I've mixed feelings about XP - needless to say I want to avoid it as long as possible - but I can think of several reasons why I may want to use an OS that assumes I am a software pirate - for example:
Lots of software is pirated... and XP would prove a good platform on which software with more flexible licences could be developed. If I wanted to work with a particular expensive piece of specialist software for just a day, at present I'm often faced with spending thousands to buy a copy, or to be forced to use "go illegal". XP hails a promise of a new legal solution without having to buy outright software I don't need to keep.
If XP brings about cryptographic authentication of users (as hideous as this sounds) it brings a significant bonus to vendors of bespoke systems. By providing a mechanism to reliably identify users/machines, it should dramatically reduce the (possibly unintentional) illegal duplication of systems - hence enabling a fair distribution of charges between customers and enabling developers to provide unencumbered systems, which can be easily audited.
XP's registration paves the way towards authentication of people, and hopefully towards the widespread use of secure digital signatures in mainstream computing. I would see this as a big bonus as it would improve my confidence in received messages. Hence, XP registration can be seen as being central to widespread use of this enabling technology.
The fact that every installation will need to be purchased is likely to drive down the unit price over time... not as a direct result of competition from other vendors, but because the optimal price (to maximise profit) is likely to be lower. MS depends upon its monopoly on the desktop, and this means that it is in MS's interest to ensure that XP will be the OS on every home computer, which won't happen if they price themselves out of the market.
Hardware vendors will love this product, as it will have a psychological impact on your average user/buyer. Many people are willing to install new "borrowed" software on old hardware (as it costs nothing) but when faced with a charge are far more likely to want to upgrade their "system" as they must pay real money anyway. Once hardware vendors realise the benefit, they will have an incentive to ensure compatibility only with XP.
The real hope for open source and free software is that MS are forced to publicly release their protocols some time before they release their products, as this presents new opportunities for other players to move into the market of services, which is (IMHO) the holy grail for commercial computing software. Linux and friends don't need to be on the desktop in the foreseeable future in order to make their presence felt - providing that MS clients can interact seamlessly with non MS servers. In the long term, I can envisage a situation in which new competition may arise to compete solely for the desktop user interface, which will only be possible once operational issues are completely decoupled from the end user interface.
I don't like the idea of check digits - while I am aware that
it can be made arbitrarily unlikely that two different
messages possess the same checksum, the possibility would
reduce confidence in e-mail, which can't be a good thing!
I've wondered for quite some time why it is not a standard
feature for a POP account to decline mail which is not addressed
to one of a user-defined set of e-mail address regexps.
The vast majority of my spam doesn't even mention my name
in the address... and if it wasn't sent to me then I don't
want to know! I realise this would require additional
configuration for each user (particularly if you frequently
join/leave mailing lists, but I for one would immediately
see the benefits if I didn't need to wait downloading
such obvious spam over dialup!
This is quite a strange question, in so far as I'd argue that it misses the point, but maybe it just sounds strange from my perspective. Any "expert programmer" should be quite able to make the conceptual leap to adopt whatever programming language is most appropriate for the task at hand, and the wider the experience of the programmer, the more effective (s)he becomes at solving new problems.
That said, as a matter of professional interest, C/C++ is a serious option as far as computationally demanding applications are concerned (e.g. server side work, systems programming, medium scale embedded systems etc.) - This is likely to be particularly valuable as it is within this paradigm that a vast quantity of existing software has been written, but IMHO anyone deciding to use C/C++ to write GUIs these days require their head examining. As mentioned in previous posts, Java is a serious player in the marketplace, and I was very impressed with Jbuilder 4.0, the least comprehensive version of which is freely available. C# has also been mentioned, and while I feel this development is not without merit, I am concerned that its proprietary status may draw programmers into the same "dead ends" as VB. Python, to me, falls into the same basket as TCL/TK did several years ago - a scripting language which is not directly comparable with the other suggestions here.
If this post is asking about "cool" languages to try out, you could go much further astray than to consider ML, (an established and capable functional language) or maybe Smalltalk There are a whole host of alternatives out there, and it's silly to pick your language before you pick your problem.
Is it only me that smells a rat? Or maybe this should read is it only
me that is so cynical? Think for a few moments... this is a games
company, who are developing, but have not yet released a new game, and
they are now in the process of publicising their new toy.
The line "the company rep said 'because other companies would try
to steal our employees'." to me sounds comical. Surely this is
no more than a lame attempt to hype their latest efforts? Lets face
it, what self respecting employer would ever consider a candidate
worthwhile because they've been on TV? Conversely, generating an
air of secrecy and excitement may prove a productive ploy - and has
the desirable side effect of masking a bunch of callow youths, who
quite possibly look far more impressive incognito - especially
if any of them sport the western programmers' appearance after a long
night of code cranking.
As far as I can see, what is required is a highly visible strategy to alert administrators of security vulnerabilities. I suspect that a standardised service to log arbitrary security messages - with time of day, etc. would be sufficient to illicit a culture for writing vaccines, if you will, which inform administrators of problems, and inform the world who is responsible.
