You could get Playboy in my (Catholic!) high-school library if you were 18.
You forget that Catholic schools have an overwhelming desire to raise children according to a certain moral/ethical code. The best way to affect this (according to the more enlightened) is to expose them to stimuli that are considered tasteless and/or questionable and discuss it in an elightened manner in a classroom setting. While Playboy may not be considered educational in any way, certain works of literature that the less enlightened want banned (i.e., The Chocolate War or Smilla's Sense of Snow) due to some adult material are read and discussed in classrooms so that students can learn about what goes on in the real world, and be prepared to deal with it according to the moral/ethical standard espoused by the Catholic Church. The English Literature teacher at my sisters' (Catholic, all-girl) high school said almost exactly what I just wrote (in many fewer words) when a group of parent's tried to get the aforementioned Smilla's Sense of Snow banned from the curriculum. Needless to say, their effort failed. The nice thing about private schools is that if you do not like the way things are being taught, you have every right to withdraw your children from that school. Unfortunately, not everyone has that option, but I won't get into the whole school voucher thing (which now that I think about it, would doom the private schools to the same Luddite fate as the public institutions).
I had the music (all of 'em, but especially the 'B' music, ahh, I can hear now) playing in my head within a day or two of getting my first GameBoy (the one that came bundled with Tetris ahh, for the days when consoles came bundled with a game). Couldn't get it out of my skull for weeks. Definitely the most addictive game ever.
First, it means that if I try to do further science using the entire body of data, I will be bound by this agreement not to redistribute the very data that I use to form my conclusions. If I can't release the data I use for theoretical work, can I really claim that I'm engaging in science, myself? Any attempt at scientific work that is based on Celera's data is tainted in an important way.
The thing is, I don't want to have to put Celera's data in my paper. It is much easier to refer the reader to a previously published article or online database of their sequence. And boy, once the Celera article is published, it's gospel. All I have to do is cite the previous work, as in; "...as published previously, in Celera, et. al...."
Anyway, imagine that you saw a flaw in Celera's reasoning and published your own paper using Celera's own data to refute their conclusions and make them look silly. Suppose that Celera's investors have sunk a great deal of money into a product based on some bit of flawed "science" that they've attempted to do.
...what happens when Celera has second thoughts about continuing to make these data availible to the scientific community. Since you can't publish Celera's data in your research, you have no choice but to point your readers to Celera to get their own copies of the data that are the foundation of your research. But if Celera is embarassed and stops making these data availible to the scientist who are reading your paper...
If I were to have found a flaw in their research, I am perfectly free to report that (provided it makes it past the review process). Once their paper is published, presenting their methods and results, there ain't no way in hell they'd be able to retract that (the data maybe, not the paper). I've read articles that have glaring basic errors, and let me tell you, I pity the author of that paper. Anytime I hear the authors' names now, all I can think of is, "duh." This is the same thing that would happen to any more research coming out of Celera; it would always be mistrusted.
This is why Science is holding a copy of the data in escrow, to prevent just such an occurence. This is one thing I do trust, since Science has too good a reputation to let one company tell them how to run things. Even if that one company is one of only two major players in the field of human genome sequencing.
I'd focus on data storage first; without that, it really doesn't matter what kind of pretty GUI face you put on the front of it.
I think the concept of soups is OK for some types of applications, but overall I prefer the structured file system route. For example, on the PalmOS devices, the databases for memos, appointments, etc. are OK (who uses more than one appointment book?), especially when syncing them with desktop applications, but with productivity apps, like spreadsheets, you have to start shoe-horning files into a database structure, making the program take into account things like file/record organization (i.e., my gas milage spreadsheet goes into my personal directory, but expense accounts go into the business directory). Filesystems already do this well, and re-inventing the wheel is very rarely a good thing. As for data corruption, I think my Psion has crashed 1 time in the 2 years I've had it. The system also (I believe) does not use file system buffers, as the buffer and writing to disk take the same amount of time and memory (shared memory/file system model). If you save to CF disk, there's nothing that's killing that data short of a fire.
I posit that the interface and accessiblity of data is paramount (much more so than on the desktop), once we have established an easy way of storing and accessing the data. These are devices we use everyday, and we don't want to look like an idiot when writing down a phone number or address because it takes twice as long as finding a pen and paper. Access to the data is the key. WinCE failed in its initial incarnations for this very reason. Damn, was it clunky.
I think that one thing left out on a lot of these Linux handheld solutions is thought about the design. Not intending to start a flame war, but it seems as if these all try (with the exception of the X+TWM+XTerm solution at handhelds.org, which , as the article stated, completely ignores any usability guidelines) to more or less copy the Palm way of doing things, and then graft on neat things like skins, etc. While mimicry is a good way to start, I think one thing to look at is what you really need in a handheld OS. From my point of view as an avid user of EPOC, I think you need to decide on the following things:
Input Method: This is key. If you decide handwriting is the way to go, then by all means do it, but make it easy. I'm personally a fan of the write-anywhere-on-the-screen technique, but that's due to my large handwriting.
