If your OO experience is in Java then I would suggest you take a look at Scala. It compiles to the JVM and has good support for functional programming, but also supports your familiar OO techniques and interoperates with the huge amount of java code out there.
I'd still suggest you play with Haskell (probably the most pure functional language that's widely available) while you learn; it's fun and will really make you think about programming problems differently -- even if you never produce a line of production code using it.
Let me break it down for them and you: - Kids in the 1950s did not study what we study today - Kids today did not study what kids studied back in the 1950s
Check out the RSC's report, it gives much more detail of their methodology. For the most part their test choose questions from parts of the syllabus which were not different.
Of course they did choose the questions by hand and not randomly, and it's easy to criticise aspects of their study. But the overall point that exams in earlier decades tested more multistep, complex reasoning and that it has become de-emphasised in recent years is good.
The RSC claim that these are important skills for higher level chemistry, and that even the top GCSE students in the country are not gaining them (the study took the form of a competition which schools entered their best Chemistry students into).
I'm sick of this "Kids in the 1950s were smarter than today" rubbish. I know that these old accidemics studied back then and want to feel smart but making kids feel dumb today is wrong and they should feel ashamaned.
It would seem likely to me that the kids of today are about as smart overall as their 1950s counterparts. It's possible that, in comparison to the post-war 50s, better levels of nutrition available today mean smarter kids (or I suppose that greater consumption of junk food means dumber kids), but overall I expect things are similar.
Which is why it's strange that the Government has reported annual increases in exam marks for over two decades now, while rejecting the idea that the exam grades are getting easier to attain.
I certainly believe there has been grade inflation i.e. that a grade A in Chemistry means less than it did some decades ago. That doesn't necessarily mean that current students are thicker though just that some people who would have got Bs are now getting As (hence the need for the new A* classification).
The RSC is also claiming that University entrants (for Chemistry) have less developed skills in multi-step mathematical reasoning, and that does make a lot of sense: It's not being tested for now, and there's greater pressure to use teaching time to train for the exams than there used to be. I don't have anything to do with secondary education directly, but my mother is a school governor, and what I hear does agree with this view.
PS: I am curious about these accidemics; I'm imagining they're people who ended up in an academic career by chance?
One problem with standardising on Debian is that you don't really know how long your machines will be supported with security updates. Debian's policy for security updates is the release of a new stable plus 12 months, without a clear schedule of when that might be. Sure, for woody, this ended up being almost forever, but there's no way to plan ahead for this.
It looks like there's a new debian-stable just around the corner and your 12 months start ticking to get your whole enterprise upgraded,
Personally, I dislike updating working systems (software testing on a different OS version, taking hardware out of service etc. - yuck) and so I like to know the support envelope when I install. Preferably staying with an OS version for the expected life of the hardware.
At work I do currently manage a couple of Debian systems along with a mixture of Redhat and Ubuntu (plus a few SuSE). Over time, the Ubuntu has grown to dominate the desktop & workstations and users familiarity with it on the desktop means it's beginning to gain ground over RHEL on servers too. It's rpm free too. Maybe it's not as stable as etch, but I do like having a 5 year support span on the Ubuntu LTS releases.
This is not really true. There are multiple standards to solve this problem.
One is the TLS server name indication (SNI) in RFC 3546. This allows the browser to specify which VHOST it wants as it's setting up the TLS connection allowing the web server to give it the right cert.
Another approach is TLS-upgrade (RFC 2817). This allows a browser to make an unencrypted HTTP connection then upgrade to secure TLS (in the same way that encrypted SNMP and other protocols work without requiring 2 different ports). This will allow secure or insecure both over port 80.
The problem is lack of browser support (and to a degree lack of server support, but people would extend/upgrade their apache/IIS/etc installs for the added functionality *if* the browsers supported it). FF2+ and IE7+ support RFC 3546, and FF3 will introduce support for RFC2817, but it'll be many years before browser share is high enough for public web sites to consider using it.