The reason why I believe this will work is the social hypothesis that many hackers are principally motivated by technical challenge rather than a desire for anarchic destruction. It would seem sensible to embrace the efforts of such people by providing a safe mechanism by which they can showcase their talents on live systems (which should be assumed to be under attack in any case.) If this were to be coupled with a standard strategy for distributing patches and updates to software, then it would seem harmless enough to allow "hackers" to keep tally of the number of systems each "vaccine virus" had successfully attacked, and which now had the security deficiency patched - thereby replacing the drudgery of securing systems with an engaging game which could be played by anyone with sufficient time, knowledge and inclination. Finally, an opt-in strategy would ensure the ostriches of this world are not offended, and in a sense 'legalise' non-malicious hacking.
I recall from my Undergrad days that all the Asian students were accused of plagiarism after a relatively simple open-ended task. These students all submitted the same elegant solution, even with the same variable names and it all seemed an open and shut case. The students were known to study together, and all submitted near identical answers.
The reason? These foreign students took their grades very seriously (they were paying for this course themselves) and attended all the available tutorials hoping to gain an edge over their fellow students. In one of these optional tutorial sessions, an identical problem was used as a worked example, and the students present took copious notes.
While phrase matching may assist an investigation into cheating, I feel it is vital that academics don't fall into the popular trap of believing that everything churned out by a computer is necessarily the whole story. Unfortunately, as the quality of automated systems improves, so does misplaced confidence.
It would appear to me that there should be no conflict of interest when archiving Usenet if the authors' identities were to be masked. This would allow a vast quantity of potentially interesting material to be available for reference without raising the inevitable problems when contributors no-longer wish to concern themselves with a discussion from a decade previous.
I would only be interested in the argument/information contained within posts, and the authorship is irrelevant to me. I realise that removing all names would be problematic as other contributors may have chosen to embed others' names into their articles, but surely it would be a simple compromise to strip headers and trailing signatures from well formed news posts?
I'd like to avoid the holy-debate regarding weather or not C++ is a good language - I use it because it is well suited to many of the programs I wish to write... A well-written C++ program exhibits reasonable performance, and judiciously applying the available language features make for reasonably clear programs. There are better languages when comparing C++ against others, but being a ubiquitous C derivative makes it a good compromise for some projects (most notably server side service implementations). That said, there are many facets of C++ which I find infuriating, many of which could be easily resolved at the language level... The top 5 on my personal wish list include:
1) Standardised ordinals - signed and unsigned - of known platform independent sizes - maybe we could have something like int8, int16, int32, int64, nat8, nat16, nat32 and nat64?
2) A mechanism by which overflow can be detected when performing arithmetic on numbers (particularly for ordinals above)
3) Run-time polymorphism - not only would this satisfy language zealots who claim it's impossible to implement the identity operation as a compiled C++ function - but would also make it possible to easily implement more flexible containers.
4) A type system to constrain template instantiations. When I template a class with another, I want to be explicit about what behaviours my template parameter type should support, and to have this aspect of my program statically verified.
5) An improved scheme for enumerations whereby new enumerated types can be easily inherited from existing ones.
I've other things I could moan about, but I don't want to see C++ done an injustice - it performs well with many real problems, and I would like to see it developed in such a way as to eliminate as many of its idiosyncrasies as possible.
I am less concerned with the idea of centrally stored data than other readers, though I too remain cynical about those who plan to introduce it. To my mind, a more significant problem than some nefarious character easily discovering facts about me (which are already in public record) are the array of information collecting agencies, whose data is less easily available - even to those to whom it refers.
For example, credit reference agencies store highly personal details regarding use of, say credit cards, but without sufficient detail to give a complete picture of a customer, which can easily lead to discrimination against the unwary. For example, if you consider buying a car on credit, and agree to let several companies to make enquire, someone with an exemplary record can easily become a suspicious character as no record is made of their decision to decline an offer. There is a conflict of interest here, as the vendor would like customers it has chosen to accept to be considered bad risk elsewhere as this increases their chance of making a sale. A similar conflict arises with credit card companies. Capital-One has been in the news in the UK for delaying dispatch of statements, giving customers only a few hours to make payments on time. A late payment acts as a black mark against a customer on a central register, a number of which can prevent the customer switching to another lender. Hence, for both long and short term prosperity, credit card companies need to identify low risk customers, and to engineer a situation in which their customer makes late payments, as this involves large charges and also effectively ties in the customer for the foreseeable future. Amusingly, as assets are not part of a credit history, an individual's record can brazenly show multiple late payments on a £10,000 store card, yet fail to mention that this is in relation to a dispute over, say a £5.00 misapplied charge falsely giving them a balance, and woefully fails to mention a consistent large credit balance at the customer's bank. Furthermore, once the customer has been declined credit, this can be a further obstacle in the future - even should it be proven that the decision was unfounded.
