I appreciate the link, but I think their example CV is horrible. (Speaking as someone whose mother is a professional careers advisor at a UK university, and so has had good CV writing drilled into him since forever...)
The parent is right on the money, except that I think the stat is much greater than 50% in my area (East Anglia, UK).
There are basically four ways to go about getting a job:
Use an on-line jobs board.
Send a resume and covering letter to likely companies.
Reply to job ads, or use an agent who serves much the same purpose.
Get a contact through networking and word-of-mouth.
In my experience, these are listed in increasing order of likelihood of success, and the first two options are way behind the other two.
My other advice would be to consider aiming for a small company first, particularly if you're good. You're much more likely to have someone technical read your resume and any covering letter you send, rather than to be filtered out by some buzzword-craving DB. If you write a good resume -- most people really don't, and I've posted advice on this subject around here before -- then so much the better.
You probably won't get a top notch salary at a small company, but you'll get a decent average for someone with your experience over the first year or two at most of them, and you'll get a much more personal experience from those you work for and with, which is good for developing your early career. Again, this is particularly useful if you really are good, either technically or in your attitude, as this is far more likely to be noticed in a smaller, more personal environment.
After a couple of years in the business, you'll have had chance to establish a solid track record with a company, and to see which skills are really useful and not just hype. If you choose to move on from there, you'll be much better placed than you are right now.
Final tip: do consider staying on and getting more qualified while the market is tough. NB: I'm mostly talking about serious qualifications, not random certificates from marketing departments, though the latter rarely hurt. I got a long way based not only on a good maths degree, but also on the one year postgrad diploma in CS I took to go with it. Aside from being a darned useful course, it distinguished me from other random graduates in my early career. If you can get some sort of funding or sponsorship to do such a course, so much the better, obviously. It gives you a way to ride out the current wave of poor IT recruitment, and good experience to boot.
If you're looking to do software development as a serious career, supporting skills in things like maths or management do no harm at all. If you're after sysadmin type work, you could do worse than having some electrical or communications engineering skills as well (and those random marketroid-driven certificates are probably worth something, at least in some cases). Either way, the extra edge does no harm.
Sorry, but you're wrong, too. The correct phrasing, at least in the UK, is "Would you like to go large with your fries?" Implication selling is a tried and tested technique, and works much better than telling your customer what to do. Just ask any used car salesman.;-)
That would be when you have the source code, dickwit.
And how, exactly, is that supposed to help? Do you expect everyone in the world who uses e-mail to learn to program in whatever language some client is written in, and then critically review the client code, and then rebuild it themselves to make sure the binary is legit?
And don't fucking piss and moan about 'not getting a virus from reading your email' in the context that the experts were discussing, the email was PLAIN TEXT. There was no executable code. Get the idea?
Exactly. It was plain text, and you could trust it, because there was no executable code. And then MS changed that, and everyone's expectations were suddenly dangerously wrong.
One of us is definitely missing the point here, but it's not me.
This gives you a brief description of each package that it will be downloading, and lets you decide whether or not to download/install this package. up2date give you far more control than the equivalent program in Windows XP.
Erm... You have exactly that much control in Windows XP. If updates are available, you are invited to download them, at which point you can follow links to relevant descriptions, KB articles, or what have you. You can then opt to install (or not) on a selective basis. And it checks for updates automatically as well.
I'm sorry, but your post is nothing but pure, unadulterated FUD.
I installed mozilla on my sister's machine... installed an IE-lookalike skin on mozilla and she hasn't noticed the difference yet. (It's been about a month now.)
Wow. Your IE must have been much more stable than mine, and your Moz much better at rendering badly formed pages.;-)
MS has only had a week or two with the knowledge of this bug (article mentions that MS learned in November aka this month some time). For such a huge exploit, I'd suspect it'll take a week to pinpoint the code error, a week to fix the code, and two to four weeks of testing it.
That seems way over the odds to me. I've spent the last couple of weeks fixing several bugs in a product within about 24-48 hours, when all that was at stake was a business deadline. When I couldn't find the bug myself, I called on other members of the team to help out, but in all cases, we had a satisfactory solution well inside a week, and usually the same day.