It's true. I had been using Redhat linux, and later fedora, as my main desktop since Redhat 3.0.3 days. The month that fedora legacy changed its policy from 1-2-3 and out (which effectively meant 18 months or so) to only 12 months support I installed my first Ubuntu desktop, and fairly quickly moved all my desktops to Ubuntu. It took a little while to become familiar with the subtleties of Debian rather than Redhat ways, but the software support was as good as fedora. RHEL didn't have the 3rd party software RPMs of fedora even if it did have long term updates.
I'm not at all convinced that this was a good move by Redhat. My Redhat skills get more rusty and I feel more and more confident with Debian. I did use a mixture of RHEL, Scientific Linux and CentOS on work servers, but since switching to Ubuntu I now use almost exclusively Debian based distros (apart from a small cluster running the SuSE install it was deleviered with).
It's possible that since I switched RHEL has improved, i.e. there are more 3rd party packages for CentOS, but I wouldn't consider fedora -- I don't have time for extra upgrades. It's easier to lose a user than to regain one.
The UK voltage is 240. 220v is the standard on most (all?) of continental europe. My understanding is that the European Union wanted to "harmonise" the power supply and so standardised on 230v +-10v, allowing everyone to continue as before and still be within spec.
UK sockets are 13A so do indeed deliver almost twice the power of US sockets. It's clear that this was chosen with great foresight by those building the power infrastructure to allow us all a quick cup of Tea. Using 240v also means that we need less copper to wire our houses.
An open design from a project at the University of Bath. It has OSS control software and is specifically designed to be self replicating, using only 400 of materials.
Mind you the blog site for the book that started going into detail about why Singleton (the most obvious fucking "pattern") is so "beautiful". I could have blown chunks right there. Effective and elegant if used correctly? Yes. A neat technique? Yes. Beautiful? Shit I almost spilled my coffee when I followed the link and saw that. However, when I read the whole blog posting it was about how Singleton might seem beautiful but is actually seductive and dangerous; much more sensible just dressed up with a somewhat trollish title.
Mind you there was another post about the beauty of Java reflections that seemed to be serious!
(Oh and I loved K&R when I learnt C from it, but then I was coming from a degree in Mathematics so it seemed like light relief)
True, and yet I'm not sure that this is the best decision for Fedora/RedHat.
I've been using RedHat for over 10 years -- since RH3.03. Initially, I ran downloaded RH on my workstations and bought a licenced copy for any servers I deployed. After RH9, I ran Fedora on my workstations and either RHEL or one of the respins on servers. The Fedora Legacy project provided valuble support for production RH7/RH9 servers in the changeover period (somewhat badly handled by RHAT), and extended support for Fedore Core releases by a valuable few months.
In recent months it's become obvious that Fedora Legacy wasn't functioning any more leaving FC with just 12 months support. With a 6 month release rate this means that when I come to install a machine I have on average 9 months of updates left. This just feels too short. I choose Fedora over RHEL or clones on desktop/laptop because it supports more hardware, uses more uptodate components, and more (OSS) 3rd party software comes with FC downloads/support. I can understand Redhat wanting to keep some clear water between their Fedora project and Enterprise products, but 12 month support (or the recently suggested 13) is too little. I use my machines to get things done, not just as a toy, and don't wish to spend time on upgrades for at least a year after installation.
In the last two months I've started install Ubuntu on new machines and switching instead of upgrading old. The amount of available software and community support are comparible, it's just meant getting used to a few Debianisms instead of the RHisms I've been used to. My servers currently still run RH family OSs, but as I get more familiar with Debian like systems and do development on Ubuntu systems, new services will naturally be deployed to Debain/Ubuntu rather than RH.
I wish no ill will to Fedora, but it no longer meets my needs. We'll have to see about Canonical/Ubuntu long term but their stated goals with LTS versions sound ideal. In retrospect I wish I'd contributed to the Legacy project but time as ever was too pressured.
There are examples of Open Source Software bettering commercial in accessibility.