Application Launching: This is very dependent upon the form factor of the device, and very influential in the overall operation of the device. For example, the Psion palmtops are currently all landscape based with a keyboard. There is a large (relatively) screen, and the device focuses on documents. There are very few applications that you must start from the launcher as opposed to opening a file or asking the OS to create a new one. This works very well for these devices. The Psion is also multi-tasking, and you can have boatloads of programs and documents open at the same time, copying and pasting from one to another. Most Psion users tend to leave open the most frequently used documents and switch between them. At the opposite end of the spectrum is the Palm, which uses a program centric model, with the data all tucked away from the user in databases. Only one program runs at a time. This seems to work well for this device. There are also differences in the way apps are launched. Palm has screens of apps to launch, and the associated documents are contained therein. On Psions, you have screens of documents, and bars of applications that pop up over the running app. Both work for the intended function, but the thing common to both (and part of the reason they work well and quickly) is that it takes two taps to open any application (more or less). Start menu type launchers are a lot slower because you have to navigate nested menus. This is OK for infrequently used functions, but too slow for common apps. PocketPC is just now getting this right, but some of the standard programs don't have an option to close them, so they just sit in memory and eat up space (the iPaq has a app terminate on its special 'Q' button). If you only use a spreadsheet app (I'm just using this as an example, I don't know if PocketExcel has a close option or not) once a month, it's a pretty big waste of space to keep it open all the time.
Output Method: Some people think color is the way to go. If you're going to implement color, make it worth my time and money. Psion and Palm color devices seem to add color to the OS as an afterthought. Since MPEG players and the like don't exist for either platform, I'm not so certain that I need color to get my work done (although it is pretty). This is the one area where MS has it right. PocketPC uses color very well in its interface. There is such thing as too much color, though. Don't overload the interface with different shades of green and blue, and require that color be important in navigation. There is a reason white golf balls against green grass and black ink on white paper have worked well for years, and that reason is contrast. When you're looking for something, contrast makes the job a lot faster. Again, PocketPC on the iPaq works really well here. Another output method you don't think about too much is sound. Sometimes the sound feedback from tapping the screen is important, but sometimes the bells, beeps, and dings you get when tapping on things gets unbearable. PocketPC really sucks at using sound well as an indicator of action (at least out of the box).
Screen Space Management: On devices with small screens, you absolutely must devote as much space to data and program function as possible. One of the big gripes Psion users had about the original WinCE was that there were toolbars and taskbars and menu bars that ate up 10-20% of the screen real estate, and you might only use them once while running the program. Psion and Palm devices float the menu over the app only when the menu is called, and this goes a long way toward saving space. The Psion also allows you (in all standard and most add on apps) to turn off all the tool bars, which makes viewing as much data as possible on the screen much easier. So, for example, you can draw on the entire screen in the Sketch app, or see all of the gas milage data in Sheet without scrolling. Viewing a lot of data at once is important on small devices with small screens.
In summary, you need a fast, easy input method, with a fast, intuitive way to launch apps, and a way to view all the necessary data at the same time (color and screen use both come into play here). After all, what are these things for if not fast, easy access to the data we can't go 2 minutes without having access to? I'm curious, though, what features do people think are really necessary in these things? I for one think MP3 players and most network connectivity are overrated. I also only care about getting the files off the device and being able to use them in other applications. Syncing, for me, is just another buzzword, as I keep my phone book and appointments in only one place (on the handheld). I know others will take issue with these points, though. Flame on.
Some of these items have been covered in other posts since I started writing this tome, but I'm going to post it anyhow since they weren't covered in the detail I wanted to see.
MS and IBM were working on OS/2 together, at first (v1.0). Then IBM started getting wonky about things, and MS said "stuff it, we'll do our own thing." They then went and wrote NT v1.0.
A friend of mine used to work for IBM, and he was telling me some of the things about the IBM environment at the time. He said that there were feature request sheets that IBM's groups would fill out, and these features would have to be implemented, regardless of how they conflicted with other feature requests. In the end, MS just got frustrated. If I recall correctly, there is also some discussion of this in Bill Gates' The Road Ahead (I know, I know, but I was young and impressionable then! I swear I didn't finish the book!).
I wouldn't recommend storing the books upright (spines up.) Especially on glue-bound books, the binding gets bent and breaks after prolonged storage, especially in a humid spot
One thing not explicitly stated, storing the books spine down works really well. Nothing pulls away from the spine, and you can get any of the books out with ease (as opposed to laying them flat, when you have to unpack the whole box to get to that good one on the bottom). I used this when I moved to New Orleans, and all of my old technical manuals (ca. 1930's CRC math tables book and ca. 1950's Mark's Handbook for Mechanical Engineers) that I rescued from the Engineering Library at my Alma Mater survived unscathed.