Security flaws are usually caused by careless errors that could easily be prevented. They can often be fixed in a few minutes once identified, and tested shortly afterwards. Companies who provide widely used critical software like operating systems or communications tools really shouldn't have a problem getting things turned around within 24 hours. If they do, either their code is so screwed up that it's totally unsuitable for use in a potentially vulnerable environment (granted, Microsoft have actually made exactly this claim about several versions of Windows in recent months) or they seriously need to reconsider how they run their response to security vulnerabilities that are reported.
Of course not, and bitching at APIs and scripting languages is retarded. If you choose to run untrusted code, then You Have Lost, on Windows or on Unix.
And how, exactly, do you know when you're running "trusted code"? For years, security experts recited the mantra that you couldn't get an e-mail virus just from reading your mail, and you had to actually run an attachment to get infected. Then MS screwed up with the scripting in things like Outlook (Express), and suddenly all the non-techies in the world, trusting their techie colleagues about the virus thing, are getting caught. Whose fault is this? I sure as hell don't want or expect to run any code automatically just because it's part of an e-mail I'm reading, but MS left me no choice if I use that product, and of course many have no choice about that, either.
No, I think bitching about scripting and APIs that let code run on my box when I neither want nor expect it to is quite justified, thanks very much.
You make a very convincing argument, as many pro-gun-control people do. Unfortunately, most of it is totally unsupported by the facts available, which is invariably where such arguments break down.
Muggers do not generally shoot first. The cycle of hate you describe, where surviving muggers learn their lesson and get more vicious next time, doesn't generally happen. Disarming everyone might be a great idea, except that it's totally impractical (said the poster from the UK, where only criminals have guns and gun crime is up 50+% since they banned the few handguns that were allowed before).
I have reservations about a massively pro-gun attitude, such as many US citizens appear to have, but I don't think mass disarming is even remotely plausible from your current starting point. That being the case, I'm more in favour of letting everyone have basic firearms than just the bad guys.
"If they can use it, they can steal it, therefore they should pay without having the privilidge of using it, so that we can continue to create more stuff along the lines of Brittney Spears"
And what, exactly, is wrong with having more stuff along the lines of Britney? I think she's cute, and one of her isn't enough to go around...;-)
So basically, providing non-comprehensive documentation seems to make it more likely you'll get hired again. Interesting.
It will also, given time, turn your reputation to shit. You will not survive forever supporting a small-scale development you once did, and no-one sane will hire you to write a larger one in the future, knowing that you'll tie them in. If board members of the biggest companies in the world are questioning when MS does this, you sure as hell aren't going to get far as a single contractor that way.
OTOH, if you do a good job, and leave it well-documented and maintainable, then (a) you'll have a much more pleasant time if you're the guy doing that maintenance a year later, and (b) you're going to develop a reputation as someone who does a good job, which is about the most valuable thing you can possibly have in the contracting business.
Most distros come with automout support, too, so you don't have to manually mount your removable media.
This is exactly the problem: you think it's impressive that most Linux distros have now reached the point of overcoming a problem that Windows never even had.
While I understand your perspective as someone who presumably likes Linux, a quick look down this thread shows how willing Linux enthusiasts are to overlook things that would put most potential converts off in a heartbeat:
I do not want to run Neverwinter Nights in a few weeks, maybe. I wanted to run it six months ago, when it came out.
I do not want to buy a new hardware modem and a PostScript printer. My Winmodem and GDI printer already work quite happily.
I do not want to go through X Configuration Hell(TM) to get fonts that look almost as good as Windows. I want fonts that look as good as Windows, out of the box.
I do not want to rely on an office suite that exports something 95% compatible with Word. I want a Word format document, because that's what's running at work. No, PDF will not do as an alternative.
I do not want to run Mozilla all the time, and give up using several very helpful web sites that happen only to work in IE. On my Win98 box, which I'm using now, I run Moz as my primary browser, but IE is just a click away if a site doesn't work in it.
Do pro-Linux people see the common thread here? Linux is mostly good at doing most of what Windows can do within a few months of being able to do it on Windows, with mostly compatible results. To a hobbyist geek, like many of us, that might be good enough, and we're prepared to put up with some modest inconvenience in exchange for the benefits you get from using Linux. However, to most Windows, users, it just ain't good enough.
