Signing makes it possible to have some level of security and some level openness. So at least Apple might be able to do *the right thing* eventually. Then again perhaps not.
Personally, I want see people using their phones on the train in the morning for something other than calls. The only people I currently see doing that are ****berry owners. I think that, given the sophistication of the devices, they are being grossly underutilised. Why? Well, simply because the features aren't good enough for the cost and trouble of using them. This is partly because phones have a very short sales life. The next model and the next model must come out to satisfy fashion. None can ever be perfected - the struggle is to keep everything working on the new hardware and to support the new "features." So there isn't enough time to spend on perfecting the user experience of phone X.
Using the same OS and UI on many phones does help because at least most of the improvements get propagated onwards but with new hardware and form factors turning up all the time there will always be problems.
When the majority of people have a smartphone this will all change if they discover one or two critically useful applications. This will introduce a slowdown where one will want to be sure that one's favoured software works on the next device. Currently it's the telephony parts of phones that users consider critical but as other things take off (e.g. you need flash and a browser to work properly for your mobile **space access) it won't be so easy to change radically all the time and a degree of stability will be created just because the large number of users will *need* it.
Basically the user has to approve anything that might cost them money so an EXE or DLL that wants internet access will trigger the installer to ask the user for permission. Programs can't change their capabilities once installed, and the data of one app can be stored such that others are not able to see or access it. Various bits of the filesystem are also inaccessible to insufficiently privileged apps etc etc. So apps are insulated from each other and the OS is shielded from apps. This makes viruses much harder to propagate. Software must be installed to work and only the system installer has the privileges to do that - and it asks the phone user and checks signing.
If some software was found to be "bad" even after testing then it's certificate could be revoked, although I'm not sure if much infrastructure has been devoted to this possible course of action.
The best a virus writer might hope for is to write something in an application's own macro language e.g. javascript in the browser. They might be able to infect the application and cause trouble but at least the damage would be limited to the capabilities of that application. So a virused browser wouldn't be able to cause a Denial of Service on the Mobile Network by doing something horrible with the phone's radio. This is still pretty cool.
A lot of effort has been put into reducing the trouble and bother of platsec and into helping developers and you can read about it on the link. e.g. there's a thing called a devcert that lets you sign your own apps but these will only install on a couple of specific phones. This allows a developer to test software on their own phone(s) up to the point of release.
My very personal opinion is that there are other problems with the structure of the market that, once solved, will set loose a flood of applications. I would probably be unwise to say anything more on that though.
Disclaimer: I work for Symbian but everything I write is my personal opinion only.
Brilliant - mod up please. This kind of control is already present on all Symbian 9.x phones (e.g. Series 60 3rd ed) -in the form of "Platsec" or Platform Security. Apps have to be signed for the "capabilities" they use. The simple-to-understand capabilities can be granted by the phone user. The more "dangerous" capabilities have to be granted by getting an external test house, with appropriate authority, to sign the application. If you want to access deep parts of Nokia's or Sony Ericsson's GSM telephony stacks the you need to get a signature from them.
It has sparked lots of heated debates because the signing costs money but it offers protection against the kinds of problems that PCs have had with the internet. Without it, phones would be a very attractive ecosystem for malicious software.
Apple seem not to have this kind of architecture in place yet so their only option is to be more restrictive for now.
I agree that the current GUI paradigm is old. It presents a model of the computer and the world. Users need to know this model to be able to plan what they want to achieve.
I have noticed that people seem to start with a very simple mental model and then when they notice some inadequacy they expand it to cope with the new needs. To begin with one doesn't tell one's granny that she can run out of disc space when "saving files" - because it's too much detail and unlikely to happen immediately on her new computer. For a beginning she can imagine her 100GB hard disc to be infinite in size.
This is just like learning to drive - you don't get into a car for the first time and drive it down the M25 (subtitute name of favorite big motorway/highway/autobahn/autostrada).