Additionally, I used some boxes I got from a Pharmacy (my Dad is a Pharmacist) that were initially used to ship vials and their lids. They are sturdy, one piece, and about the size of a copier paper box. They also fold up easily (designed to do so), and I have kept all of them for future moves (when finished with Grad school). The additional benefit of boxes this size is that when full of books, they are still easily managed by one person. I might ask at a Pharmacy about these type of boxes.
One day, management decided to pipe some happy-happy shit^H^H^H^Hmusic
through the PA system.
For a time, I worked at a power plant (Mechanical Engineer by training), and in the office the phones were hooked into a radio system of some sort, and the music played over a speaker on the front of the phone. The secretaries were in control of this, and they always had that damn thing tuned into the adult contemporary station that I swear had about ten songs that they cycled through. The worst part of it all was that while the speakers on the individual phones could be turned off, it seemed as if everyone else in the whole damn facility loved that tripe. The office was (surprizingly for being in a 1500+ MW generating station) remarkably quiet, and I could hear every God-forsaken tune over the cubical walls. I couldn't get away from it, and I couldn't control it. Thankfully, I only spent about a month (total) working in that facility. I think that has to be the single most irritating thing about any office environment; a completely dictated atmosphere. At least let me control what (if anything) I listen to.
The other problem I have is with TV's. Those things suck my brain right out of my head through my eyeballs, and I can't maintain a train of thought for more that 30 seconds in front of one. It doesn't matter what is on it. I have to turn the thing off to look at pr0n effectively, let alone try to get real work done.
I want something the size of a psion (preferably with a small keyboard) which can run gcc, has a serial port, uses AA batteries (so I can replace them with a quick visit to a newsagent), has at least a 24hr battery life, has a PCMCIA slot (probably for a 340mb IBM microdrive or an ethernet card) and fits nicely in my pocket.
It doesn't meet all of your specs, but you can run Linux on a Psion Series 5. If you can handle the larger size, the netBook (or Series 7, I can't remember which) will take a MicroDrive.
I also would love to try and "roll my own" with a LART, but I also lack the hardware skills to implement my dream machine.
I ran Doom (DOS version, 1.9) on NT4.0 just last Friday (got nostalgic). No problems whatsoever.
The only way I can see a program like Virtual PC being "more realistic" is if it allows direct hardware calls, which AFAIK, WinNT did not allow from version 1.0
Another game (Lucas Arts as well) that fits in with this is 'Grim Fandango.' It also has 3D graphics, an intriguing story line, and engaging characters. It also took a different step in terms of interface in that Manny (the main character) was the interface, and anything he looked at (generally, his head would track the object until it was out of his field of view) was fair game for any of the usage options. Another great game all around.
The only programming that I have done on my Series 5 is some OPL coding, and some vanilla text editing of FORTRAN code for use on larger machine. The keyboard and portability make it VERY nice. There is a 'sort of' programmer's editor available here that does auto-indenting, etc. As far as Java goes, I do not know the language myself, however:
There is a Java compiler implemented in Java in the works here.
And (according to the web page) a fully working one here.
I THINK that both of these should run in the JVM for EPOC, but I suggest checking before running out and buying a 5mx (~$500, but well worth it) strictly for Java use.
The other thing to keep in mind is the limited memory on these devices. You may need to add a CF card to hold all of your development environment. YOU HAVE BEEN WARNED!
Perl is available for handhelds running EPOC (i.e., the Psion Series 5/5mx/7/netBook/Revo, the Oregon Scientific Osiris, and the Ericsson MC218). The installation is LARGE for a device of this caliber (~4MB of storage and running space). You can check out the author's web page for the port here.
Alternatively, most of the machines I listed (with the exception of the Revo and possibly the Osiris) come with a programming language editor and byte-code compiler built in. The language is called OPL32 (organizer programming language 32-bit). OPL is a relatively powerful language (even if it is interpreted), with the capability to expand (i.e., graphical printing functions, multi-page dialogs) via.OPX libraries.
The newer machines (NOT the Series 5, and maybe not the Revo or Osiris) can also run Java. Which may also meet your needs.
If you would like to learn more about programming the EPOC OS, you can download OPL and Java SDK's, and the included emulator, from the Symbian site (you have to register to download, but it's free. It does run only under Windows, though). One additional note, the Perl port will likely NOT run under the emulator (compiled OPL will run under both, C++, will not). You have been warned.
I agree. One of the things I actually like about Win95/98 is that I can use the keyboard to start about anything. In my case, WinKey (or Ctrl-Esc)-p-o-w starts up Word [1]. Replace the w with an e, and I'm running Excel. Discounting start-up time, I estimate it takes me at most a quarter the time of a mouse user to start my Office apps.
There is also the WinKey-r series for a 'shell prompt' of sorts. I never look through the accessories menu for the tiny calculator icon, instead I hit WinKey-r and type calc. I personally think that these two features work really, really well. When I ran Win3.1, I had a command prompt open on startup for some tasks. I've found that in Win95/98, I don't always need that prompt there (even with the added integration). Try and take the prompt away from me, though (damn Macs) and you'll lose a hand.