I for one would welcome any transparent system - and would like to see a minimum requirement that agencies explicitly inform subjects when they store personal information; that the agency undertakes the burden of proof in cases of disputes, and that they are required to prove the accuracy of their information within a predetermined timescale. At present, in the UK (and I understand the USA too) the procedure is that individuals must contact the agency, paying a nominal charge, on any occasion they feel concerned, which is invariably too late. Naturally, the agency says it's only storing information provided to it by third parties, and if this information is wrong then the customer should take this matter up with the originator of the "information" who, naturally, insists that their version of events is accurate unless the customer can demonstrate that they adhered to reams of nebulous fine-print on every occasion in their distant history... nigh impossible!
Maybe I too would have dismissed the topic if I realised that it was still April Fools day in other continents.
Conversely, I don't see that a rapid seduction technique is the solution to the plight of technically minded thinking young men. From my perspective, computing has an atrocious image problem - something which should be addressed. I find it ridiculous to prejudge a "bright technical elite" as boring - which is what happens. You become a geek if you're interested in technological and scientific developments, work in the affluent IT industry and are capable of understanding technical issues. Why should this have to be a millstone? One hypothesis is for a society mainstream swayed by a technically inept minority who expect to benefit from discrimination. Another is that geeks become so consumed with passion for their subjects that they forget to wash themselves, dress appropriately, or to exercise common courtesy (I've yet to meet a discourteous "geek!") While I'm sure both versions have some grounding in truth, I find it distasteful to make sweeping generalisations, or to suggest there is a "quick fix" - something I can easily do to make my personality more affable to the fairer sex - frankly is ridiculous.
Anyhow, thanks for your well wishes which I'm sure will come in handy on my next chance encounter with an attractive cerebrally functional unattached girlie of my generation.
I was asked in my local pub last week if I was a geek, I replied
"Probably" - well I was drinking "Calsberg" - and the reaction I
received was fantastic! It was only after a few minutes I realised
this girl thought I'd declared some irredeemable perversion, which I
now suppose I was expected to keep hidden if it were true. By way of
clarification, my local pub is a small affair, where everyone knows
everyone... mixed sexes, however all the female patrons I find
attractive are much elder, and happily married!
A difficult question I've been asked is "Why are you single?" and so
far I've only found one honest answer: "I'm more fussy than I am
attractive" - brutal but true. You see, I have met women with whom
I've fallen head over heals in love - and since the age of 16, have
been sufficiently confident to ensure at the very least a "date".
Unfortunately, on every (easily countable) occasion so far, things
were not "meant to be" and I'm convinced that no action on my part
could reverse the inevitable tragic result. You see, when I'm
convinced that I want a feminine non-smoking brunette librarian who's
interested in fine art, wines and architecture - I'm unlikely to find
such a person without her own similarly stringent expectations... and
if they're as resrtrictive as mine, the concequences of Bayesian theory
don't bear thinking about... especially as these days I seldom meet a
new person, let alone a new woman, given the unfortunate technology skew
and my own long hours.
A year or so ago I joined a local dating agency - and my friends
thought that this was well worth while, purely because they enjoyed
the stories which made most mainstream comedy seem pretty lame. I met
about half a dozen women before I failed to renew... As a dual
consequence of my wanting to avoid legal action and my own failing
memory, I'm inventing the names of the most memorable liaisons:
Terrible Tracy sat with me outside her favourite pub - I
realised something was wrong when she insisted on staring at a fence
post for half an hour - I'm not sure if she was more bored or boring -
sort of pretty - but absolutely no cerebral activity - It respires,
but its not life as we know it. Then there was Nancy, the Nazi Nanny,
who sticks in my mind for making it quite clear that she believed that
children should be beaten as a matter of course... which I just had
to take seriously as she continued to murmur right-wing propaganda
through her grubby chunky jumper. Of course I shouldn't forget Slobby
Suzie, whose resounding line was "You're too classy for me!" I have
to admit that I hope I was. I refrained from verbalising that a bag
of cut price, extra lard chips and a quick shag behind the bus shelter
was probably too classy for her. My final example is Ladette Lucy, and
we just can't go there - it was set up that we should meet in a bar,
but as soon as Lucy entered the room I was convinced she was male...
It is my hunch that Lucy was actually in the early stages of
trans-sexual changes... I had no wish to investigate further... so
bluffed for an hour, keeping a careful emotional and physical
distance, then left. With that my brush with random blind dates came
to a close...
These days, I think I can safely say that a different approach is
required. A change of social circle would be good - my neighbours are
great people, but are all drawing pensions. The females in my local
pub are attached, double my age or uninspiring (though rarely all three) and at the office I'm
surrounded by "Geeks" - a phrase I use loosely in the most
affectionate of ways. So my "ask slashdot" has to be for suggestions
as to where to meet fantastic women considering my intense distaste
for most sport (or conversely - where would women look for someone
fantastic like me?) So far I've considered Salsa dancing (Pro:
Inverse gender skew Con: I dance like a pregnant duck suffering from a
bad hair day) and a local humanist society (about which I know very
little)...or... has anyone had dates turn out to be bigger prunes?
I have to disagree with the popular idea that these days "everyone is a computer person" - which is clearly the intended sentiment. I find computing fascinating for exactly the converse reason. A good computer person (be that computer scientist, programmer or whatever) is someone who is able to command a comprehension of a wide range of disciplines and identify informatic problems in foreign subjects and to establish solutions.