Not Open Source, but there's always Borland Kylix, which now supports C++. Its debugger has all of the features you'd expect.
Does it? I haven't used anything else seriously for a while, but at the last count, Visual Studio's debugger was comfortably ahead of the opposition on useful features. Granted that in their quest to.NETify everything they've stuffed up a couple of the basics for vanilla C++ programmers like me, but still, the little things make all the difference. For example, in the VS debugger, you have the AUTOEXP.DAT file to customise how variables of user-defined types are displayed. Does the Kylix debugger have such things these days?
(This is a serious question, not a troll. I've always quite liked Borland's dev apps, but haven't used C++ Builder seriously for a good 3-4 years now, whereas I've had the latest and greatest(?) Visual Studio at work all the time. I'd like to know if Borland really have caught up.)
It's not a question of "can't be", it's a question of "doesn't benefit from being".
An open source approach has produced some great developments, but almost universally when:
it's a widely useful project (OS, office suite, common networking tools, etc.)
there is no alternative already available for an affordable price and with no major drawbacks
a community has developed, resulting in enough people contributing to get some serious work done (including developers, but also users, testers, etc.)
there have been talented project leaders and early developers to get the ball rolling and keep it rolling on a sensible path.
If you meet these criteria, then OS may be the way forward. If not, maybe you need to look at why so many people still do things the old-fashioned way.
You need open installation, open distribution, open setup, open guards, open data transmission/collection and open results. Otherwise there is no assurance.
<sigh> You'll be wanting open government next...
General purposes: relational vs OO
on
Design Patterns
·
· Score: 2
Regarding "general purpose". OO is no more general purpose than databases are. [...] I don't think the percentage of projects/domains that OO helps is significantly more than those which can make nice use of RDB's.
Now that I have to take issue with. There are, as we have demonstrated in this thread and elsewhere, numerous fields where databases simply aren't a useful programming tool. There are very few fields where OO is not applicable if you choose to use it, though there may be specialist tools that would do the job better. (What are those tools written in, though?)
I don't really want to get into this whole argument all over again, but just consider: I could sit down, right now, with your typical OO development system, and write you a database (either an actual DBMS or a custom biz app based on a database), and I could do this without using any Oracle, MySQL, or whatever. I can therefore do anything a database can, though perhaps not as cleanly until I've built the tools to reinvent a database.
On the other hand, I rather suspect that you couldn't sit down with your typical database package, right now, and write me an OO compiler. In fact, I rather suspect that you couldn't sit down with your typical database package, right now, and write any of the applications I've ever worked on.
This in itself should tell you something about the relative expressive power of the two approaches, and how general-purpose they really are.
They way content is delivered (aka "Style") is a subtancial part of the content itself. If I write an Essay and print it out in the Font "Impact" you'll know what I mean.
Ironically enough, you make my point for me. If you don't like an essay that's printed in Impact, you can change the style to print it in a more readable font without affecting the actual content at all. The presentation is improved, without changing any of the information contained within that presentation.
Re:All data structures are *not* equal!
on
Design Patterns
·
· Score: 2
Is your major argument that people are now starting to try to write what are essentially database apps using only general purpose OO languages, and not database-specific tools?
I am not sure what you mean by "database app". The DB engine itself, or an app that *uses* an existing database or database engine?
I meant what you call "custom biz apps", which basically seem to be databases with some UI on top.
I don't agree [that OO is a general-purpose approach]. I suppose anything that is Turing Complete could be called "general purpose", but I don't agree that OO philosophy is the best fit for every problem space.
I didn't say OO philosophy was the best fit for everything. I said it's a general-purpose approach, as in, it can be used to solve problems in many domains, not one specific one. Clearly a domain-specific tool, such as a database, will normally be more helpful for solving a problem in its own specific domain than any general-purpose approach. The smart programmer identifies any domain-specific tools that will help with his task, and then typically uses those in combination with general-purpose techniques, the latter providing functionality where no more appropriate tool is available, and providing glue between everything.
Re:All data structures are *not* equal!
on
Design Patterns
·
· Score: 2
Take the "structure taxonomies" that are popular with OO fans: stack, queue, set, bag, hash, dictionary, etc. Are these really "structures" or "protocols"?