[1] In case you aren't familiar with the feature, the Start menu can be set up to use keystrokes easily. If only one item on the current menu starts with the letter you press, it is selected and opened/executed; more than one and the first, then next is selected with subsequent key presses, and enter opening/executing. Right and left keys open and close menus, and up/down navigating.
Finally, this discussion comes to light without me having to ask it. I've been thinking about trying this for about 2 or 3 months, the problem I have is that my design seems rather novel, and I havn't seen any of the small boards/systems (like LART, TINI, TIQIT, uCLinux) that would support it hardware-wise, let alone getting the software working for it. Most of the following rant is nothing more than a pipe dream, but if you know of something to make this dream a reality, by God post it.
I envision a bifold device, about the size of a deck of cards, with rounded edges. The device flips open (either spring loaded to a set angle ala Psion, or with a friction clutch holding it at any angle), and both inside faces are LCDs with digitizers on them. The side that you hold in your hand is the input area,which can be changed to fit the application (for example, all buttons for a calculator app, or a few buttons with a writing area for a notepad app). The top half is the display area.
Two displays/digitizers may seem like a waste, but an infinitely configurable input area seems like it might be just the kind of thing to make this badboy very easy and fast to use. The other nice thing about two halves is that you get about twice the screen real estate, as well as a measure of screen protection. Additionally, the bottom half in the hand, with the top half above, allows you to write while resting your writing hand against the holding hand, which would make input more natural (i.e., just moving the wrist, as opposed to the whole arm I envision this system using Quikwriting as the primary input method, but implemetning other software would also be necessary, especially the stuff with the input area and receiving input. I could imagine that a writing area with four general buttons would be the default, and a special call would be executed to change this, that way, only programs that need to have a different input method would have it. Other programs could act as if they were receiveing input from a keyboard.
I already have the case design in my head (at least the design for individual manufacture). The problem I've run into is the lack of suitable hardware. Some of the devices listed above are small enough, but do not include any way (that I can see) of having a configurable input area that is separate from the display screen.
I am a Mechanical and Biomedical engineer by training, so most of the aspects of specialized hardware implementation are beyond me, and I have been looking into using pre-built boards and stuffing them into a novel case (which my training does let me specialize). So I ask you, fellow/.ers, "what do you think?" Am I a deranged loon, or does an idea like this have some merit (if not, I'll still want to build one for myself)? Can anyone offer me guidance for hardware and even some software? Anyone want to help build one?
--Copyright, 2000 by WhyCause (just in case something pans out)
This also works in older versions of Communicator. You can also view the source of the page, which I saved to an HTML file that I currently use as my home page in IE5
Standing Up at 70,000 lb. Any animal has to be able to lift its own weight off the ground, i.e. stand up, with no more difficulty than Kazmaier experiences doing a 1000 lb. squat. Consider, however, what would happen to Mr. Kazmaier, were he to be scaled up to 70,000 lb., the weight commonly given for the brontosaur. Kazmaier's maximum effort at standing, fully warmed up, assuming the 1000 lb. squat, was 1340 lb. (1000 for the bar and 340 for himself). The scaled maximum lift would be a solution to:
1340/340^.667 = x/70,000^667 or 47,558 lb..
He'd not be able to lift his weight off the ground!
This makes so much sense..
bzzt, try again. This makes no sense whatsoever. What you have to remember is that things like lifting force, running speed, etc. DO NOT SCALE DIRECTLY WITH WEIGHT! More important things to keep in mind in any discussion like this are questions such as:
How far from the center of the joint are the tendon insertions (i.e., how much more mechanical advantage did the animal's muscles have when attempting to move that bulk vs. Mr. Kazmaier)?
How much larger/stronger were the animal's ligaments versus Mr. Kazmaier's (if you've ever seen the pictures of a powerlifter blowing out his knee as he attempts the stand part of the lift, you'll know exactly what I'm talking about)?
Was the cortical bone of the animal thicker (relatively) than that of a human? What about the density?
How well balanced was the animal's weight (remeber, we're constantly shifting muscle force to keep upright. Turtles have no such problems)?
The list goes on, but I'm tired of ranting. As for the questions about speed, think about a horse. It weighs a good bit more than I do, and they can outrun me, no problem (I'd best be wary, I've lain a trap for myself with my previous comments). The question comes then, why don't they run faster than they do? Answer, they don't need to. Why do we think the T. Rex had to run fast? Answer, he probably needed to in order to eat. Evolution is a way to effectively address problems in novel ways (granted, it is the brute force method, but it works). No engineer on the planet can compare in terms of the solutions mother nature has concocted, and with stuff like this masquerading as science, I am beginning to wonder if we ever will.
As to the different gravity "theories"... hmmm. I thought we lived in an enlightened age.