Re:Who cares about performances with Winblows98 ?
on
Emergence of SMT
·
· Score: 1
These days, typically only ardent graphicophiles require a powerful machine on their desk, given a high bandwidth network connection and the availability of suitable shared resources. Sure it can be handy for a developer to sit next to his obscenely over-specified machine, but similar results can be achieved in different ways.
Unfortunately, a similarly dismissive attitude is misplaced when considering server architectures. As real-life presents opportunity to build ever larger scale interactive systems, the speed of execution of critical sections becomes every more important.
There are more useful questions to ask than "What proportion of users browse with IE on Windows" - for example
1. What proportion of users use unconnected (i.e. no ISP/LAN) machines?
2. What proportion of users require to use client server software to access centralised resources?
3. Are users satisfied with the performance of servers when the access shared resources at times of high contention?
Or... in a much more down to earth fashion - do these users of IE also use Exceed (or other win32 X-server), or telnet?
Here in England, when I did my undergrad, these two strands were Computer Science (resulting in a BSc) or Computer Systems Engineering (resulting in a BSc or BEng at the candidates discretion). I was a CS student, but used my "optional courses" to take CSE courses in signal processing. The tricky aspect to CS was that coursework was less well defined, and the courses contained more algebraic-maths. Conversely, CSE was more "practical" and lectures had more of a "how to make things work" feel to them. I wholeheartedly agree with previous posts suggesting that you should start courses to "suck it and see" but my word of warning would be that neither CS nor CSE are primarily concerned with training programmers, or even "software engineers" and in my experience most practical knowledge is self-learned by graduates from all disciplines. It does help, however, to have a sound background in how computer systems work, as it helps you avoid some pitfalls less obvious to the empirically self-taught.
I would be interested to hear from any corporate users who honestly state that the long-term effects of a virus were positive. It's hard to deny that viruses have a massive potential cost, not limited only to diminished direct productivity and lost data, but more significantly with regard to credibility of the organisation. That aside, I suspect users most at risk from viruses often coincide with those who would benefit most from a "clean start."
It would seem to me, that some organisations for whom a skirmish with a virus might be just what is required to provoke adoption of better practices. How can you factor in the reduced costs associated with an ability to recover from a hardware failure a year later? Is it better to suffer from a virus sooner or later? Can anyone associate a cost with the technically inept associating divine accuracy of anything on-screen? Conversely, what long term costs are incurred when people actively avoid technology for fear of its inherent problems? [Remind me again, what were your virus costs, Mr. Norton?]
I follow, but feel that you are taking me out of context. When I brazenly stated that lower utilisation results in lower latencies, I accept that this requires several (not unrealistic) assumptions. I agree whole-heartedly that it is possible to statically schedule without loss of performance (providing that capacity exceeds demand), and can see your perspective when you state that latencies are not lower in synchronous systems.
My fundamental objection is that the overwhelming majority of communications systems, and all general-purpose systems, are not synchronous... and for good reason. While we see a proliferation of synchronous transmission media, we must also consider the nature of the processes on the computers effecting communication over these networks, which are almost universally asynchronous, not to mention the effects of truly external events, such as communication triggered by user actions.
An interesting concept I've previously considered myself... about which I'd like to make a few observations:
The paper correctly asserts that TCP as a transport introduces significant overheads to IPC in distributed systems, but glosses over the detail. The generic nature of TCP assumes that the underlying network technology is imperfect, which as far as I can tell, the WDD does not. Hence, the cost of TCP's ability to tolerate failures is an overhead of protocol processing to establish a reliable stream transport. TCP is not ideal for high performance message passing systems, but this should not come as any surprise - just look at the inefficiencies inherent in the HTTP protocol (particularly early versions).
It would be fairer to compare the likely performance of the WDD with a messaging system based upon a transport without retransmission support - message passing over UDP being an obvious contender. Here, protocol processing is drastically reduced, but a new communications issue must be raised: what happens if a packet is lost in-transit owing to corruption? (I admit that this is far less likely in typical optical networks of today, than yesterdays less robust CSMA/CD offerings, but this problem must be considered, particularly if the stability of a large scale distributed processing system depends upon it!
When I read the WDD proposal I see a clear comparison with token ring systems, the only significant difference being that the "transmitting node" is able to pre-empt what data requires to be sent. To me, the storage capacity of a WDD renders it more comparable to a shared memory,. This highlights a need for a strategy to decide what to store in the shared medium, and what to store on traditional media... a page (packet) replacement algorithm if you will. A successful PRA would present a considerable performance improvement, as it would significantly reduce latency - the curse of distributed computing! Conversely, such a PRA could equally well be applied to all communications, as it effectively presents a mechanism to pre-empt what data requires transmission, and everyone likes clairvoyant systems.
This aside, my primary objection to the usefulness of this sort of facility is that the authors assume that a WDD would present a use for spare bandwidth. I would argue that there is no such thing. When utilisation of a shared resource (the transport) is low, then latencies are lower, and this is of fundamental importance to the performance of processes interacting over a communications channel. Cf. write-times on an empty vs. full hard disk:-)
To summarise - an interesting idea from which I can imagine many practical spin offs, but surely it is frivolous to suggest that this will revolutionise computer communication or shared data storage.