There are data structures: various types of array, linked list, tree, etc. Then there are interfaces used to abstract those data structures: stack, queue, support of iteration to various degrees, etc. Logically enough, the latter are commonly referred to as "abstract data types". All ADTs must be implemented using some sort of data structure(s) under the hood, but a stack interface could be (and often is) implemented using a variety of different data structures depending on the particular performance characteristics needed for the problem at hand.
So if something wants to process the nodes independently of nodes created via a "stack" protocol (outside of push and pop), then that *disqualifies* it from being a "stack"? I find that a bit abrupt.
That is irrelevant. You cannot look inside a stack using the stack interface. I defy you to cite a single authoritative source anywhere in computing literature whose definition of a stack allows this.
For example, the subroutine calling "structure" in an interpreter is often considered a "stack", yet the debugger can see into the middle of it, and perhaps change stuff in the middle. Does this X-ray cabability disqualify it in your mind?
It is commonly referred to as a "call stack" because function calls work like a stack. When you call a function, you add a new function to the top of the stack. When you return from a function, you remove the function currently at the top of the stack. There is nothing else you can do. Even exceptions, as used in many languages, only perform the equivalent of a series of "pop" operations.
Obviously, a debugger can instrument what's going on in more detail if it wishes to do so, but it won't be using a stack interface to do it. What you see in the debugger window isn't being generated from the actual call stack, but from the debugger's own internal data structures.
I never wanted to change the tools of "low level" programmers. It just seems that their needs are ruining the tool choices of other domains which can take advantage of higher abstraction-based tools, such as relational modeling and protocols.
This is the thing I don't understand. Aside from disagreeing with your definition of "low-level" (is anything that is not a database high level in your terms?) I wonder what you mean when you write "ruining" here. If I wanted to write a database app, I would use database tools. Depending on what else it had to do, I might choose to write it in another language, or in multiple languages, but even then, I would use a database tool to do the database work.
Is your major argument that people are now starting to try to write what are essentially database apps using only general purpose OO languages, and not database-specific tools? If so, then I am sorry that I have missed your point for so long, and I agree entirely that that particular idea is idiotic. OO is a general-purpose approach, and it makes no sense to ignore database-specific tools if databases are what you're doing. This doesn't mean you can't write parts of your program in an OO language, of course, just that it would be sensible to control a database from your program to do the database work, and not to try creating a whole specific database in OO terms to mimic the same effect.
And to all the freaks here jumping to conclusions about good-looking sites == no quality:
Um... Who was that?
I've seen lots of people indicate that they prefer content to style. (The/. population might well not be representative of the overall population in this respect, of course.) I don't think I've seen anyone suggest that a quality site can't be good looking as well, though.
Is it just me or does everyone interpret sites with tiny font sizes as credible? (The article in point, too.)
On the contrary. Like you, I regard sites that use small font sizes and fix them there using stylesheets as being irritating rather than useful. Like many computer users, I don't have perfect vision, and appreciate a clear layout that I can read without eye strain. I simply won't attempt to read through an article of any length in small type. If a site provides an alternative stylesheet, or defines its font sizes relatively so that I can adjust them to be comfortable, it scores bonus points. Otherwise, it had better hope it works with Moz, where I can override the "fixed" font size anyway.
I agree that any site that requires Flash or cookies gets a big thumbs down. I don't think it's the credibility that's harmed, though. It's the usability (principally because of the download times for us modem users in the Flash case).
If the same site let me see a simple front page with the same content but no Flash animation, so I can download it and use it some time today, then maybe I'd rate it very highly if it still gave me the information I wanted easily.
If a site provides reasonable defaults without cookies, which it should be able to do if it just uses them to store my preferences as they were intended, then again, WTP? If I want to store the preferences, I'll enable cookies for the site. If not, I won't.
Even ads don't damage a site's credibility in my eyes, if they're done responsibly. I don't mind a banner ad or two that support a page. I find pop-up ads irritating, but these mostly seem to be put there by web hosting companies rather than the actual authors of a page I'm reading, so I tend to discount them as well. The only ads thing that really hurts a page in my eyes is being nasty about it. If I visit a travel agent's site, and when I've finished I close the window to find seventeen different ad windows for holidays I haven't even asked about, I'm never going back.