By The Way, IAABME (I am a Biomedical Engineer), and yes, I have looked at some of this stuff. If you're interested in some of the problems associated with drawing comparisons between different critters, I suggest looking into the field of morphometrics.
One other point not covered in this post; Compaq sells WinCE based handhelds. It could be they don't want to compete with themselves, or it could be they have signed some sort of exclusivity agreement with regards to marketing handhelds.
I don't know anything about EPOC it may be as bad, but what I know about PalmOS makes me amazed at how stable the apps under it are. And it makes me lery of wanting to try to write my own!
EPOC (as far as all the press goes) is a fully moderm, 32-bit pre-emptive multitasking, multi-threaded OS. The OS is based around a tightly coded kernel, with a window manager (EIKON in the case of my Series 5) designed to run on top of it. The way I understand it, the window manager is easily changed (more on this later), but no manager other than EIKON is currently available. The OS has inherent memory protection, and I believe all of its predecessors did as well (16-bit Psion OS's). The result is a platform that is as stable as, if not more so, than Linux. I've heard tales of individuals not restarting their Psions for a number of years, and months is very common. Even with using a large amount of home-grown software (my own included), I have never had to hit the reset button to stop a rouge app. In fact, using the built in OPL interpreter, programming on my Series 5 is easy, and I have access to almost all of the system functions through API calls. There are very few things (multi-page dialogs for one) that aren't readily available to the home programmer. The only real problem I've seen with EPOC is that if an app dies horribly enough, there may be a slight memory leak. (can you tell I'm an EPOC advocate?)
As to the description of the OS, this reminds me a good bit as to the setup used in Linux. A tight kernel (that can have support for unecessary items left out at compile time), with a windowing system on top. I think that using Linix on a handheld device can be a Very Good Thing (TM), if done properly. The trick with using any handheld is that your apps come up quickly, and you don't have to fool around with the OS much unless you really need to. The problem therefore becomes a pure UI design issue. This has been discussed elsewhere on this page, so I will rant no longer.
Lawyer-Talk is NOT good. My roommate is currently in law school, and he says the emphasis is on writing documents that anyone can read, not just hose with a J.D.
I wrote a timekeeping application for the engineering department of the company at which I co-op'ed. The project codes changed regularly, but once used, they were generally used for a few months. I used a combo box to allow the user to select from the list of previously entered project numbers, or add a new one when appropriate. I thought it to be a very effective system, and the engineers thought so as well.
The only problem is that this is the likely the new release (R6) of EPOC, termed QUARTZ. The shell looks like it's a lot different, so, unfortunately, the programs I have on my Series 5 will likely not run on this device. Too bad really, since I'm waiting on the Ericsonn QUARTZ phones before I get a cell phone.
Slashdot Mirror
...As in you can buy an election with enough money...
Nahh, Steve Forbes has proven this wrong at least twice now.
You could get Playboy in my (Catholic!) high-school library if you were 18.
You forget that Catholic schools have an overwhelming desire to raise children according to a certain moral/ethical code. The best way to affect this (according to the more enlightened) is to expose them to stimuli that are considered tasteless and/or questionable and discuss it in an elightened manner in a classroom setting. While Playboy may not be considered educational in any way, certain works of literature that the less enlightened want banned (i.e., The Chocolate War or Smilla's Sense of Snow) due to some adult material are read and discussed in classrooms so that students can learn about what goes on in the real world, and be prepared to deal with it according to the moral/ethical standard espoused by the Catholic Church. The English Literature teacher at my sisters' (Catholic, all-girl) high school said almost exactly what I just wrote (in many fewer words) when a group of parent's tried to get the aforementioned Smilla's Sense of Snow banned from the curriculum. Needless to say, their effort failed. The nice thing about private schools is that if you do not like the way things are being taught, you have every right to withdraw your children from that school. Unfortunately, not everyone has that option, but I won't get into the whole school voucher thing (which now that I think about it, would doom the private schools to the same Luddite fate as the public institutions).
I had the music (all of 'em, but especially the 'B' music, ahh, I can hear now) playing in my head within a day or two of getting my first GameBoy (the one that came bundled with Tetris ahh, for the days when consoles came bundled with a game). Couldn't get it out of my skull for weeks. Definitely the most addictive game ever.
If I were to have found a flaw in their research, I am perfectly free to report that (provided it makes it past the review process). Once their paper is published, presenting their methods and results, there ain't no way in hell they'd be able to retract that (the data maybe, not the paper). I've read articles that have glaring basic errors, and let me tell you, I pity the author of that paper. Anytime I hear the authors' names now, all I can think of is, "duh." This is the same thing that would happen to any more research coming out of Celera; it would always be mistrusted.
This is why Science is holding a copy of the data in escrow, to prevent just such an occurence. This is one thing I do trust, since Science has too good a reputation to let one company tell them how to run things. Even if that one company is one of only two major players in the field of human genome sequencing.