Respect is a strange fish, all told, which comes in many shapes and sizes. Respect means something completely different to different people. Is "respect" found in the tribute paid to a mafia boss, or in the reverence you may hold for a local Pasteur? It seems to me that this overloaded term is not particularly helpful. Technology too is a funny game - how many top executives need the help of toddlers to tune their video recorders?
It's my opinion that the real crux of this matter is not whether or not you are respected by your employers, but rather an issue of personal relationships - a different matter altogether. Its fine and large to be on-the-ball and technically able, however don't be fooled - being diligent, reliable and able to provide timely answers is only a small part of the battle, so to speak.
Have you heard the anecdote where a 7-year-old girl asks: "Daddy, what's politics about?" After some thought he replies thoughtfully "Politics is about people and persuading them to do what you want them to." A few seconds later: "I see Daddy... a bit like sexual politics then?" I suspect that you are experiencing a similar, if somewhat less dramatic problem of cultural difference. We all know that the pace of modern technological change is phenomenal - the significance of this is something a young person can easily overlook as a result of a limited historical perspective - whereas a wrinkley will have greater context for their subjective responses.
I wonder if you have (professionally) personal relationships with your work colleagues? [No... I'm not suggesting you try to bed the secretary... or maybe that would be a good idea anyhow - never mind.] In some circumstances, you might need to learn to play golf, in others you need to "do lunch" from the right place. Find some non-work related common ground and get the true size of these people. From my experiences, the disrespect to which you allude tends to result from the fears of those who expect to remain senior, coupled with the less honed social skills of the whippersnapper.
I wish you the best of luck with your job and a world of other fascinating things you need to be experience.
In line with previous replies, beauty in code certainly is in the eye of the beholder, or at least in the realm of the problem domain. To my mind the only definition of beautiful code is code without ugly features, which are somewhat easier to define. To my mind, the principle causes of ugly code are:
Failing to meet (nebulous) requirements fully
Inappropriate (or even worse, misleading or contradictory) documentation/comment.
Inappropriate encapsulation
Exhibiting redundant source code
Inconsistency of notational style and inappropriate naming conventions
Employing inappropriate algorithmic techniques
So, logically speaking beautiful code should emerge when the above errors are avoided for an interesting problem. The first point is trivial and has been covered by other replies "The software must work". The second point is dear to my heart - absence of commentary is bad, but banal, contradictory or otherwise invalid annotation is far worse! Encapsulation must be considered a central concept in any software engineering effort, and this black art should be considered fundamental. Redundant code is misleading and can have performance implications too. Adherence to house style for code is a considerable cause of strife when programmers collaborate... however a simple consistent style is important to avoid confusion. Finally, employing appropriate techniques form the conceptual link between beautiful code and beautiful programs.
A while ago I read a great book: Programing Pearls, which discussed some of the most elegant aspects of programing the author had encountered (along with historical context). It's a great read and in my opinion highlights some fascinating historic revelations in software development... by no means a "how-to" book, nor up-to-date, but enlightening non the less.
Slashdot Mirror
The real hope for open source and free software is that MS are forced to publicly release their protocols some time before they release their products, as this presents new opportunities for other players to move into the market of services, which is (IMHO) the holy grail for commercial computing software. Linux and friends don't need to be on the desktop in the foreseeable future in order to make their presence felt - providing that MS clients can interact seamlessly with non MS servers. In the long term, I can envisage a situation in which new competition may arise to compete solely for the desktop user interface, which will only be possible once operational issues are completely decoupled from the end user interface.
I've wondered for quite some time why it is not a standard feature for a POP account to decline mail which is not addressed to one of a user-defined set of e-mail address regexps. The vast majority of my spam doesn't even mention my name in the address... and if it wasn't sent to me then I don't want to know! I realise this would require additional configuration for each user (particularly if you frequently join/leave mailing lists, but I for one would immediately see the benefits if I didn't need to wait downloading such obvious spam over dialup!
That said, as a matter of professional interest, C/C++ is a serious option as far as computationally demanding applications are concerned (e.g. server side work, systems programming, medium scale embedded systems etc.) - This is likely to be particularly valuable as it is within this paradigm that a vast quantity of existing software has been written, but IMHO anyone deciding to use C/C++ to write GUIs these days require their head examining. As mentioned in previous posts, Java is a serious player in the marketplace, and I was very impressed with Jbuilder 4.0, the least comprehensive version of which is freely available. C# has also been mentioned, and while I feel this development is not without merit, I am concerned that its proprietary status may draw programmers into the same "dead ends" as VB. Python, to me, falls into the same basket as TCL/TK did several years ago - a scripting language which is not directly comparable with the other suggestions here.
If this post is asking about "cool" languages to try out, you could go much further astray than to consider ML, (an established and capable functional language) or maybe Smalltalk There are a whole host of alternatives out there, and it's silly to pick your language before you pick your problem.
