I think that a single menu bar makes sense if you have a small screen. It saves screen real-estate and you will always know where to find the menu bar.
But.. for a larger screen (17" and up), I think it makes more sense to have the menu bar attached to the window in question so you don't have to move the pointer so far.
But then.. many people maximize all their windows anyway, removing the benefits of either approach.
2. I'm, sorry but I don't think you have a point here. In Linux, you have to open() a file before you can mmap() it. Once you have a file descriptor, you can call fsync() on it any time you want to. I don't see why an OS with orthogonal persistence would have to be different.
3. I think capabilities is the best tool to use to solve this problem, because it is fine-grained. See 4 below.
4. I agree that it is harder to get a view of the access matrix with capabilities than with the POSIX model. I believe there are two primary reasons for this: first because people are afraid of capabilities simply because they are unfamiliar to them, second because it is not too common for capability systems to have a good way to revoke capabilities that no longer have to be alive, and this causes more uncertainty. However, I think mcrbids and Peaker are looking at capabilities the wrong way as a model, when they should look at capabilities as primitives with which to build security models. In fact, you could build the POSIX model using capabilities as building blocks. But you also get the choice to create other models.
I believe that for DRM to "work" (on all so many levels), then the key will be to implement a metafor of a physical object. Let's imagine that there is a virtual world parallel to ours, in which these digital objects (DRM'ed files) live, and that your PC and your portable media players are only interfaces into that world. If you own a record in the real world, then you can play it, skip tracks, move it, sell it, lend it to your friends etc.. Similiarly, if you own a record in the virtual world, then you should be to play it, skip tracks, move it, sell it, lend it your friends etc. in the virtual world - using your PC and/or portable media player. I think this is a model that both the industry and the public could agree on using.
Another problem is the volatility of digital objects. We are so used to our apps crashing, files getting lost etc. that we do not value digital files as much as physical objects. If something is important then we want hardcopy! People are ready to pay for hardcopy, but not for digital files. DRM'd files need to feel safe, and be safe.
For years I used a pirated version of TVPaint for Windows. It had the DPaint look and feel but worked in 24bit RGB and had optional antialiasing. It later changed name to Aura and became more about animation, more for professional video work.
The thing I liked most about DPaint and TVPaint was that you had a large canvas, and you could pick up, replicate, move around, load and save parts of it easily - which was great for painting icons that had many similiar items. Another simple UI thing was that you erased with the right mouse button, using the current brush.
I agree about GTK+. Something changed in GTK+ 1.3 way back when. Betatesters complained about sluggishness back then, were promised that things would improve when the kinks were straighted out, but nothing changed. Eventually people got faster computers, and experienced the problem less.
If you don't believe me, get a 200 MHz computer, run a GTK+ 1.2 app from back then (e.g. X-chat), then run the same app that has been ported to GTK+ 2.0. You [i]will[/i] notice a difference!
About MacOS X, I wouldn't be surprised if some delays are [i]features[/i] in the OS, to emulate the feel of a 680x0 Mac from '85... I bought a Mac last year, but sold it not long afterwards. I just could not stand it, even after having tweaked obscure values in plists for maximum responsiveness. For instance, the mouse acceleration is strange and needs time to get used to. You have to get a new mouse that installs its own driver/subsystem if you want it to behave like in Windows or X.
I don't see that any of the concepts that Singularity is built on are particularly innovative. The idea of using a safe language to replace memory protection is decades old (JavaOS anyone?). There has also been a lot of research into code verification so there is a lot to build on there. There are a huge lot of research systems that have various ways of synchronizing processes in a safe manner - both languages and OSes. etc. etc.
What is special about Singularity is the combination of several of these concepts into a whole. It could be interesting to see how the different technologies would fit together. What I dislike about Singularity (apart being from Micro$oft) is its rigidity. In a few ways, it is very limiting. To be successful, it must be able to evolve to new (better) technologies. An example of that is memory management, where programming languages research today is going back to explicit deallocation - now controlled by the compiler. That is not possible in Singularity.
I don't see why an open-source operating systems using the same concepts couldn't be just as good, or better than Singularity. It probably would be.
Seed banks are nothing new. There have been seed banks in every western country for at least a hundred years, and there still are, so there is a lot of redundancy right there. This seed bank is a little bit bigger, that's the only difference.
Seed banks need to replenish their stocks every few years by planting, because seeds will eventually degenerate even when stored in sub-zero temperatures.
I anticipate the text would look OK on many Windows systems with their default DPI, text size settings etc., but it too tiny on my system, and I usually like small text. Also, on my system, the Times font is not antialiased (i.e. blurred) as Arial/Helvetica is, which further adds readability.