I'd focus on data storage first; without that, it really doesn't matter what kind of pretty GUI face you put on the front of it.
I think the concept of soups is OK for some types of applications, but overall I prefer the structured file system route. For example, on the PalmOS devices, the databases for memos, appointments, etc. are OK (who uses more than one appointment book?), especially when syncing them with desktop applications, but with productivity apps, like spreadsheets, you have to start shoe-horning files into a database structure, making the program take into account things like file/record organization (i.e., my gas milage spreadsheet goes into my personal directory, but expense accounts go into the business directory). Filesystems already do this well, and re-inventing the wheel is very rarely a good thing. As for data corruption, I think my Psion has crashed 1 time in the 2 years I've had it. The system also (I believe) does not use file system buffers, as the buffer and writing to disk take the same amount of time and memory (shared memory/file system model). If you save to CF disk, there's nothing that's killing that data short of a fire.
I posit that the interface and accessiblity of data is paramount (much more so than on the desktop), once we have established an easy way of storing and accessing the data. These are devices we use everyday, and we don't want to look like an idiot when writing down a phone number or address because it takes twice as long as finding a pen and paper. Access to the data is the key. WinCE failed in its initial incarnations for this very reason. Damn, was it clunky.
In summary, you need a fast, easy input method, with a fast, intuitive way to launch apps, and a way to view all the necessary data at the same time (color and screen use both come into play here). After all, what are these things for if not fast, easy access to the data we can't go 2 minutes without having access to? I'm curious, though, what features do people think are really necessary in these things? I for one think MP3 players and most network connectivity are overrated. I also only care about getting the files off the device and being able to use them in other applications. Syncing, for me, is just another buzzword, as I keep my phone book and appointments in only one place (on the handheld). I know others will take issue with these points, though. Flame on.
Some of these items have been covered in other posts since I started writing this tome, but I'm going to post it anyhow since they weren't covered in the detail I wanted to see.
I thought IBM wrote OS/2
MS and IBM were working on OS/2 together, at first (v1.0). Then IBM started getting wonky about things, and MS said "stuff it, we'll do our own thing." They then went and wrote NT v1.0.
A friend of mine used to work for IBM, and he was telling me some of the things about the IBM environment at the time. He said that there were feature request sheets that IBM's groups would fill out, and these features would have to be implemented, regardless of how they conflicted with other feature requests. In the end, MS just got frustrated. If I recall correctly, there is also some discussion of this in Bill Gates' The Road Ahead (I know, I know, but I was young and impressionable then! I swear I didn't finish the book!).
I use the 'about:mozilla' page source, stored as a local file, as my startup page in IE. It still makes me chuckle sometimes.
I wouldn't recommend storing the books upright (spines up.) Especially on glue-bound books, the binding gets bent and breaks after prolonged storage, especially in a humid spot
One thing not explicitly stated, storing the books spine down works really well. Nothing pulls away from the spine, and you can get any of the books out with ease (as opposed to laying them flat, when you have to unpack the whole box to get to that good one on the bottom). I used this when I moved to New Orleans, and all of my old technical manuals (ca. 1930's CRC math tables book and ca. 1950's Mark's Handbook for Mechanical Engineers) that I rescued from the Engineering Library at my Alma Mater survived unscathed.
Additionally, I used some boxes I got from a Pharmacy (my Dad is a Pharmacist) that were initially used to ship vials and their lids. They are sturdy, one piece, and about the size of a copier paper box. They also fold up easily (designed to do so), and I have kept all of them for future moves (when finished with Grad school). The additional benefit of boxes this size is that when full of books, they are still easily managed by one person. I might ask at a Pharmacy about these type of boxes.
One day, management decided to pipe some happy-happy shit^H^H^H^Hmusic through the PA system.
For a time, I worked at a power plant (Mechanical Engineer by training), and in the office the phones were hooked into a radio system of some sort, and the music played over a speaker on the front of the phone. The secretaries were in control of this, and they always had that damn thing tuned into the adult contemporary station that I swear had about ten songs that they cycled through. The worst part of it all was that while the speakers on the individual phones could be turned off, it seemed as if everyone else in the whole damn facility loved that tripe. The office was (surprizingly for being in a 1500+ MW generating station) remarkably quiet, and I could hear every God-forsaken tune over the cubical walls. I couldn't get away from it, and I couldn't control it. Thankfully, I only spent about a month (total) working in that facility. I think that has to be the single most irritating thing about any office environment; a completely dictated atmosphere. At least let me control what (if anything) I listen to.
The other problem I have is with TV's. Those things suck my brain right out of my head through my eyeballs, and I can't maintain a train of thought for more that 30 seconds in front of one. It doesn't matter what is on it. I have to turn the thing off to look at pr0n effectively, let alone try to get real work done.
It doesn't meet all of your specs, but you can run Linux on a Psion Series 5. If you can handle the larger size, the netBook (or Series 7, I can't remember which) will take a MicroDrive.