Rockbox, the replacement firmware for archos mp3 players, is not only a lot better than the original, but more accessible too. See http://www.rockbox.org/twiki/bin/view/Main/BlindFA Q. I understand that this makes it the only off the shelf, inexpensive, mp3 player with such accessibility. Specialist units available at the time were much more expensive.
This was due to the involvment of blind users in the development mailing lists. One or two initial blind users started using it when it got basic support for identifying where you were in the menus aurally (usefull to the sighted if your mp3 player is in your pocket), and with their feedback it quickly improved.
With Rockbox, input to the development team from disabled users was very well received, and I think this would be the case with very many OSS projects. OSS depends on community input and that doesn't just include coders. Clear descriptions of requirements; design discussions on mail-lists; testing of features, they're are all useful. If projects have disabled voices within their development community then I'm sure that accessible OSS software will result.
but so many people consider their internet connection a permanent fixture and don't worry about the negatives of downtime or a possible catastrophic host failure that deletes all their archives.
Most users in the category you describe are going to be better off using a service provider's storage, the chances of failure being much lower than on their home administered PCs. At least with a webmail provider there's a good chance of some sort of backup regime being in place (in the case of google a cool replicating distributed filesystem across multiple datacentres).
Not so sure I'd want google's huge cluster analysing my email archives though.
I have the 2Gb version (which accounts for the price, Adamond can't command the volume discounts of Apple). It takes a single AAA battery, has a small but sharp OLED display which is totally unable to display JPEGs, or even thumbnail video clips, and records to wav on the built in mic and MP3 though the line in. I've used it with Linux over USB2 and it plays my ogg vorbis files fine.
The earphones it came to weren't great, but it sounds great with a pair of sennheiser earbuds.
Although it would be hard to judge from the way this has been reported in the media, student and national. Your speculation about the covering up of security flaws, known or unknown, is wide of the mark.
In fact, they didn't uncover any major security flaw which the University IT support were unaware of. As I understand it, some traffic was sniffed on an old unswitched hub. I believe, the last one in use at that college, and which was scheduled to be replaced with switched connections. Though that hadn't yet been implemented partly due to the budgetary constraints mentioned in the article. Even with a switched network people playing games with ARP can sniff traffic, though at least that's an active attck which can be detected by diligent admins.
Lo and behold, when the students looked at the traffic they found IM content being sent in the clear and a whole lot of Outlook users collecting their mail by POP/IMAP rather than IMAPS. This is no surprise to anyone in IT support though it may well have shocked some of the more clueless users,
This is certainly against the University's computer use policy, and as such they are being investigated by the Proctors. They do have the authority to suspend student's access to University buildings and facilities (or Rusticate them, in local terms), but as far as I know no decision on what sanction, if any, they will face has been reached.
IT staff at the University do try to keep users informed about network security, and students are told to use secure methods to access email servers, but obviously more education could always be done. Much effort has been needed recently in keeping Windows users up to date with security patches, and AV software. The more effort is spent on communicating these matters the less attention students have left to listen to more general security messages.
Do you think people should be able to sell the risk of something bad happening to you? If so, then do you think both you
and an insurance company should have an equal shot at measuring the risk? Remember, you can get a genetics test without
telling the insurance company.
That's not the case. The insurance companies aren't trying to make you take a test (not yet, anyway), just asking you (as is already the case with HIV tests) to disclose whether you have already had one. If you're at risk, you stand a good chance of having taken the test, and not disclosing it would be grounds for voiding your insurance.
In the 80's merely having taken an HIV test (not tested +ve) was grounds for being refused life insurance, mortgages, etc., due to logic that if you had to ask whether you were HIV +ve you must somehow be high risk. I'm not sure, but I believe the increasing number of people taking HIV tests (e.g. for employers benifit, rather than their own risk) means that you wouldn't necessarily be refused outright these days, but it's a similar situation and not one I'm very comfortable with.