I think that the problem with computer UIs is that they have no understanding of the person who is using them so they don't know how much detail to offer them today and how much more to offer tomorrow. They aren't really conversing with the user so they can't become truly effective for that particular user.
I think the English name might be "Flintlock Pistol" which refers to the firing mechanism - a spring-loaded chip of flint (a stone commonly used to make sparks) which is released by pulling the trigger.
. . . just to clarify. In Japan, for example, Symbian is sold with the MOAP environment.
I am a Symbian employee though what I say is purely my own opinion: The fact that firmware updates are freely available for the newer Nokia phones is a fairly new phenomenon and has a major effect. Problems are being dealt with swiftly.
Symbian is a much much more powerful operating system than it's competitors and it's really quite sophisticated inside. The pressure on us employees to get things right simply gets more and more tremendous. We don't do everything - not device drivers for example because the hardware isn't ours, after all - but the product delivered to our customers (the phone manufacturers) is very good and enables them to deliver amazing products like the N95 which will change ways of living. A shitload of kewl work has been put into v9 for your pleasure and it has bedded down so you will see it wipe the floor with it's opponents.
It does static analysis. It is a C++ compiler that turns your code into a form that can be easily analyed by scripts that look for particular faults. It has a low rate of false positives and it is able to find bugs even in very unusual code paths which are very hard to reproduce. This means that it is much less likely that one's customers will experience faults that you cannot reproduce (these are very difficult and expensive to fix). It should also mean that your software will pass QA at the first or second try and thus you can get it to market earlier.
The last time I checked, the fee was about 250K for a license to use it on your source base of 100K lines of code. Coverity is a very small company and it's software is particularly useful to people working on fairly large projects so it makes sense for them to charge a stiff fee.
When I think of some of the bugs that it would have found and total up the time of all the test, support, and software engineers on product X that I was working on at company Z a couple of years ago, I think that this could easily have paid for itself on one project and left us with a much more reputable product.
Splice is the open source equivalent - perhaps you should "call" for that. I wonder who has to actually do the work to satisfy you?
OK, the opinions here are my own personal ones and in no way represent any other person or entity. It's also not guaranteed to be accurate.
I have seen the issue from both sides and I basically agree that getting in the way of developers is very bad. I think that the issue will become much less important thanks to the "firming up" of some APIs and the fact that manufacturers support them better.
The certificate stuff is pretty much an answer for the security argument against releasing APIs. It will stop a legion of trojans and worms from wriggling through GPRS/3G/bluetooth/wifi connections and infesting the world's mobile phone "fleet". The only alternative is to lock down phones completely and that is what might happen to the other OSes if something similar isn't done for them. I think that this is actually critical to enabling third-party software and quite positive for everyone. Symbian have put in a huge effort to make it friendly so don't be disheartened - read on to find out why it'll be ok.
The first thing to note is that you can get a free certificate (DevCert) that gives you very "deep" access but it's locked to your personal phone. This means that you can develop whatever you like on your own device.
If your application uses the 60% of the API which is considered harmless then it can run without signing.
There are a number of capabilities which are considered to be simple enough that the user can understand them and grant them during installation. So these also don't require signing.
If you use some of the extended capabilities or Phone Manufacturer Approved capabilities and you wish to make your software generally available on all phones then you'll need a certificate.
If your software is free (i.e. you make no financial gain from it) then Symbian has made free testing and free certificates available. If your software is not free then you can probably afford a "couple" of hundred dollars for it to be signed.
I am an "interested party" but I still think I can clarify this somewhat.
Symbian has a big API which is understandable given that it covers multimedia, UIs and a lot of different types of networking. The documentation for almost all of that is freely available as are the SDKs.
One has to pay to get a copy of the source code, however, and access to the more volatile APIs that aren't evenly supported by all devices or that might pose some degree of security risk. This tends to include some of the telephony stuff. The source code allows you to understand what's happening ia very specific way although it doesn't include the drivers written by each manufacturer (Symbian doesn't own them after all).