I also would love to try and "roll my own" with a LART, but I also lack the hardware skills to implement my dream machine.
I ran Doom (DOS version, 1.9) on NT4.0 just last Friday (got nostalgic). No problems whatsoever. The only way I can see a program like Virtual PC being "more realistic" is if it allows direct hardware calls, which AFAIK, WinNT did not allow from version 1.0
Another game (Lucas Arts as well) that fits in with this is 'Grim Fandango.' It also has 3D graphics, an intriguing story line, and engaging characters. It also took a different step in terms of interface in that Manny (the main character) was the interface, and anything he looked at (generally, his head would track the object until it was out of his field of view) was fair game for any of the usage options. Another great game all around.
There is a Java compiler implemented in Java in the works here.
And (according to the web page) a fully working one here.
I THINK that both of these should run in the JVM for EPOC, but I suggest checking before running out and buying a 5mx (~$500, but well worth it) strictly for Java use.
The other thing to keep in mind is the limited memory on these devices. You may need to add a CF card to hold all of your development environment. YOU HAVE BEEN WARNED!
I also had a helluva lot of fun kicking the crap out of my friends with Vader's light saber in "Masters of Teras Kasi"
Perl is available for handhelds running EPOC (i.e., the Psion Series 5/5mx/7/netBook/Revo, the Oregon Scientific Osiris, and the Ericsson MC218). The installation is LARGE for a device of this caliber (~4MB of storage and running space). You can check out the author's web page for the port here.
.OPX libraries.
Alternatively, most of the machines I listed (with the exception of the Revo and possibly the Osiris) come with a programming language editor and byte-code compiler built in. The language is called OPL32 (organizer programming language 32-bit). OPL is a relatively powerful language (even if it is interpreted), with the capability to expand (i.e., graphical printing functions, multi-page dialogs) via
The newer machines (NOT the Series 5, and maybe not the Revo or Osiris) can also run Java. Which may also meet your needs.
If you would like to learn more about programming the EPOC OS, you can download OPL and Java SDK's, and the included emulator, from the Symbian site (you have to register to download, but it's free. It does run only under Windows, though). One additional note, the Perl port will likely NOT run under the emulator (compiled OPL will run under both, C++, will not). You have been warned.
CLIs aren't dying, they're just mutating.
I agree. One of the things I actually like about Win95/98 is that I can use the keyboard to start about anything. In my case, WinKey (or Ctrl-Esc)-p-o-w starts up Word [1]. Replace the w with an e, and I'm running Excel. Discounting start-up time, I estimate it takes me at most a quarter the time of a mouse user to start my Office apps.
There is also the WinKey-r series for a 'shell prompt' of sorts. I never look through the accessories menu for the tiny calculator icon, instead I hit WinKey-r and type calc. I personally think that these two features work really, really well. When I ran Win3.1, I had a command prompt open on startup for some tasks. I've found that in Win95/98, I don't always need that prompt there (even with the added integration). Try and take the prompt away from me, though (damn Macs) and you'll lose a hand.
[1] In case you aren't familiar with the feature, the Start menu can be set up to use keystrokes easily. If only one item on the current menu starts with the letter you press, it is selected and opened/executed; more than one and the first, then next is selected with subsequent key presses, and enter opening/executing. Right and left keys open and close menus, and up/down navigating.
Finally, this discussion comes to light without me having to ask it. I've been thinking about trying this for about 2 or 3 months, the problem I have is that my design seems rather novel, and I havn't seen any of the small boards/systems (like LART, TINI, TIQIT, uCLinux) that would support it hardware-wise, let alone getting the software working for it. Most of the following rant is nothing more than a pipe dream, but if you know of something to make this dream a reality, by God post it.
/.ers, "what do you think?" Am I a deranged loon, or does an idea like this have some merit (if not, I'll still want to build one for myself)? Can anyone offer me guidance for hardware and even some software? Anyone want to help build one?
I envision a bifold device, about the size of a deck of cards, with rounded edges. The device flips open (either spring loaded to a set angle ala Psion, or with a friction clutch holding it at any angle), and both inside faces are LCDs with digitizers on them. The side that you hold in your hand is the input area,which can be changed to fit the application (for example, all buttons for a calculator app, or a few buttons with a writing area for a notepad app). The top half is the display area.
Two displays/digitizers may seem like a waste, but an infinitely configurable input area seems like it might be just the kind of thing to make this badboy very easy and fast to use. The other nice thing about two halves is that you get about twice the screen real estate, as well as a measure of screen protection. Additionally, the bottom half in the hand, with the top half above, allows you to write while resting your writing hand against the holding hand, which would make input more natural (i.e., just moving the wrist, as opposed to the whole arm I envision this system using Quikwriting as the primary input method, but implemetning other software would also be necessary, especially the stuff with the input area and receiving input. I could imagine that a writing area with four general buttons would be the default, and a special call would be executed to change this, that way, only programs that need to have a different input method would have it. Other programs could act as if they were receiveing input from a keyboard.