I guess you could argue that even this last case is really usability, but there comes a point where a site is sufficiently hard to use, or disregards the feelings of its visitors to such an extent, that it becomes a credibility problem as well.
Slashdot Mirror
I appreciate the link, but I think their example CV is horrible. (Speaking as someone whose mother is a professional careers advisor at a UK university, and so has had good CV writing drilled into him since forever...)
The parent is right on the money, except that I think the stat is much greater than 50% in my area (East Anglia, UK).
There are basically four ways to go about getting a job:
In my experience, these are listed in increasing order of likelihood of success, and the first two options are way behind the other two.
My other advice would be to consider aiming for a small company first, particularly if you're good. You're much more likely to have someone technical read your resume and any covering letter you send, rather than to be filtered out by some buzzword-craving DB. If you write a good resume -- most people really don't, and I've posted advice on this subject around here before -- then so much the better.
You probably won't get a top notch salary at a small company, but you'll get a decent average for someone with your experience over the first year or two at most of them, and you'll get a much more personal experience from those you work for and with, which is good for developing your early career. Again, this is particularly useful if you really are good, either technically or in your attitude, as this is far more likely to be noticed in a smaller, more personal environment.
After a couple of years in the business, you'll have had chance to establish a solid track record with a company, and to see which skills are really useful and not just hype. If you choose to move on from there, you'll be much better placed than you are right now.
Final tip: do consider staying on and getting more qualified while the market is tough. NB: I'm mostly talking about serious qualifications, not random certificates from marketing departments, though the latter rarely hurt. I got a long way based not only on a good maths degree, but also on the one year postgrad diploma in CS I took to go with it. Aside from being a darned useful course, it distinguished me from other random graduates in my early career. If you can get some sort of funding or sponsorship to do such a course, so much the better, obviously. It gives you a way to ride out the current wave of poor IT recruitment, and good experience to boot.
If you're looking to do software development as a serious career, supporting skills in things like maths or management do no harm at all. If you're after sysadmin type work, you could do worse than having some electrical or communications engineering skills as well (and those random marketroid-driven certificates are probably worth something, at least in some cases). Either way, the extra edge does no harm.
Sorry, but you're wrong, too. The correct phrasing, at least in the UK, is "Would you like to go large with your fries?" Implication selling is a tried and tested technique, and works much better than telling your customer what to do. Just ask any used car salesman. ;-)
And how, exactly, is that supposed to help? Do you expect everyone in the world who uses e-mail to learn to program in whatever language some client is written in, and then critically review the client code, and then rebuild it themselves to make sure the binary is legit?
Exactly. It was plain text, and you could trust it, because there was no executable code. And then MS changed that, and everyone's expectations were suddenly dangerously wrong.
One of us is definitely missing the point here, but it's not me.
Erm... You have exactly that much control in Windows XP. If updates are available, you are invited to download them, at which point you can follow links to relevant descriptions, KB articles, or what have you. You can then opt to install (or not) on a selective basis. And it checks for updates automatically as well.
I'm sorry, but your post is nothing but pure, unadulterated FUD.
Wow. Your IE must have been much more stable than mine, and your Moz much better at rendering badly formed pages. ;-)
That seems way over the odds to me. I've spent the last couple of weeks fixing several bugs in a product within about 24-48 hours, when all that was at stake was a business deadline. When I couldn't find the bug myself, I called on other members of the team to help out, but in all cases, we had a satisfactory solution well inside a week, and usually the same day.
Security flaws are usually caused by careless errors that could easily be prevented. They can often be fixed in a few minutes once identified, and tested shortly afterwards. Companies who provide widely used critical software like operating systems or communications tools really shouldn't have a problem getting things turned around within 24 hours. If they do, either their code is so screwed up that it's totally unsuitable for use in a potentially vulnerable environment (granted, Microsoft have actually made exactly this claim about several versions of Windows in recent months) or they seriously need to reconsider how they run their response to security vulnerabilities that are reported.
And how, exactly, do you know when you're running "trusted code"? For years, security experts recited the mantra that you couldn't get an e-mail virus just from reading your mail, and you had to actually run an attachment to get infected. Then MS screwed up with the scripting in things like Outlook (Express), and suddenly all the non-techies in the world, trusting their techie colleagues about the virus thing, are getting caught. Whose fault is this? I sure as hell don't want or expect to run any code automatically just because it's part of an e-mail I'm reading, but MS left me no choice if I use that product, and of course many have no choice about that, either.