I can well believe that if they decided today to start on a transmeta-like CPU it would take years to come to market as Transmeta claim (maybe not 5, but close to it). The question is whether Transmeta's approach really does provide the best way forward? Even if it does, is it enough of an improvement when Intel have so many more resources to throw at increasing the clock speed of their Pentium series?
I code for the palm, using XEmacs, the gnu toolchain based prc-tools, and debug on the Palm emulator (a GPL application being activly supported by Palm) using gdb via the DDD front end. This provides a very nice development environment, for which I have all the source, and paid nothing.
I doubt that Agenda, have as got their offerings as well set up (how long will it be before they have a polished emulator?) for programming using an open source platfrom, or not.
There is a considerable amount of open source software available for PalmOS, as well as the huge amount of shareware/commercial software - mostly written by Windows users.
66 MHz machines running X windows must be dog-slow, too.
Nonsence, most 66MHz processors are more than capable of running X windows. I have a 66MHz 486DX in the next room which does a perfectly adequate job (although not at 1600x1200 x32bit - obviously). At quarter vga mono, it would have no trouble. X windows was around well before 66MHz processors were, even in desktops.
I thought that most similarities between them were due to the fact that they both, in part, derived their UI and style guidelines from IBM's CUA, but I don't have any supporting references at the moment.
Anyone have a memory of pre 3.0?
I remember using windows v2 in the mid-late 80s. I don't have many specific memories, apart from that there were quite a number of shells which sat on top of DOS around at the time, and since all the useful programs were DOS based, win2 was effectively just another one. We dumped it after a quick try, and stuck with a text mode based shell Norton Commander style.
However, Linux is slowly entering the realm of the non-geeks. They will run their system as preinstalled. They will not update their system when a bug is discovered (and fixed) They will run as root if it saves them the trouble of remembering *two* passwords.
These are all problems that can be (pretty easily) solved with distributions aimed at end users. It would take little difficulty to config a box to allow sessions at the console to login without requiring a user password, and for the programs which alter system settings to ask for an administration password then su to root as necessary. This would not put a lot of obstacles in the way of a non-geek. In fact this is the exact way that my SGI O2 can be set up out of the box.
Both Debian and Redhat already have package management utilities which can automatically check for new upgrades. There is no reason why these ahouldn't be built into the system so that (at least in the case of security fixes) any installation connected to the network will periodically check and upgrade. This might go against the grain to a hard-core geek, but would suit a consumer dist just fine.
Hey, these are the people who has their password on a post it note beside the screen!
Personally, I don't see why people make such a fuss over this point. People should take appropriate care of their passwords. In the case of a personal system, with little confidential information, keeping the password to hand may well be acceptable. If an adversary is prepared to break into your home to get at your computer, an inexperienced user wouldn't have much of a chance anyway.
In my mind, for instance, it is better that a workstation user carries there password in their wallet than chooses the name of their dog for fear of forgetting anything more secure. On the other hand if I were to leave the root passwords to our servers on a scrap of paper in my coat pocket, then as the system admin I would be behaving incompetently. Security is a matter of balancing the threats, and potential damage, against convenience of use. Too many annoying security rules which users can't see the need for only leads to laxness.
Microsoft develops Windows NT using a directory of text files held together by makefiles.
I'm not sure you've picked the best example. Did win2000 come in on time and on budget? Of course this doesn't mean that it's MS's choice of dev tools that are to blame;-)
BTW I use Xemacs for all my development (works X-platform and X-language). What I think we could do with are tools which provide some of the back-end work provided by IDE's (E.g. syntax analysis, context sensitive completion, incremental compilation, etc.) through a well defined interface/protocol. Editors such as emacs/vi could link to these to provide services, and for those that want it graphical front ends, could provide the whole IDE experience. This would be much more flexible, and more in keeping with Unix/Linux philosophy, than the shove it all in one product, GUI and all, approach of commercial vendors products.
Slashdot Mirror
If your OO experience is in Java then I would suggest you take a look at Scala. It compiles to the JVM and has good support for functional programming, but also supports your familiar OO techniques and interoperates with the huge amount of java code out there.