It was much harder to get the equivalent API details for WinMob - in fact in some instances impossible. Good luck getting the same details out of Motorola!
Symbian made it's first ever profit in Q1 and according to it's website it makes around $2.5-3 per phone. That doesn't seem too evil.
If a facility is useful, a very large proportion of people seem to not give a stuff about security e.g. Outlook.
I use different computers at work, at home on holiday, when I am out and about. They are all different systems and I don't own them myself necessarily. I have a powerful PDA phone but I don't always carry it with me and it has limited capacity anyhow.
The more "devices" that we end up using, the more desirable is will become to be able access one's information without having to be involved in the mechanics of how it moves around. I think so because I appreciate simplicity and I think that other people do too unless they are hooked on "messing around with gadgets."
I find Gmail useful because I can get to all my messages from anywhere and I don't need to stuff my pockets with devices or manage the disc space on them them or find a charger that works in country X or put up with the sometimes rubbishy software.
Gmail is like a worldwide clipboard too - you can be anywhere and put new links, travel details, addresses and phone numbers etc into it (a draft message) and can search through later on to find something - much better than post-it notes.
So - good for Paul Maritz and his efforts. I had better declare 2 reasons for bias, though: 1) He was born in Zimbabwe like me although he was brought up in South Africa. 2) He has flown a Hawker Hunter at Thunder City (a place in Cape Town where you can fly old British combat jets) and in some odd way I admire him for doing what I would like to do. I want to fly in the awesome BAC Lightning, though. http://www.thundercity.com/tiger_paul.htm
No. Symbian costs about $2.50 per phone - one reason why Symbian have only recently turned a profit.
Symbian is pretty open from a developer's point of view and bigger partners can get access to the source code for the kernel and servers (without the GUI or all the device drivers because these belong to the manufacturers).
From what I can see this openness is not entirely pleasing to mobile phone operators whose thirst for "Average Revenue per User" or ARPU is unquenchable - they would rather like to monopolise their customers.
I'm from Zimbabwe and of Irish descent and I know what it is to be put down and joked about - it's something that everyone tries to do to everyone and based on many things and not only nationality.
It is much much harder to not ever put anyone else down than it is to complain when people do it to you. i.e. people always notice insults to themselves and ignore their own insults to others.
If you can honestly get rid of your prejudgement of Americans then you have some right to ask for the same lack of prejudgement of Welsh people.
Thus spake the Assembler programmers of the first C compilers . . . . . . and the C programmers of the C++ compilers . . . . . and the C++ programmers of Java . . .
It would be interesting to know if ALP can support single core CPUs - where the same CPU runs the GSM/CDMA stack and all the applications as well. This is going to be a critical factor in the total cost of the operating system + handset and is an area where Symbian has a long head start over Windows Mobile.
Symbian was sold on 10.9 million handsets in the last quarter - making it an attractive target for developers such as myself. I am not denying Palm's product could be awesome but it seems to have a hell of a long way to catch up.
I had to find a bug that caused a crash approximately every 2 weeks. It was terrible to debug because we couldn't replicate it initially. I eventually wrote a system test that shunted about 2 weeks data through the system in an hour but using random rates and delays and I simulated comms failures by killing the test process and restarting randomly. I found the bug in a comms error handling routine using purify (you could use valgrind). It was a double-free that corrupted the stack and caused a crash somewhere else later on, sometimes. The customer account was saved.
The lesson, to me, is to write a test program which will simulate heavy but randomised use of your system and to then to use this against the system whilst running it in a memory debugger.
In addition, I would suggest that you don't really need to split the program into several processes but would probably benefit from restarting the single process every day at some low usage period or by starting a new process and handing off to it from the old one. This idea always invites derision from my colleagues but they do make just as many mistakes as me and most of these mistakes wouldn't result in costly support time and patch creation if the software was restarted often.