No, I think bitching about scripting and APIs that let code run on my box when I neither want nor expect it to is quite justified, thanks very much.
So does Windows XP, and look at all the good press that's gotten Microsoft.
You make a very convincing argument, as many pro-gun-control people do. Unfortunately, most of it is totally unsupported by the facts available, which is invariably where such arguments break down.
Muggers do not generally shoot first. The cycle of hate you describe, where surviving muggers learn their lesson and get more vicious next time, doesn't generally happen. Disarming everyone might be a great idea, except that it's totally impractical (said the poster from the UK, where only criminals have guns and gun crime is up 50+% since they banned the few handguns that were allowed before).
I have reservations about a massively pro-gun attitude, such as many US citizens appear to have, but I don't think mass disarming is even remotely plausible from your current starting point. That being the case, I'm more in favour of letting everyone have basic firearms than just the bad guys.
And what, exactly, is wrong with having more stuff along the lines of Britney? I think she's cute, and one of her isn't enough to go around... ;-)
It will also, given time, turn your reputation to shit. You will not survive forever supporting a small-scale development you once did, and no-one sane will hire you to write a larger one in the future, knowing that you'll tie them in. If board members of the biggest companies in the world are questioning when MS does this, you sure as hell aren't going to get far as a single contractor that way.
OTOH, if you do a good job, and leave it well-documented and maintainable, then (a) you'll have a much more pleasant time if you're the guy doing that maintenance a year later, and (b) you're going to develop a reputation as someone who does a good job, which is about the most valuable thing you can possibly have in the contracting business.
This is exactly the problem: you think it's impressive that most Linux distros have now reached the point of overcoming a problem that Windows never even had.
While I understand your perspective as someone who presumably likes Linux, a quick look down this thread shows how willing Linux enthusiasts are to overlook things that would put most potential converts off in a heartbeat:
Do pro-Linux people see the common thread here? Linux is mostly good at doing most of what Windows can do within a few months of being able to do it on Windows, with mostly compatible results. To a hobbyist geek, like many of us, that might be good enough, and we're prepared to put up with some modest inconvenience in exchange for the benefits you get from using Linux. However, to most Windows, users, it just ain't good enough.
Does it? I haven't used anything else seriously for a while, but at the last count, Visual Studio's debugger was comfortably ahead of the opposition on useful features. Granted that in their quest to .NETify everything they've stuffed up a couple of the basics for vanilla C++ programmers like me, but still, the little things make all the difference. For example, in the VS debugger, you have the AUTOEXP.DAT file to customise how variables of user-defined types are displayed. Does the Kylix debugger have such things these days?
(This is a serious question, not a troll. I've always quite liked Borland's dev apps, but haven't used C++ Builder seriously for a good 3-4 years now, whereas I've had the latest and greatest(?) Visual Studio at work all the time. I'd like to know if Borland really have caught up.)
[Contemplates the various swipes at the US Presidential election currently running through the minds of /.ers around the world]
Nah, that's just too easy... ;-)
It's not a question of "can't be", it's a question of "doesn't benefit from being".
An open source approach has produced some great developments, but almost universally when:
If you meet these criteria, then OS may be the way forward. If not, maybe you need to look at why so many people still do things the old-fashioned way.
<sigh> You'll be wanting open government next...
Now that I have to take issue with. There are, as we have demonstrated in this thread and elsewhere, numerous fields where databases simply aren't a useful programming tool. There are very few fields where OO is not applicable if you choose to use it, though there may be specialist tools that would do the job better. (What are those tools written in, though?)
I don't really want to get into this whole argument all over again, but just consider: I could sit down, right now, with your typical OO development system, and write you a database (either an actual DBMS or a custom biz app based on a database), and I could do this without using any Oracle, MySQL, or whatever. I can therefore do anything a database can, though perhaps not as cleanly until I've built the tools to reinvent a database.
On the other hand, I rather suspect that you couldn't sit down with your typical database package, right now, and write me an OO compiler. In fact, I rather suspect that you couldn't sit down with your typical database package, right now, and write any of the applications I've ever worked on.