The line "the company rep said 'because other companies would try to steal our employees'." to me sounds comical. Surely this is no more than a lame attempt to hype their latest efforts? Lets face it, what self respecting employer would ever consider a candidate worthwhile because they've been on TV? Conversely, generating an air of secrecy and excitement may prove a productive ploy - and has the desirable side effect of masking a bunch of callow youths, who quite possibly look far more impressive incognito - especially if any of them sport the western programmers' appearance after a long night of code cranking.
The reason why I believe this will work is the social hypothesis that many hackers are principally motivated by technical challenge rather than a desire for anarchic destruction. It would seem sensible to embrace the efforts of such people by providing a safe mechanism by which they can showcase their talents on live systems (which should be assumed to be under attack in any case.) If this were to be coupled with a standard strategy for distributing patches and updates to software, then it would seem harmless enough to allow "hackers" to keep tally of the number of systems each "vaccine virus" had successfully attacked, and which now had the security deficiency patched - thereby replacing the drudgery of securing systems with an engaging game which could be played by anyone with sufficient time, knowledge and inclination. Finally, an opt-in strategy would ensure the ostriches of this world are not offended, and in a sense 'legalise' non-malicious hacking.
The reason? These foreign students took their grades very seriously (they were paying for this course themselves) and attended all the available tutorials hoping to gain an edge over their fellow students. In one of these optional tutorial sessions, an identical problem was used as a worked example, and the students present took copious notes. While phrase matching may assist an investigation into cheating, I feel it is vital that academics don't fall into the popular trap of believing that everything churned out by a computer is necessarily the whole story. Unfortunately, as the quality of automated systems improves, so does misplaced confidence.
I would only be interested in the argument/information contained within posts, and the authorship is irrelevant to me. I realise that removing all names would be problematic as other contributors may have chosen to embed others' names into their articles, but surely it would be a simple compromise to strip headers and trailing signatures from well formed news posts?
1) Standardised ordinals - signed and unsigned - of known platform independent sizes - maybe we could have something like int8, int16, int32, int64, nat8, nat16, nat32 and nat64?
2) A mechanism by which overflow can be detected when performing arithmetic on numbers (particularly for ordinals above)
3) Run-time polymorphism - not only would this satisfy language zealots who claim it's impossible to implement the identity operation as a compiled C++ function - but would also make it possible to easily implement more flexible containers.
4) A type system to constrain template instantiations. When I template a class with another, I want to be explicit about what behaviours my template parameter type should support, and to have this aspect of my program statically verified.
5) An improved scheme for enumerations whereby new enumerated types can be easily inherited from existing ones.
I've other things I could moan about, but I don't want to see C++ done an injustice - it performs well with many real problems, and I would like to see it developed in such a way as to eliminate as many of its idiosyncrasies as possible.
For example, credit reference agencies store highly personal details regarding use of, say credit cards, but without sufficient detail to give a complete picture of a customer, which can easily lead to discrimination against the unwary. For example, if you consider buying a car on credit, and agree to let several companies to make enquire, someone with an exemplary record can easily become a suspicious character as no record is made of their decision to decline an offer. There is a conflict of interest here, as the vendor would like customers it has chosen to accept to be considered bad risk elsewhere as this increases their chance of making a sale. A similar conflict arises with credit card companies. Capital-One has been in the news in the UK for delaying dispatch of statements, giving customers only a few hours to make payments on time. A late payment acts as a black mark against a customer on a central register, a number of which can prevent the customer switching to another lender. Hence, for both long and short term prosperity, credit card companies need to identify low risk customers, and to engineer a situation in which their customer makes late payments, as this involves large charges and also effectively ties in the customer for the foreseeable future. Amusingly, as assets are not part of a credit history, an individual's record can brazenly show multiple late payments on a £10,000 store card, yet fail to mention that this is in relation to a dispute over, say a £5.00 misapplied charge falsely giving them a balance, and woefully fails to mention a consistent large credit balance at the customer's bank. Furthermore, once the customer has been declined credit, this can be a further obstacle in the future - even should it be proven that the decision was unfounded.
I for one would welcome any transparent system - and would like to see a minimum requirement that agencies explicitly inform subjects when they store personal information; that the agency undertakes the burden of proof in cases of disputes, and that they are required to prove the accuracy of their information within a predetermined timescale. At present, in the UK (and I understand the USA too) the procedure is that individuals must contact the agency, paying a nominal charge, on any occasion they feel concerned, which is invariably too late. Naturally, the agency says it's only storing information provided to it by third parties, and if this information is wrong then the customer should take this matter up with the originator of the "information" who, naturally, insists that their version of events is accurate unless the customer can demonstrate that they adhered to reams of nebulous fine-print on every occasion in their distant history... nigh impossible!
Conversely, I don't see that a rapid seduction technique is the solution to the plight of technically minded thinking young men. From my perspective, computing has an atrocious image problem - something which should be addressed. I find it ridiculous to prejudge a "bright technical elite" as boring - which is what happens. You become a geek if you're interested in technological and scientific developments, work in the affluent IT industry and are capable of understanding technical issues. Why should this have to be a millstone? One hypothesis is for a society mainstream swayed by a technically inept minority who expect to benefit from discrimination. Another is that geeks become so consumed with passion for their subjects that they forget to wash themselves, dress appropriately, or to exercise common courtesy (I've yet to meet a discourteous "geek!") While I'm sure both versions have some grounding in truth, I find it distasteful to make sweeping generalisations, or to suggest there is a "quick fix" - something I can easily do to make my personality more affable to the fairer sex - frankly is ridiculous.