The usual theory is that mobile IP will be the nucleus of IPv6 and may actually end up being larger in terms of the number of devices than the rest of the internet since lots of people have phones but not necessarily PCs.
Even if there were less phones, there would still be an incentive for business to IPv6-enable their software and hardware if there was a large market of mobile users to supply with servies.
Lots of people feel angry about failing to make a free product work but that's the way it is. Your kind of complaints have been made about every part of RedHat and then Fedora. First the installer, then the desktop then the applications . . . . they get fixed and people move on to complain bitterly about something else.
Of course . . they would never put any effort in themselves . ..no . . . someone in the Fedora community must do it for free.
Re:Let me know when it stops sucking
on
GCC 4.1 Released
·
· Score: 1
As you know, it hasn't actually been released yet to 4.1 is not ready. The most recent round of dramatic changes were from 3.4 to 4.0. 4.02 seems very good and I haven't seen a Fedora development update for it for quite a long time.
4.1's main new features relate to an improved use of the optimisation architecture that was implemented in 4.0 - so 4.1 is not radically different. It just produces faster object code. It will probably be of extremely high quality but, as ever, you can wait for 4.1.1 if you wish to play it safe.
I am looking forward to it a lot and even more to 4.2 which should reach a peak of sophistication and performance before the next big advance that is even now being discussed on the GCC dev mailing list - the possible integration of LLVM.
The computer to student ratio might no be good but it allows this to happen:
10000 "information sources" : 1 teacher
And I feel that that can have a much bigger effect than giving a dictionary to every kid in a class. Textbooks are a huge expense for african schools and keeping them up-to-date is hard unless they are core subjects like languages and maths. There is arguably a need for a greater emphasis on practical subjects and information for these has to be updated frequently to remain relevant.
In Zimbabwe (where I went to school), schools seemed to spend a lot of time teaching maths and science which reflects the cultural bias that working in a university, a bank or other office is more respectable than running a garage or a small construction business.
The convenience of not having to carry all those separate bits of equipment is considerable.
If you ever have opportunities to take a photo of something that you didn't expect to see, you might be glad that although you don't have a camera, you do always have your phone with you. This has happened to me on numerous occasions and even though my camera-phone is not nearly as good as the N90's, it has still been useful.
I have used an N90 (I develop mobile phone software) and although the crispness of the screen is it's most "standout" feature, I think that the biggest technology jump is really the video recorder which, unlike older phones, seems to be able to record for as long as there is space left. The quality is good by video standards and the compression is mpeg4 so the files are small. I used to think that videos clips were a gimmick until I took a few - I think that a short video clip is a lot more evocative than a photo, especially if it is of people.
This phone is a bit too clunky for me but I think it is a good preview of the features that are to come in smaller packages in future. They are all desirable once the cost comes down and you have a chance to try them.
Regards,
Tim
Re:emacs and vim are too difficult to use
on
Vim 6.4 Released
·
· Score: 1
Many people have predicted the death of Vi and emacs . . . they are now old.
I have used many editors, starting with Turbo Pascal 3.0's Wordstar-like one and including most of the windows editors out there now. Even though I was familiar with them first I was still "hooked" by Vim. Probable causes of this loyalty are: * It's independence: ** it's useful for any editing task not just writing Java or C++ ** some version can be found on the biggest systems or the smallest ** it can be used graphically or over a telnet-session ** one can find it on many platforms * Regular expression search-and-replace * Ability to repeat commands a number of times by prefixing them with a number
Slashdot Mirror
Very interesting link - thanks!
Signing makes it possible to have some level of security and some level openness. So at least Apple might be able to do *the right thing* eventually. Then again perhaps not.
Personally, I want see people using their phones on the train in the morning for something other than calls. The only people I currently see doing that are ****berry owners. I think that, given the sophistication of the devices, they are being grossly underutilised. Why? Well, simply because the features aren't good enough for the cost and trouble of using them. This is partly because phones have a very short sales life. The next model and the next model must come out to satisfy fashion. None can ever be perfected - the struggle is to keep everything working on the new hardware and to support the new "features." So there isn't enough time to spend on perfecting the user experience of phone X.