I propose that the body text should be the default text size so that we could all customize it. We are all power-users here, after all.
I really like this proposal overall. It is clean. I like the ideas of using the bottom curve to frame the story, how quoted text is indicated and the look of sidebars.
An LCD display that needs power only when the image is changes is old. Check out Kent Displays. It was spun off in 1993 from a project at Kent State University. I wonder why it never took off.
According to this article the motion sensor is actually a three-axis accelerometer.
I doubt that George Lucas would sue, but I know of a guy who claims that he has a US patent on the use of an accelerometer in a lightsaber toy to trigger sound events.
I suggested the use of an accelerometer on an Internet messageboard, when another user had asked for advice for building a high-end lightsaber toy. I directly got snotty remarks that it would be infringing on this guy's patent. He was then selling circuits with an accelerometer that were connected via radio-link to a PC that played sounds.
I then told him that I was building such a circuit myself and knew of other people who also had designed similiar circuits independently from eachother, because it was quite straightforward design if you know your way around electronics. I pointed out that I thought that such straight-forward patents were stealing from the community, especially if this was a software patent but that it wouldn't apply to me anyway because he was in the US and I was in Europe. For this I was banned from the msgboard, for "software piracy" (!).. apparently the admins did not distinguish between different types of intellectual property, or they were friends with this guy. I don't know.
For larger web sites I separate the code into a Presentation Layer and a Business Layer. The Presentation Layer is simply PHP/JSP/whatever-code embedded in HTML. The Business Layer does not contain any markup at all, being mostly a wrapper around database calls but with an API that fits the application. Could be written in another language. For instance, for a web site that requires user accounts, User objects are owned by the Business Layer.
Then let the Presentation Layer have a common library for common markup actions, where some of the look of the web site is defined. Ifever you have misdesigned the system so that the Business Layer has to do markup anyway, it should use functions from that library to do it.
I actually sold my Mac Mini because I did not like the OS, to build a new Linux/x86 computer out of almost precisely the same components as in the article, except for a few things: What I liked the most about the Mini was that it was 1) small, 2) stylish and 3) quiet.
1) is almost impossible with PC components, if you want any reasonable performance. I will be choosing a MATX board, but I am trying to find one that is smaller than 24.4×22 cm and which still has Firewire and DVI/TV out like the Mac Mini. The PSU will be external like for the Mini. (btw. Mac Mini's PSU gets very hot.. and it does not have any sink drain either)
2) I am solving by building the case myself. I am no stranger to working with aluminum.
3) I am trying to solve by replacing the internal heat sinks and chassi fan(s) with one large heat sink that will become one side of the case. Heat transfer from each component to the heatsink will be through copper blocks, but some heatpipe-like features in these blocks would be better. Of course, this decision requires that all i/o is integrated... But there are still drawbacks compared to the Mini, while costing as much: - still twice as big as the Mac Mini, - the Mac Mini has dedicated graphics memory while the integrated PC boards share memory between CPU and GPU with reduced performance. - the Mac has a slot-in DVD. I can not find any reasonably-priced slot-in DVD burner anywhere close... (I live in Sweden, btw)
I have also done that. It is fun and is a good deed. Because you are holding them back from harassing others, you are doing a service to your local community.
I wish... It has been quite difficult to find a new 3-button mouse lately, that does not have a scrollwheel instead of a real button.
btw: * Safari support the third mouse button! It opens a link in a new window. * Apple's mouse driver sucks. Acceleration is very weird. Open source though, as it is part of Darwin.. but I have found only one replacement, which also has weird behaviour. Logitech and Microsoft override Apple's driver with their own.
How about tracking retina motion and moving the cursor. This would be great for advertising too Advertizing!?... I really hope not!
What you are suggesting is not interfaces, really. I think all of the suggested features need good interaction with the user to weed out errors or they will be too annoying to use - and there you are back to the user interface again.
I think that we do not need smart interfaces as much as smart programmers that make interaction simple and intuitive for the users.
Slashdot Mirror
I think that a single menu bar makes sense if you have a small screen. It saves screen real-estate and you will always know where to find the menu bar.
But.. for a larger screen (17" and up), I think it makes more sense to have the menu bar attached to the window in question so you don't have to move the pointer so far.
But then.. many people maximize all their windows anyway, removing the benefits of either approach.
Oh yeah.. I remember a third-party accessory for the "puck" mouse: An external plastic shell that you clipped onto the puck mouse to make it oval.