I'd still suggest you play with Haskell (probably the most pure functional language that's widely available) while you learn; it's fun and will really make you think about programming problems differently -- even if you never produce a line of production code using it.
Let me break it down for them and you:
- Kids in the 1950s did not study what we study today
- Kids today did not study what kids studied back in the 1950s
Check out the RSC's report, it gives much more detail of their methodology. For the most part their test choose questions from parts of the syllabus which were not different.
Of course they did choose the questions by hand and not randomly, and it's easy to criticise aspects of their study. But the overall point that exams in earlier decades tested more multistep, complex reasoning and that it has become de-emphasised in recent years is good.
The RSC claim that these are important skills for higher level chemistry, and that even the top GCSE students in the country are not gaining them (the study took the form of a competition which schools entered their best Chemistry students into).
I'm sick of this "Kids in the 1950s were smarter than today" rubbish. I know that these old accidemics studied back then and want to feel smart but making kids feel dumb today is wrong and they should feel ashamaned.
It would seem likely to me that the kids of today are about as smart overall as their 1950s counterparts. It's possible that, in comparison to the post-war 50s, better levels of nutrition available today mean smarter kids (or I suppose that greater consumption of junk food means dumber kids), but overall I expect things are similar.
Which is why it's strange that the Government has reported annual increases in exam marks for over two decades now, while rejecting the idea that the exam grades are getting easier to attain.
I certainly believe there has been grade inflation i.e. that a grade A in Chemistry means less than it did some decades ago. That doesn't necessarily mean that current students are thicker though just that some people who would have got Bs are now getting As (hence the need for the new A* classification).
The RSC is also claiming that University entrants (for Chemistry) have less developed skills in multi-step mathematical reasoning, and that does make a lot of sense: It's not being tested for now, and there's greater pressure to use teaching time to train for the exams than there used to be. I don't have anything to do with secondary education directly, but my mother is a school governor, and what I hear does agree with this view.
PS: I am curious about these accidemics; I'm imagining they're people who ended up in an academic career by chance?
One problem with standardising on Debian is that you don't really know how long your machines will be supported with security updates. Debian's policy for security updates is the release of a new stable plus 12 months, without a clear schedule of when that might be. Sure, for woody, this ended up being almost forever, but there's no way to plan ahead for this.
It looks like there's a new debian-stable just around the corner and your 12 months start ticking to get your whole enterprise upgraded,
Personally, I dislike updating working systems (software testing on a different OS version, taking hardware out of service etc. - yuck) and so I like to know the support envelope when I install. Preferably staying with an OS version for the expected life of the hardware.
At work I do currently manage a couple of Debian systems along with a mixture of Redhat and Ubuntu (plus a few SuSE). Over time, the Ubuntu has grown to dominate the desktop & workstations and users familiarity with it on the desktop means it's beginning to gain ground over RHEL on servers too. It's rpm free too. Maybe it's not as stable as etch, but I do like having a 5 year support span on the Ubuntu LTS releases.
John
This is not really true. There are multiple standards to solve this problem.
One is the TLS server name indication (SNI) in RFC 3546. This allows the browser to specify which VHOST it wants as it's setting up the TLS connection allowing the web server to give it the right cert.
Another approach is TLS-upgrade (RFC 2817). This allows a browser to make an unencrypted HTTP connection then upgrade to secure TLS (in the same way that encrypted SNMP and other protocols work without requiring 2 different ports). This will allow secure or insecure both over port 80.
The problem is lack of browser support (and to a degree lack of server support, but people would extend/upgrade their apache/IIS/etc installs for the added functionality *if* the browsers supported it). FF2+ and IE7+ support RFC 3546, and FF3 will introduce support for RFC2817, but it'll be many years before browser share is high enough for public web sites to consider using it.