Using the same OS and UI on many phones does help because at least most of the improvements get propagated onwards but with new hardware and form factors turning up all the time there will always be problems.
When the majority of people have a smartphone this will all change if they discover one or two critically useful applications. This will introduce a slowdown where one will want to be sure that one's favoured software works on the next device. Currently it's the telephony parts of phones that users consider critical but as other things take off (e.g. you need flash and a browser to work properly for your mobile **space access) it won't be so easy to change radically all the time and a degree of stability will be created just because the large number of users will *need* it.
This is the ultimate resource. You might find that there's even too much detail.
https://www.symbiansigned.com/app/page
Basically the user has to approve anything that might cost them money so an EXE or DLL that wants internet access will trigger the installer to ask the user for permission. Programs can't change their capabilities once installed, and the data of one app can be stored such that others are not able to see or access it. Various bits of the filesystem are also inaccessible to insufficiently privileged apps etc etc. So apps are insulated from each other and the OS is shielded from apps. This makes viruses much harder to propagate. Software must be installed to work and only the system installer has the privileges to do that - and it asks the phone user and checks signing.
If some software was found to be "bad" even after testing then it's certificate could be revoked, although I'm not sure if much infrastructure has been devoted to this possible course of action.
The best a virus writer might hope for is to write something in an application's own macro language e.g. javascript in the browser. They might be able to infect the application and cause trouble but at least the damage would be limited to the capabilities of that application. So a virused browser wouldn't be able to cause a Denial of Service on the Mobile Network by doing something horrible with the phone's radio. This is still pretty cool.
A lot of effort has been put into reducing the trouble and bother of platsec and into helping developers and you can read about it on the link. e.g. there's a thing called a devcert that lets you sign your own apps but these will only install on a couple of specific phones. This allows a developer to test software on their own phone(s) up to the point of release.
My very personal opinion is that there are other problems with the structure of the market that, once solved, will set loose a flood of applications. I would probably be unwise to say anything more on that though.
Disclaimer: I work for Symbian but everything I write is my personal opinion only.
Brilliant - mod up please. This kind of control is already present on all Symbian 9.x phones (e.g. Series 60 3rd ed) -in the form of "Platsec" or Platform Security. Apps have to be signed for the "capabilities" they use. The simple-to-understand capabilities can be granted by the phone user. The more "dangerous" capabilities have to be granted by getting an external test house, with appropriate authority, to sign the application. If you want to access deep parts of Nokia's or Sony Ericsson's GSM telephony stacks the you need to get a signature from them.
It has sparked lots of heated debates because the signing costs money but it offers protection against the kinds of problems that PCs have had with the internet. Without it, phones would be a very attractive ecosystem for malicious software.
Apple seem not to have this kind of architecture in place yet so their only option is to be more restrictive for now.
I agree that the current GUI paradigm is old. It presents a model of the computer and the world. Users need to know this model to be able to plan what they want to achieve.
I have noticed that people seem to start with a very simple mental model and then when they notice some inadequacy they expand it to cope with the new needs. To begin with one doesn't tell one's granny that she can run out of disc space when "saving files" - because it's too much detail and unlikely to happen immediately on her new computer. For a beginning she can imagine her 100GB hard disc to be infinite in size.
This is just like learning to drive - you don't get into a car for the first time and drive it down the M25 (subtitute name of favorite big motorway/highway/autobahn/autostrada).
I think that the problem with computer UIs is that they have no understanding of the person who is using them so they don't know how much detail to offer them today and how much more to offer tomorrow. They aren't really conversing with the user so they can't become truly effective for that particular user.
It's a one size-fits everyone UI at the moment.
Cheers,
Tim
I think the English name might be "Flintlock Pistol" which refers to the firing mechanism - a spring-loaded chip of flint (a stone commonly used to make sparks) which is released by pulling the trigger.