Anyhow, thanks for your well wishes which I'm sure will come in handy on my next chance encounter with an attractive cerebrally functional unattached girlie of my generation.
A difficult question I've been asked is "Why are you single?" and so far I've only found one honest answer: "I'm more fussy than I am attractive" - brutal but true. You see, I have met women with whom I've fallen head over heals in love - and since the age of 16, have been sufficiently confident to ensure at the very least a "date". Unfortunately, on every (easily countable) occasion so far, things were not "meant to be" and I'm convinced that no action on my part could reverse the inevitable tragic result. You see, when I'm convinced that I want a feminine non-smoking brunette librarian who's interested in fine art, wines and architecture - I'm unlikely to find such a person without her own similarly stringent expectations... and if they're as resrtrictive as mine, the concequences of Bayesian theory don't bear thinking about... especially as these days I seldom meet a new person, let alone a new woman, given the unfortunate technology skew and my own long hours.
A year or so ago I joined a local dating agency - and my friends thought that this was well worth while, purely because they enjoyed the stories which made most mainstream comedy seem pretty lame. I met about half a dozen women before I failed to renew... As a dual consequence of my wanting to avoid legal action and my own failing memory, I'm inventing the names of the most memorable liaisons: Terrible Tracy sat with me outside her favourite pub - I realised something was wrong when she insisted on staring at a fence post for half an hour - I'm not sure if she was more bored or boring - sort of pretty - but absolutely no cerebral activity - It respires, but its not life as we know it. Then there was Nancy, the Nazi Nanny, who sticks in my mind for making it quite clear that she believed that children should be beaten as a matter of course... which I just had to take seriously as she continued to murmur right-wing propaganda through her grubby chunky jumper. Of course I shouldn't forget Slobby Suzie, whose resounding line was "You're too classy for me!" I have to admit that I hope I was. I refrained from verbalising that a bag of cut price, extra lard chips and a quick shag behind the bus shelter was probably too classy for her. My final example is Ladette Lucy, and we just can't go there - it was set up that we should meet in a bar, but as soon as Lucy entered the room I was convinced she was male... It is my hunch that Lucy was actually in the early stages of trans-sexual changes... I had no wish to investigate further... so bluffed for an hour, keeping a careful emotional and physical distance, then left. With that my brush with random blind dates came to a close...
These days, I think I can safely say that a different approach is required. A change of social circle would be good - my neighbours are great people, but are all drawing pensions. The females in my local pub are attached, double my age or uninspiring (though rarely all three) and at the office I'm surrounded by "Geeks" - a phrase I use loosely in the most affectionate of ways. So my "ask slashdot" has to be for suggestions as to where to meet fantastic women considering my intense distaste for most sport (or conversely - where would women look for someone fantastic like me?) So far I've considered Salsa dancing (Pro: Inverse gender skew Con: I dance like a pregnant duck suffering from a bad hair day) and a local humanist society (about which I know very little) ...or... has anyone had dates turn out to be bigger prunes?
I have to disagree with the popular idea that these days "everyone is a computer person" - which is clearly the intended sentiment. I find computing fascinating for exactly the converse reason. A good computer person (be that computer scientist, programmer or whatever) is someone who is able to command a comprehension of a wide range of disciplines and identify informatic problems in foreign subjects and to establish solutions.
Unfortunately, a similarly dismissive attitude is misplaced when considering server architectures. As real-life presents opportunity to build ever larger scale interactive systems, the speed of execution of critical sections becomes every more important.
There are more useful questions to ask than "What proportion of users browse with IE on Windows" - for example
1. What proportion of users use unconnected (i.e. no ISP/LAN) machines?
2. What proportion of users require to use client server software to access centralised resources?
3. Are users satisfied with the performance of servers when the access shared resources at times of high contention?
Or... in a much more down to earth fashion - do these users of IE also use Exceed (or other win32 X-server), or telnet?
Here in England, when I did my undergrad, these two strands were Computer Science (resulting in a BSc) or Computer Systems Engineering (resulting in a BSc or BEng at the candidates discretion). I was a CS student, but used my "optional courses" to take CSE courses in signal processing. The tricky aspect to CS was that coursework was less well defined, and the courses contained more algebraic-maths. Conversely, CSE was more "practical" and lectures had more of a "how to make things work" feel to them. I wholeheartedly agree with previous posts suggesting that you should start courses to "suck it and see" but my word of warning would be that neither CS nor CSE are primarily concerned with training programmers, or even "software engineers" and in my experience most practical knowledge is self-learned by graduates from all disciplines. It does help, however, to have a sound background in how computer systems work, as it helps you avoid some pitfalls less obvious to the empirically self-taught.
