel marketing power, not technical merit.
OTOH, that hasn't stopped inferior techology before. I'm writing this on a G3 macintosh with EIDE drives, when EIDE ousted SCSI _not_ because it was in any real way better, but because the fact that it is the 'standard' in PCs made the drives a lot cheaper. I hope we don't also get the 'nearly good enough' technology of USB 2.0 in a similar way.
Roy Ward
In terms of pure clock speed, they probably are giving us the best that they can, but clock speed is not the whole story.
... I'm just using PowerPC as the example I know the best.
The clock speed is the most visible part of the specification of a chip whereas other things like cache size/speed and bus speed are not so obvious to the less technically inclined.
Clock speed being the obvious identifier for marketing, what has happened? There is this race to get the highest number possible before the 'MHz'. Now increasing the clock speed _does_ make the chip faster (all other things being equal), but there's bound to be tradeoffs, and are they doing this at the expense of being able to increase the cache size/speed that might actually make the chip perform better?
For an example of this (admittedly for a different architecture), I was looking at buying a Macintosh or clone a couple of years ago (about the time the clones got killed), and compiled a big list of all the PowerPC models, their CPU, clock speed, bus speed (where I could find it), and cache size/speed, and MacBench scores. I was surprised to find that the cache size/speed was the biggest determinant of MacBench CPU (non floating-point) speed.
Then the G3 machine came out, where the G3 is essentially a 603e with better caching, and a 233 MHz G3 performed comparably with a 350 MHz 604e (which is in many ways a superior chip) simply because the G3 has a fast (for those days) backside L2 cache.
And then of course (still in the PowerPC line) there's things like Altivec, which don't increase the clock speed at all, but in many cases can significantly increase performance.
Oh, and please note that this post is staying out of the minefield of comparing x86 and PowerPC performance/MHz
So while Intel and AMD might be offering you the fastest clock speeds that they can, they might not be selling the 'fastest' chips possible, because it's high MHz and not performance that markets the best.
>What about all the people who need something low power? It will drive
>prices down for other chips, but who needs a 1ghz cpu for home?
If you want something low power, and don't mind switching architecture, you can always try a PowerPC G3/G4. Runs nice and cool. It'll be a while before they reach 1 GHz though.
Of course, clock speed isn't everything.
>Anyone else see some corrolaries between RAM
...
>prices and gas prices? I just saw gas below 60
>cents/litre today for the first time in months.
>I think both are ripping the consumer off
>because we've no choice.
No choice? Of course you've got a choice
(1) Don't buy RAM
(2) Walk, cycle, or use public transport.
The choices might not be very palatable to some people ((2) might require relocating/changing jobs), but don't confuse that with not having a choice.
I'm waiting for prices to go to something sensible before getting a RAM upgrade. That's my choice.
Roy Ward.
A few inaccuracies with the post I am replying to (Most of these are just nit-picks from someone familiar with the instruction set):
:-)
* It is SIMD, in that each altivec operation does the same thing to each part of the altivec register.
* Altivec has 32 Altivec (128 bit) registers, not 64, in addition to the 64-bit floating point and 32-bit integer registers.
* Altivec registers can be treated as:
- 16 8-bit, 8 16-bit or 4-32-bit integers
- 4 32-bit floats
- 128 bits for bitwise operations
There is no support for 64-bit integers, or for 16-bit or 64-bit floats - only 32-bit floats are supported.
* There is no explicit support for fixed point operations, although there is provision for converting between floating and fixed point.
* Altivec instructions take 1,2 _or_3_ registers as source, and one as destination.
It is also rather fun to program (even if only using an emulator at the moment)
- Roy Ward.