2. I'm, sorry but I don't think you have a point here. In Linux, you have to open() a file before you can mmap() it. Once you have a file descriptor, you can call fsync() on it any time you want to. I don't see why an OS with orthogonal persistence would have to be different.
3. I think capabilities is the best tool to use to solve this problem, because it is fine-grained. See 4 below.
4. I agree that it is harder to get a view of the access matrix with capabilities than with the POSIX model.
I believe there are two primary reasons for this: first because people are afraid of capabilities simply because they are unfamiliar to them, second because it is not too common for capability systems to have a good way to revoke capabilities that no longer have to be alive, and this causes more uncertainty.
However, I think mcrbids and Peaker are looking at capabilities the wrong way as a model, when they should look at capabilities as primitives with which to build security models. In fact, you could build the POSIX model using capabilities as building blocks. But you also get the choice to create other models.
I believe that for DRM to "work" (on all so many levels), then the key will be to implement a metafor of a physical object.
Let's imagine that there is a virtual world parallel to ours, in which these digital objects (DRM'ed files) live, and that your PC and your portable media players are only interfaces into that world.
If you own a record in the real world, then you can play it, skip tracks, move it, sell it, lend it to your friends etc..
Similiarly, if you own a record in the virtual world, then you should be to play it, skip tracks, move it, sell it, lend it your friends etc. in the virtual world - using your PC and/or portable media player.
I think this is a model that both the industry and the public could agree on using.
Another problem is the volatility of digital objects. We are so used to our apps crashing, files getting lost etc. that we do not value digital files as much as physical objects. If something is important then we want hardcopy! People are ready to pay for hardcopy, but not for digital files. DRM'd files need to feel safe, and be safe.
Sadly, this seems to be the norm, even after you have been to an interview.
You would think it would be common courtesy to do so, but nooooo.
You can change the playback speed of CD's on a lot of DJ gear. Some even have mock turntables.
For years I used a pirated version of TVPaint for Windows. It had the DPaint look and feel but worked in 24bit RGB and had optional antialiasing. It later changed name to Aura and became more about animation, more for professional video work.
The thing I liked most about DPaint and TVPaint was that you had a large canvas, and you could pick up, replicate, move around, load and save parts of it easily - which was great for painting icons that had many similiar items.
Another simple UI thing was that you erased with the right mouse button, using the current brush.
I agree about GTK+. Something changed in GTK+ 1.3 way back when. Betatesters complained about sluggishness back then, were promised that things would improve when the kinks were straighted out, but nothing changed. Eventually people got faster computers, and experienced the problem less.
...
If you don't believe me, get a 200 MHz computer, run a GTK+ 1.2 app from back then (e.g. X-chat), then run the same app that has been ported to GTK+ 2.0. You [i]will[/i] notice a difference!
About MacOS X, I wouldn't be surprised if some delays are [i]features[/i] in the OS, to emulate the feel of a 680x0 Mac from '85
I bought a Mac last year, but sold it not long afterwards. I just could not stand it, even after having tweaked obscure values in plists for maximum responsiveness. For instance, the mouse acceleration is strange and needs time to get used to. You have to get a new mouse that installs its own driver/subsystem if you want it to behave like in Windows or X.
There is not much to see here, no.
The mockups were made by fans of Apple, and Apple fans love flashy new features Features FEATURES! That is what you can find in these mockups.
If you are looking for good UI design, look elsewhere.
Anyone looking to improve MacOS X should look here first: http://www.asktog.com/ .
I don't see that any of the concepts that Singularity is built on are particularly innovative. The idea of using a safe language to replace memory protection is decades old (JavaOS anyone?). There has also been a lot of research into code verification so there is a lot to build on there. There are a huge lot of research systems that have various ways of synchronizing processes in a safe manner - both languages and OSes. etc. etc.
What is special about Singularity is the combination of several of these concepts into a whole. It could be interesting to see how the different technologies would fit together.
What I dislike about Singularity (apart being from Micro$oft) is its rigidity. In a few ways, it is very limiting. To be successful, it must be able to evolve to new (better) technologies.
An example of that is memory management, where programming languages research today is going back to explicit deallocation - now controlled by the compiler. That is not possible in Singularity.
I don't see why an open-source operating systems using the same concepts couldn't be just as good, or better than Singularity. It probably would be.
Seed banks are nothing new. There have been seed banks in every western country for at least a hundred years, and there still are, so there is a lot of redundancy right there.
This seed bank is a little bit bigger, that's the only difference.
Seed banks need to replenish their stocks every few years by planting, because seeds will eventually degenerate even when stored in sub-zero temperatures.