Just phone your own land-line and then say, "Binladenbinladenbinladen" 10 times.
Wait 30 minutes.
If there are no black helicopters after 30 minutes then you probably aren't being bugged.
. . . just to clarify. In Japan, for example, Symbian is sold with the MOAP environment.
I am a Symbian employee though what I say is purely my own opinion:
The fact that firmware updates are freely available for the newer Nokia phones is a fairly new phenomenon and has a major effect. Problems are being dealt with swiftly.
Symbian is a much much more powerful operating system than it's competitors and it's really quite sophisticated inside. The pressure on us employees to get things right simply gets more and more tremendous. We don't do everything - not device drivers for example because the hardware isn't ours, after all - but the product delivered to our customers (the phone manufacturers) is very good and enables them to deliver amazing products like the N95 which will change ways of living. A shitload of kewl work has been put into v9 for your pleasure and it has bedded down so you will see it wipe the floor with it's opponents.
That's my personal opinion, anyhow.
Cheers.
http://www.lightblueoptics.com/
These guys do that. Somehow they have apparently demontrated colour too.
It does static analysis. It is a C++ compiler that turns your code into a form that can be easily analyed by scripts that look for particular faults. It has a low rate of false positives and it is able to find bugs even in very unusual code paths which are very hard to reproduce. This means that it is much less likely that one's customers will experience faults that you cannot reproduce (these are very difficult and expensive to fix). It should also mean that your software will pass QA at the first or second try and thus you can get it to market earlier.
The last time I checked, the fee was about 250K for a license to use it on your source base of 100K lines of code. Coverity is a very small company and it's software is particularly useful to people working on fairly large projects so it makes sense for them to charge a stiff fee.
When I think of some of the bugs that it would have found and total up the time of all the test, support, and software engineers on product X that I was working on at company Z a couple of years ago, I think that this could easily have paid for itself on one project and left us with a much more reputable product.
Splice is the open source equivalent - perhaps you should "call" for that. I wonder who has to actually do the work to satisfy you?
Hi,
n ed_evolved_with_Symbian_OS_v9.pdfq
OK, the opinions here are my own personal ones and in no way represent any other person or entity. It's also not guaranteed to be accurate.
I have seen the issue from both sides and I basically agree that getting in the way of developers is very bad. I think that the issue will become much less important thanks to the "firming up" of some APIs and the fact that manufacturers support them better.
The certificate stuff is pretty much an answer for the security argument against releasing APIs. It will stop a legion of trojans and worms from wriggling through GPRS/3G/bluetooth/wifi connections and infesting the world's mobile phone "fleet". The only alternative is to lock down phones completely and that is what might happen to the other OSes if something similar isn't done for them. I think that this is actually critical to enabling third-party software and quite positive for everyone. Symbian have put in a huge effort to make it friendly so don't be disheartened - read on to find out why it'll be ok.
The first thing to note is that you can get a free certificate (DevCert) that gives you very "deep" access but it's locked to your personal phone. This means that you can develop whatever you like on your own device.
If your application uses the 60% of the API which is considered harmless then it can run without signing.
There are a number of capabilities which are considered to be simple enough that the user can understand them and grant them during installation. So these also don't require signing.
If you use some of the extended capabilities or Phone Manufacturer Approved capabilities and you wish to make your software generally available on all phones then you'll need a certificate.
If your software is free (i.e. you make no financial gain from it) then Symbian has made free testing and free certificates available. If your software is not free then you can probably afford a "couple" of hundred dollars for it to be signed.
These links have more detail:
https://www.symbiansigned.com/How_has_Symbian_Sig
https://www.symbiansigned.com/app/page/freewareFa
Cheers,
Tim
I am an "interested party" but I still think I can clarify this somewhat.
Symbian has a big API which is understandable given that it covers multimedia, UIs and a lot of different types of networking. The documentation for almost all of that is freely available as are the SDKs.