John
It's true. I had been using Redhat linux, and later fedora, as my main desktop since Redhat 3.0.3 days. The month that fedora legacy changed its policy from 1-2-3 and out (which effectively meant 18 months or so) to only 12 months support I installed my first Ubuntu desktop, and fairly quickly moved all my desktops to Ubuntu. It took a little while to become familiar with the subtleties of Debian rather than Redhat ways, but the software support was as good as fedora. RHEL didn't have the 3rd party software RPMs of fedora even if it did have long term updates.
I'm not at all convinced that this was a good move by Redhat. My Redhat skills get more rusty and I feel more and more confident with Debian. I did use a mixture of RHEL, Scientific Linux and CentOS on work servers, but since switching to Ubuntu I now use almost exclusively Debian based distros (apart from a small cluster running the SuSE install it was deleviered with).
It's possible that since I switched RHEL has improved, i.e. there are more 3rd party packages for CentOS, but I wouldn't consider fedora -- I don't have time for extra upgrades. It's easier to lose a user than to regain one.
I learnt my first programming on a clone of the TRS-80 (the Video Genie).
16 Kb RAM and it had graphics too -- 128x48 pixels, I seem to remember.
Fun, even if it was MS BASIC.
The UK voltage is 240. 220v is the standard on most (all?) of continental europe. My understanding is that the European Union wanted to "harmonise" the power supply and so standardised on 230v +-10v, allowing everyone to continue as before and still be within spec.
UK sockets are 13A so do indeed deliver almost twice the power of US sockets. It's clear that this was chosen with great foresight by those building the power infrastructure to allow us all a quick cup of Tea. Using 240v also means that we need less copper to wire our houses.
Then you ought to check out the RepRap project see:
http://reprap.org/
An open design from a project at the University of Bath. It has OSS control software and is specifically designed to be self replicating, using only 400 of materials.
Mind you there was another post about the beauty of Java reflections that seemed to be serious!
(Oh and I loved K&R when I learnt C from it, but then I was coming from a degree in Mathematics so it seemed like light relief)
MS has been doing quite a lot of advertising over here of the use the London Stock Exchange makes of Windows server and MS-SQL:
s tudies/lse.mspx
http://www.microsoft.com/windowsserver/facts/case
So I guess it's swings and roundabouts in the PR game.
True, and yet I'm not sure that this is the best decision for Fedora/RedHat.
I've been using RedHat for over 10 years -- since RH3.03. Initially, I ran downloaded RH on my workstations and bought a licenced copy for any servers I deployed. After RH9, I ran Fedora on my workstations and either RHEL or one of the respins on servers. The Fedora Legacy project provided valuble support for production RH7/RH9 servers in the changeover period (somewhat badly handled by RHAT), and extended support for Fedore Core releases by a valuable few months.
In recent months it's become obvious that Fedora Legacy wasn't functioning any more leaving FC with just 12 months support. With a 6 month release rate this means that when I come to install a machine I have on average 9 months of updates left. This just feels too short. I choose Fedora over RHEL or clones on desktop/laptop because it supports more hardware, uses more uptodate components, and more (OSS) 3rd party software comes with FC downloads/support. I can understand Redhat wanting to keep some clear water between their Fedora project and Enterprise products, but 12 month support (or the recently suggested 13) is too little. I use my machines to get things done, not just as a toy, and don't wish to spend time on upgrades for at least a year after installation.
In the last two months I've started install Ubuntu on new machines and switching instead of upgrading old. The amount of available software and community support are comparible, it's just meant getting used to a few Debianisms instead of the RHisms I've been used to. My servers currently still run RH family OSs, but as I get more familiar with Debian like systems and do development on Ubuntu systems, new services will naturally be deployed to Debain/Ubuntu rather than RH.
I wish no ill will to Fedora, but it no longer meets my needs. We'll have to see about Canonical/Ubuntu long term but their stated goals with LTS versions sound ideal. In retrospect I wish I'd contributed to the Legacy project but time as ever was too pressured.
Perhaps it's conservation?
http://news.bbc.co.uk/1/hi/uk/3783491.stm
Though I'm not sure Oregon has a skylark population.