... as much as "Can I hack it or not?".
I anticipate the text would look OK on many Windows systems with their default DPI, text size settings etc., but it too tiny on my system, and I usually like small text.
Also, on my system, the Times font is not antialiased (i.e. blurred) as Arial/Helvetica is, which further adds readability.
I propose that the body text should be the default text size so that we could all customize it. We are all power-users here, after all.
I really like this proposal overall. It is clean. I like the ideas of using the bottom curve to frame the story, how quoted text is indicated and the look of sidebars.
An LCD display that needs power only when the image is changes is old.
Check out Kent Displays. It was spun off in 1993 from a project at Kent State University.
I wonder why it never took off.
I suggested the use of an accelerometer on an Internet messageboard, when another user had asked for advice for building a high-end lightsaber toy. I directly got snotty remarks that it would be infringing on this guy's patent. He was then selling circuits with an accelerometer that were connected via radio-link to a PC that played sounds.
I then told him that I was building such a circuit myself and knew of other people who also had designed similiar circuits independently from eachother, because it was quite straightforward design if you know your way around electronics. I pointed out that I thought that such straight-forward patents were stealing from the community, especially if this was a software patent but that it wouldn't apply to me anyway because he was in the US and I was in Europe. For this I was banned from the msgboard, for "software piracy" (!) .. apparently the admins did not distinguish between different types of intellectual property, or they were friends with this guy. I don't know.
For larger web sites I separate the code into a Presentation Layer and a Business Layer.
The Presentation Layer is simply PHP/JSP/whatever-code embedded in HTML. The Business Layer does not contain any markup at all, being mostly a wrapper around database calls but with an API that fits the application. Could be written in another language. For instance, for a web site that requires user accounts, User objects are owned by the Business Layer.
Then let the Presentation Layer have a common library for common markup actions, where some of the look of the web site is defined.
Ifever you have misdesigned the system so that the Business Layer has to do markup anyway, it should use functions from that library to do it.
Yeah, but the UK is already a police state.
I don't use tabs. I open lots of multiple windows that I "shade" in WindowMaker, so I have to wish for an "undo close window" option.
Window shading has the feature that there is always enough space for the whole title. I can also group pages/windows together.
I actually sold my Mac Mini because I did not like the OS, to build a new Linux/x86 computer out of almost precisely the same components as in the article, except for a few things: What I liked the most about the Mini was that it was 1) small, 2) stylish and 3) quiet.
..
1) is almost impossible with PC components, if you want any reasonable performance. I will be choosing a MATX board, but I am trying to find one that is smaller than 24.4×22 cm and which still has Firewire and DVI/TV out like the Mac Mini.
The PSU will be external like for the Mini. (btw. Mac Mini's PSU gets very hot.. and it does not have any sink drain either)
2) I am solving by building the case myself. I am no stranger to working with aluminum.
3) I am trying to solve by replacing the internal heat sinks and chassi fan(s) with one large heat sink that will become one side of the case. Heat transfer from each component to the heatsink will be through copper blocks, but some heatpipe-like features in these blocks would be better.
Of course, this decision requires that all i/o is integrated.
But there are still drawbacks compared to the Mini, while costing as much:
- still twice as big as the Mac Mini,
- the Mac Mini has dedicated graphics memory while the integrated PC boards share memory between CPU and GPU with reduced performance.
- the Mac has a slot-in DVD. I can not find any reasonably-priced slot-in DVD burner anywhere close... (I live in Sweden, btw)
I dunno about a film, but I think it could be made into a kick-ass miniseries.
There are too many things going on during a time span of a few hundred years. You can not squeeze that into a two-hour movie.
I have also done that. It is fun and is a good deed. Because you are holding them back from harassing others, you are doing a service to your local community.
I wish ...
It has been quite difficult to find a new 3-button mouse lately, that does not have a scrollwheel instead of a real button.
btw:
* Safari support the third mouse button! It opens a link in a new window.
* Apple's mouse driver sucks. Acceleration is very weird. Open source though, as it is part of Darwin.. but I have found only one replacement, which also has weird behaviour. Logitech and Microsoft override Apple's driver with their own.
Being a Star Wars fan, I have seen much better home-made R2-D2's than that. ;)
Google for "R2 Builder's Club" and be even more amazed.
You don't want to know.. ;)
Advertizing!?
What you are suggesting is not interfaces, really. I think all of the suggested features need good interaction with the user to weed out errors or they will be too annoying to use - and there you are back to the user interface again.
I think that we do not need smart interfaces as much as smart programmers that make interaction simple and intuitive for the users.