One has to pay to get a copy of the source code, however, and access to the more volatile APIs that aren't evenly supported by all devices or that might pose some degree of security risk. This tends to include some of the telephony stuff. The source code allows you to understand what's happening ia very specific way although it doesn't include the drivers written by each manufacturer (Symbian doesn't own them after all).
It was much harder to get the equivalent API details for WinMob - in fact in some instances impossible. Good luck getting the same details out of Motorola!
Symbian made it's first ever profit in Q1 and according to it's website it makes around $2.5-3 per phone. That doesn't seem too evil.
Regards,
Tim
If a facility is useful, a very large proportion of people seem to not give a stuff about security e.g. Outlook.
I use different computers at work, at home on holiday, when I am out and about. They are all different systems and I don't own them myself necessarily. I have a powerful PDA phone but I don't always carry it with me and it has limited capacity anyhow.
The more "devices" that we end up using, the more desirable is will become to be able access one's information without having to be involved in the mechanics of how it moves around. I think so because I appreciate simplicity and I think that other people do too unless they are hooked on "messing around with gadgets."
I find Gmail useful because I can get to all my messages from anywhere and I don't need to stuff my pockets with devices or manage the disc space on them them or find a charger that works in country X or put up with the sometimes rubbishy software.
Gmail is like a worldwide clipboard too - you can be anywhere and put new links, travel details, addresses and phone numbers etc into it (a draft message) and can search through later on to find something - much better than post-it notes.
So - good for Paul Maritz and his efforts. I had better declare 2 reasons for bias, though:
1) He was born in Zimbabwe like me although he was brought up in South Africa.
2) He has flown a Hawker Hunter at Thunder City (a place in Cape Town where you can fly old British combat jets) and in some odd way I admire him for doing what I would like to do. I want to fly in the awesome BAC Lightning, though.
http://www.thundercity.com/tiger_paul.htm
No. Symbian costs about $2.50 per phone - one reason why Symbian have only recently turned a profit.
Symbian is pretty open from a developer's point of view and bigger partners can get access to the source code for the kernel and servers (without the GUI or all the device drivers because these belong to the manufacturers).
From what I can see this openness is not entirely pleasing to mobile phone operators whose thirst for "Average Revenue per User" or ARPU is unquenchable - they would rather like to monopolise their customers.
I'm from Zimbabwe and of Irish descent and I know what it is to be put down and joked about - it's something that everyone tries to do to everyone and based on many things and not only nationality.
It is much much harder to not ever put anyone else down than it is to complain when people do it to you. i.e. people always notice insults to themselves and ignore their own insults to others.
If you can honestly get rid of your prejudgement of Americans then you have some right to ask for the same lack of prejudgement of Welsh people.
Regards,
Tim
Thus spake the Assembler programmers of the first C compilers . . .
. . . and the C programmers of the C++ compilers . .
. . . and the C++ programmers of Java . . .
It would be interesting to know if ALP can support single core CPUs - where the same CPU runs the GSM/CDMA stack and all the applications as well. This is going to be a critical factor in the total cost of the operating system + handset and is an area where Symbian has a long head start over Windows Mobile.
Symbian was sold on 10.9 million handsets in the last quarter - making it an attractive target for developers such as myself. I am not denying Palm's product could be awesome but it seems to have a hell of a long way to catch up.
Cheers,
Tim
I had to find a bug that caused a crash approximately every 2 weeks. It was terrible to debug because we couldn't replicate it initially. I eventually wrote a system test that shunted about 2 weeks data through the system in an hour but using random rates and delays and I simulated comms failures by killing the test process and restarting randomly. I found the bug in a comms error handling routine using purify (you could use valgrind). It was a double-free that corrupted the stack and caused a crash somewhere else later on, sometimes. The customer account was saved.
The lesson, to me, is to write a test program which will simulate heavy but randomised use of your system and to then to use this against the system whilst running it in a memory debugger.