There are examples of Open Source Software bettering commercial in accessibility.
A Q. I understand that this makes it the only off the shelf, inexpensive, mp3 player with such accessibility. Specialist units available at the time were much more expensive.
Rockbox, the replacement firmware for archos mp3 players, is not only a lot better than the original, but more accessible too. See http://www.rockbox.org/twiki/bin/view/Main/BlindF
This was due to the involvment of blind users in the development mailing lists. One or two initial blind users started using it when it got basic support for identifying where you were in the menus aurally (usefull to the sighted if your mp3 player is in your pocket), and with their feedback it quickly improved.
With Rockbox, input to the development team from disabled users was very well received, and I think this would be the case with very many OSS projects. OSS depends on community input and that doesn't just include coders. Clear descriptions of requirements; design discussions on mail-lists; testing of features, they're are all useful. If projects have disabled voices within their development community then I'm sure that accessible OSS software will result.
I heartily agree with most of this. Apart from:
but so many people consider their internet connection a permanent fixture and don't worry about the negatives of downtime or a possible catastrophic host failure that deletes all their archives.
Most users in the category you describe are going to be better off using a service provider's storage, the chances of failure being much lower than on their home administered PCs. At least with a webmail provider there's a good chance of some sort of backup regime being in place (in the case of google a cool replicating distributed filesystem across multiple datacentres).
Not so sure I'd want google's huge cluster analysing my email archives though.
I have an Adamond ZK1 which hits all the points on your list, except perhaps price -- it wasn't the cheapest out there.
a mond_zk1/
http://www.adamond.com/press/zk1.html
I have the 2Gb version (which accounts for the price, Adamond can't command the volume discounts of Apple). It takes a single AAA battery, has a small but sharp OLED display which is totally unable to display JPEGs, or even thumbnail video clips, and records to wav on the built in mic and MP3 though the line in. I've used it with Linux over USB2 and it plays my ogg vorbis files fine.
The earphones it came to weren't great, but it sounds great with a pair of sennheiser earbuds.
The register have a review:
http://www.theregister.co.uk/2005/07/07/review_ad
What was decided was dropping support for RH7.2 and RH8, to make the task of getting updates for 7.3, RH9 (and now FC1) out quicker.
Although it would be hard to judge from the way this has been reported in the media, student and national. Your speculation about the covering up of security flaws, known or unknown, is wide of the mark.
In fact, they didn't uncover any major security flaw which the University IT support were unaware of. As I understand it, some traffic was sniffed on an old unswitched hub. I believe, the last one in use at that college, and which was scheduled to be replaced with switched connections. Though that hadn't yet been implemented partly due to the budgetary constraints mentioned in the article. Even with a switched network people playing games with ARP can sniff traffic, though at least that's an active attck which can be detected by diligent admins.
Lo and behold, when the students looked at the traffic they found IM content being sent in the clear and a whole lot of Outlook users collecting their mail by POP/IMAP rather than IMAPS. This is no surprise to anyone in IT support though it may well have shocked some of the more clueless users,
This is certainly against the University's computer use policy, and as such they are being investigated by the Proctors. They do have the authority to suspend student's access to University buildings and facilities (or Rusticate them, in local terms), but as far as I know no decision on what sanction, if any, they will face has been reached.
IT staff at the University do try to keep users informed about network security, and students are told to use secure methods to access email servers, but obviously more education could always be done. Much effort has been needed recently in keeping Windows users up to date with security patches, and AV software. The more effort is spent on communicating these matters the less attention students have left to listen to more general security messages.
Does this new release restore any support for the MNG/JNG graphics formats, or is GIF still the only animated format supported?
That's not the case. The insurance companies aren't trying to make you take a test (not yet, anyway), just asking you (as is already the case with HIV tests) to disclose whether you have already had one. If you're at risk, you stand a good chance of having taken the test, and not disclosing it would be grounds for voiding your insurance.