I already have the case design in my head (at least the design for individual manufacture). The problem I've run into is the lack of suitable hardware. Some of the devices listed above are small enough, but do not include any way (that I can see) of having a configurable input area that is separate from the display screen.
I am a Mechanical and Biomedical engineer by training, so most of the aspects of specialized hardware implementation are beyond me, and I have been looking into using pre-built boards and stuffing them into a novel case (which my training does let me specialize). So I ask you, fellow
--Copyright, 2000 by WhyCause (just in case something pans out)
This also works in older versions of Communicator. You can also view the source of the page, which I saved to an HTML file that I currently use as my home page in IE5
This makes so much sense..
bzzt, try again. This makes no sense whatsoever. What you have to remember is that things like lifting force, running speed, etc. DO NOT SCALE DIRECTLY WITH WEIGHT! More important things to keep in mind in any discussion like this are questions such as:
How far from the center of the joint are the tendon insertions (i.e., how much more mechanical advantage did the animal's muscles have when attempting to move that bulk vs. Mr. Kazmaier)?
How much larger/stronger were the animal's ligaments versus Mr. Kazmaier's (if you've ever seen the pictures of a powerlifter blowing out his knee as he attempts the stand part of the lift, you'll know exactly what I'm talking about)?
Was the cortical bone of the animal thicker (relatively) than that of a human? What about the density?
How well balanced was the animal's weight (remeber, we're constantly shifting muscle force to keep upright. Turtles have no such problems)?
The list goes on, but I'm tired of ranting. As for the questions about speed, think about a horse. It weighs a good bit more than I do, and they can outrun me, no problem (I'd best be wary, I've lain a trap for myself with my previous comments). The question comes then, why don't they run faster than they do? Answer, they don't need to. Why do we think the T. Rex had to run fast? Answer, he probably needed to in order to eat. Evolution is a way to effectively address problems in novel ways (granted, it is the brute force method, but it works). No engineer on the planet can compare in terms of the solutions mother nature has concocted, and with stuff like this masquerading as science, I am beginning to wonder if we ever will.
As to the different gravity "theories"... hmmm. I thought we lived in an enlightened age.
By The Way, IAABME (I am a Biomedical Engineer), and yes, I have looked at some of this stuff. If you're interested in some of the problems associated with drawing comparisons between different critters, I suggest looking into the field of morphometrics.
One other point not covered in this post; Compaq sells WinCE based handhelds. It could be they don't want to compete with themselves, or it could be they have signed some sort of exclusivity agreement with regards to marketing handhelds.
I don't know anything about EPOC it may be as bad, but what I know about PalmOS makes me amazed at how stable the apps under it are. And it makes me lery of wanting to try to write my own!
EPOC (as far as all the press goes) is a fully moderm, 32-bit pre-emptive multitasking, multi-threaded OS. The OS is based around a tightly coded kernel, with a window manager (EIKON in the case of my Series 5) designed to run on top of it. The way I understand it, the window manager is easily changed (more on this later), but no manager other than EIKON is currently available. The OS has inherent memory protection, and I believe all of its predecessors did as well (16-bit Psion OS's). The result is a platform that is as stable as, if not more so, than Linux. I've heard tales of individuals not restarting their Psions for a number of years, and months is very common. Even with using a large amount of home-grown software (my own included), I have never had to hit the reset button to stop a rouge app. In fact, using the built in OPL interpreter, programming on my Series 5 is easy, and I have access to almost all of the system functions through API calls. There are very few things (multi-page dialogs for one) that aren't readily available to the home programmer. The only real problem I've seen with EPOC is that if an app dies horribly enough, there may be a slight memory leak. (can you tell I'm an EPOC advocate?)
As to the description of the OS, this reminds me a good bit as to the setup used in Linux. A tight kernel (that can have support for unecessary items left out at compile time), with a windowing system on top. I think that using Linix on a handheld device can be a Very Good Thing (TM), if done properly. The trick with using any handheld is that your apps come up quickly, and you don't have to fool around with the OS much unless you really need to. The problem therefore becomes a pure UI design issue. This has been discussed elsewhere on this page, so I will rant no longer.
You can check out more about the EPOC OS at: http://www.symbian.com
Lawyer-Talk is NOT good. My roommate is currently in law school, and he says the emphasis is on writing documents that anyone can read, not just hose with a J.D.
I wrote a timekeeping application for the engineering department of the company at which I co-op'ed. The project codes changed regularly, but once used, they were generally used for a few months. I used a combo box to allow the user to select from the list of previously entered project numbers, or add a new one when appropriate. I thought it to be a very effective system, and the engineers thought so as well.
The only problem is that this is the likely the new release (R6) of EPOC, termed QUARTZ. The shell looks like it's a lot different, so, unfortunately, the programs I have on my Series 5 will likely not run on this device. Too bad really, since I'm waiting on the Ericsonn QUARTZ phones before I get a cell phone.