In addition, I would suggest that you don't really need to split the program into several processes but would probably benefit from restarting the single process every day at some low usage period or by starting a new process and handing off to it from the old one. This idea always invites derision from my colleagues but they do make just as many mistakes as me and most of these mistakes wouldn't result in costly support time and patch creation if the software was restarted often.
Regards,
TIm
The usual theory is that mobile IP will be the nucleus of IPv6 and may actually end up being larger in terms of the number of devices than the rest of the internet since lots of people have phones but not necessarily PCs.
Even if there were less phones, there would still be an incentive for business to IPv6-enable their software and hardware if there was a large market of mobile users to supply with servies.
Cheers,
Tim
Lots of people feel angry about failing to make a free product work but that's the way it is. Your kind of complaints have been made about every part of RedHat and then Fedora. First the installer, then the desktop then the applications . . . . they get fixed and people move on to complain bitterly about something else.
.no . . . someone in the Fedora community must do it for free.
Of course . . they would never put any effort in themselves . .
As you know, it hasn't actually been released yet to 4.1 is not ready. The most recent round of dramatic changes were from 3.4 to 4.0. 4.02 seems very good and I haven't seen a Fedora development update for it for quite a long time.
4.1's main new features relate to an improved use of the optimisation architecture that was implemented in 4.0 - so 4.1 is not radically different. It just produces faster object code. It will probably be of extremely high quality but, as ever, you can wait for 4.1.1 if you wish to play it safe.
I am looking forward to it a lot and even more to 4.2 which should reach a peak of sophistication and performance before the next big advance that is even now being discussed on the GCC dev mailing list - the possible integration of LLVM.
Regards,
Tim
Perhaps, but if we don't listen then we will certainly never hear . .
The computer to student ratio might no be good but it allows this to happen:
10000 "information sources" : 1 teacher
And I feel that that can have a much bigger effect than giving a dictionary to every kid in a class. Textbooks are a huge expense for african schools and keeping them up-to-date is hard unless they are core subjects like languages and maths. There is arguably a need for a greater emphasis on practical subjects and information for these has to be updated frequently to remain relevant.
In Zimbabwe (where I went to school), schools seemed to spend a lot of time teaching maths and science which reflects the cultural bias that working in a university, a bank or other office is more respectable than running a garage or a small construction business.
Regards,
Tim
Perhaps you haven't given it a try?
The convenience of not having to carry all those separate bits of equipment is considerable.
If you ever have opportunities to take a photo of something that you didn't expect to see, you might be glad that although you don't have a camera, you do always have your phone with you. This has happened to me on numerous occasions and even though my
camera-phone is not nearly as good as the N90's, it has still been useful.
I have used an N90 (I develop mobile phone software) and although the crispness of the screen is it's most "standout" feature, I think that the biggest technology jump is really the video recorder which, unlike older phones, seems to be able to record for as long as there is space left. The quality is good by video standards and the compression is mpeg4 so the files are small. I used to think that videos clips were a gimmick until I took a few - I think that a short video clip is a lot more evocative than a photo, especially if it is of people.
This phone is a bit too clunky for me but I think it is a good preview of the features that are to come in smaller packages in future. They are all desirable once the cost comes down and you have a chance to try them.
Regards,
Tim
Many people have predicted the death of Vi and emacs . . . they are now old.
I have used many editors, starting with Turbo Pascal 3.0's Wordstar-like one and including most of the windows editors out there now. Even though I was familiar with them first I was still "hooked" by Vim. Probable causes of this loyalty are:
* It's independence:
** it's useful for any editing task not just writing Java or C++
** some version can be found on the biggest systems or the smallest
** it can be used graphically or over a telnet-session
** one can find it on many platforms
* Regular expression search-and-replace
* Ability to repeat commands a number of times by prefixing them with a number
Cheers,
Tim
If you do this then your software will become irrelevant like the GNU Hurd.
Cheers,
Tim