It would seem to me, that some organisations for whom a skirmish with a virus might be just what is required to provoke adoption of better practices. How can you factor in the reduced costs associated with an ability to recover from a hardware failure a year later? Is it better to suffer from a virus sooner or later? Can anyone associate a cost with the technically inept associating divine accuracy of anything on-screen? Conversely, what long term costs are incurred when people actively avoid technology for fear of its inherent problems? [Remind me again, what were your virus costs, Mr. Norton?]
My fundamental objection is that the overwhelming majority of communications systems, and all general-purpose systems, are not synchronous... and for good reason. While we see a proliferation of synchronous transmission media, we must also consider the nature of the processes on the computers effecting communication over these networks, which are almost universally asynchronous, not to mention the effects of truly external events, such as communication triggered by user actions.
The paper correctly asserts that TCP as a transport introduces significant overheads to IPC in distributed systems, but glosses over the detail. The generic nature of TCP assumes that the underlying network technology is imperfect, which as far as I can tell, the WDD does not. Hence, the cost of TCP's ability to tolerate failures is an overhead of protocol processing to establish a reliable stream transport. TCP is not ideal for high performance message passing systems, but this should not come as any surprise - just look at the inefficiencies inherent in the HTTP protocol (particularly early versions).
It would be fairer to compare the likely performance of the WDD with a messaging system based upon a transport without retransmission support - message passing over UDP being an obvious contender. Here, protocol processing is drastically reduced, but a new communications issue must be raised: what happens if a packet is lost in-transit owing to corruption? (I admit that this is far less likely in typical optical networks of today, than yesterdays less robust CSMA/CD offerings, but this problem must be considered, particularly if the stability of a large scale distributed processing system depends upon it!
When I read the WDD proposal I see a clear comparison with token ring systems, the only significant difference being that the "transmitting node" is able to pre-empt what data requires to be sent. To me, the storage capacity of a WDD renders it more comparable to a shared memory,. This highlights a need for a strategy to decide what to store in the shared medium, and what to store on traditional media... a page (packet) replacement algorithm if you will. A successful PRA would present a considerable performance improvement, as it would significantly reduce latency - the curse of distributed computing! Conversely, such a PRA could equally well be applied to all communications, as it effectively presents a mechanism to pre-empt what data requires transmission, and everyone likes clairvoyant systems.
This aside, my primary objection to the usefulness of this sort of facility is that the authors assume that a WDD would present a use for spare bandwidth. I would argue that there is no such thing. When utilisation of a shared resource (the transport) is low, then latencies are lower, and this is of fundamental importance to the performance of processes interacting over a communications channel. Cf. write-times on an empty vs. full hard disk:-)
To summarise - an interesting idea from which I can imagine many practical spin offs, but surely it is frivolous to suggest that this will revolutionise computer communication or shared data storage.
It's my opinion that the real crux of this matter is not whether or not you are respected by your employers, but rather an issue of personal relationships - a different matter altogether. Its fine and large to be on-the-ball and technically able, however don't be fooled - being diligent, reliable and able to provide timely answers is only a small part of the battle, so to speak.
Have you heard the anecdote where a 7-year-old girl asks: "Daddy, what's politics about?" After some thought he replies thoughtfully "Politics is about people and persuading them to do what you want them to." A few seconds later: "I see Daddy... a bit like sexual politics then?" I suspect that you are experiencing a similar, if somewhat less dramatic problem of cultural difference. We all know that the pace of modern technological change is phenomenal - the significance of this is something a young person can easily overlook as a result of a limited historical perspective - whereas a wrinkley will have greater context for their subjective responses.
I wonder if you have (professionally) personal relationships with your work colleagues? [No... I'm not suggesting you try to bed the secretary... or maybe that would be a good idea anyhow - never mind.] In some circumstances, you might need to learn to play golf, in others you need to "do lunch" from the right place. Find some non-work related common ground and get the true size of these people. From my experiences, the disrespect to which you allude tends to result from the fears of those who expect to remain senior, coupled with the less honed social skills of the whippersnapper.
I wish you the best of luck with your job and a world of other fascinating things you need to be experience.
In line with previous replies, beauty in code certainly is in the eye of the beholder, or at least in the realm of the problem domain. To my mind the only definition of beautiful code is code without ugly features, which are somewhat easier to define. To my mind, the principle causes of ugly code are:
So, logically speaking beautiful code should emerge when the above errors are avoided for an interesting problem. The first point is trivial and has been covered by other replies "The software must work". The second point is dear to my heart - absence of commentary is bad, but banal, contradictory or otherwise invalid annotation is far worse! Encapsulation must be considered a central concept in any software engineering effort, and this black art should be considered fundamental. Redundant code is misleading and can have performance implications too. Adherence to house style for code is a considerable cause of strife when programmers collaborate... however a simple consistent style is important to avoid confusion. Finally, employing appropriate techniques form the conceptual link between beautiful code and beautiful programs.
A while ago I read a great book: Programing Pearls, which discussed some of the most elegant aspects of programing the author had encountered (along with historical context). It's a great read and in my opinion highlights some fascinating historic revelations in software development... by no means a "how-to" book, nor up-to-date, but enlightening non the less.