In the 80's merely having taken an HIV test (not tested +ve) was grounds for being refused life insurance, mortgages, etc., due to logic that if you had to ask whether you were HIV +ve you must somehow be high risk. I'm not sure, but I believe the increasing number of people taking HIV tests (e.g. for employers benifit, rather than their own risk) means that you wouldn't necessarily be refused outright these days, but it's a similar situation and not one I'm very comfortable with.
John
Yes, but they took a long time to do it.
I can well believe that if they decided today to start on a transmeta-like CPU it would take years to come to market as Transmeta claim (maybe not 5, but close to it). The question is whether Transmeta's approach really does provide the best way forward? Even if it does, is it enough of an improvement when Intel have so many more resources to throw at increasing the clock speed of their Pentium series?
John
I code for the palm, using XEmacs, the gnu toolchain based prc-tools, and debug on the Palm emulator (a GPL application being activly supported by Palm) using gdb via the DDD front end. This provides a very nice development environment, for which I have all the source, and paid nothing.
I doubt that Agenda, have as got their offerings as well set up (how long will it be before they have a polished emulator?) for programming using an open source platfrom, or not.
There is a considerable amount of open source software available for PalmOS, as well as the huge amount of shareware/commercial software - mostly written by Windows users.
John
Nonsence, most 66MHz processors are more than capable of running X windows. I have a 66MHz 486DX in the next room which does a perfectly adequate job (although not at 1600x1200 x32bit - obviously). At quarter vga mono, it would have no trouble. X windows was around well before 66MHz processors were, even in desktops.
John
Anyone have a memory of pre 3.0?
I remember using windows v2 in the mid-late 80s. I don't have many specific memories, apart from that there were quite a number of shells which sat on top of DOS around at the time, and since all the useful programs were DOS based, win2 was effectively just another one. We dumped it after a quick try, and stuck with a text mode based shell Norton Commander style.
John
They will run their system as preinstalled.
They will not update their system when a bug is discovered (and fixed)
They will run as root if it saves them the trouble of remembering *two* passwords.
These are all problems that can be (pretty easily) solved with distributions aimed at end users. It would take little difficulty to config a box to allow sessions at the console to login without requiring a user password, and for the programs which alter system settings to ask for an administration password then su to root as necessary. This would not put a lot of obstacles in the way of a non-geek. In fact this is the exact way that my SGI O2 can be set up out of the box.
Both Debian and Redhat already have package management utilities which can automatically check for new upgrades. There is no reason why these ahouldn't be built into the system so that (at least in the case of security fixes) any installation connected to the network will periodically check and upgrade. This might go against the grain to a hard-core geek, but would suit a consumer dist just fine.
Hey, these are the people who has their password on a post it note beside the screen!
Personally, I don't see why people make such a fuss over this point. People should take appropriate care of their passwords. In the case of a personal system, with little confidential information, keeping the password to hand may well be acceptable. If an adversary is prepared to break into your home to get at your computer, an inexperienced user wouldn't have much of a chance anyway.
In my mind, for instance, it is better that a workstation user carries there password in their wallet than chooses the name of their dog for fear of forgetting anything more secure. On the other hand if I were to leave the root passwords to our servers on a scrap of paper in my coat pocket, then as the system admin I would be behaving incompetently. Security is a matter of balancing the threats, and potential damage, against convenience of use. Too many annoying security rules which users can't see the need for only leads to laxness.
John Pybus
I'm not sure you've picked the best example. Did win2000 come in on time and on budget? Of course this doesn't mean that it's MS's choice of dev tools that are to blame ;-)
BTW I use Xemacs for all my development (works X-platform and X-language). What I think we could do with are tools which provide some of the back-end work provided by IDE's (E.g. syntax analysis, context sensitive completion, incremental compilation, etc.) through a well defined interface/protocol. Editors such as emacs/vi could link to these to provide services, and for those that want it graphical front ends, could provide the whole IDE experience. This would be much more flexible, and more in keeping with Unix/Linux philosophy, than the shove it all in one product, GUI and all, approach of commercial vendors products.