This in itself should tell you something about the relative expressive power of the two approaches, and how general-purpose they really are.
Ironically enough, you make my point for me. If you don't like an essay that's printed in Impact, you can change the style to print it in a more readable font without affecting the actual content at all. The presentation is improved, without changing any of the information contained within that presentation.
I meant what you call "custom biz apps", which basically seem to be databases with some UI on top.
I didn't say OO philosophy was the best fit for everything. I said it's a general-purpose approach, as in, it can be used to solve problems in many domains, not one specific one. Clearly a domain-specific tool, such as a database, will normally be more helpful for solving a problem in its own specific domain than any general-purpose approach. The smart programmer identifies any domain-specific tools that will help with his task, and then typically uses those in combination with general-purpose techniques, the latter providing functionality where no more appropriate tool is available, and providing glue between everything.
There are data structures: various types of array, linked list, tree, etc. Then there are interfaces used to abstract those data structures: stack, queue, support of iteration to various degrees, etc. Logically enough, the latter are commonly referred to as "abstract data types". All ADTs must be implemented using some sort of data structure(s) under the hood, but a stack interface could be (and often is) implemented using a variety of different data structures depending on the particular performance characteristics needed for the problem at hand.
That is irrelevant. You cannot look inside a stack using the stack interface. I defy you to cite a single authoritative source anywhere in computing literature whose definition of a stack allows this.
It is commonly referred to as a "call stack" because function calls work like a stack. When you call a function, you add a new function to the top of the stack. When you return from a function, you remove the function currently at the top of the stack. There is nothing else you can do. Even exceptions, as used in many languages, only perform the equivalent of a series of "pop" operations.
Obviously, a debugger can instrument what's going on in more detail if it wishes to do so, but it won't be using a stack interface to do it. What you see in the debugger window isn't being generated from the actual call stack, but from the debugger's own internal data structures.
This is the thing I don't understand. Aside from disagreeing with your definition of "low-level" (is anything that is not a database high level in your terms?) I wonder what you mean when you write "ruining" here. If I wanted to write a database app, I would use database tools. Depending on what else it had to do, I might choose to write it in another language, or in multiple languages, but even then, I would use a database tool to do the database work.
Is your major argument that people are now starting to try to write what are essentially database apps using only general purpose OO languages, and not database-specific tools? If so, then I am sorry that I have missed your point for so long, and I agree entirely that that particular idea is idiotic. OO is a general-purpose approach, and it makes no sense to ignore database-specific tools if databases are what you're doing. This doesn't mean you can't write parts of your program in an OO language, of course, just that it would be sensible to control a database from your program to do the database work, and not to try creating a whole specific database in OO terms to mimic the same effect.
Um... Who was that?
I've seen lots of people indicate that they prefer content to style. (The /. population might well not be representative of the overall population in this respect, of course.) I don't think I've seen anyone suggest that a quality site can't be good looking as well, though.
On the contrary. Like you, I regard sites that use small font sizes and fix them there using stylesheets as being irritating rather than useful. Like many computer users, I don't have perfect vision, and appreciate a clear layout that I can read without eye strain. I simply won't attempt to read through an article of any length in small type. If a site provides an alternative stylesheet, or defines its font sizes relatively so that I can adjust them to be comfortable, it scores bonus points. Otherwise, it had better hope it works with Moz, where I can override the "fixed" font size anyway.
I agree that any site that requires Flash or cookies gets a big thumbs down. I don't think it's the credibility that's harmed, though. It's the usability (principally because of the download times for us modem users in the Flash case).
If the same site let me see a simple front page with the same content but no Flash animation, so I can download it and use it some time today, then maybe I'd rate it very highly if it still gave me the information I wanted easily.
If a site provides reasonable defaults without cookies, which it should be able to do if it just uses them to store my preferences as they were intended, then again, WTP? If I want to store the preferences, I'll enable cookies for the site. If not, I won't.
Even ads don't damage a site's credibility in my eyes, if they're done responsibly. I don't mind a banner ad or two that support a page. I find pop-up ads irritating, but these mostly seem to be put there by web hosting companies rather than the actual authors of a page I'm reading, so I tend to discount them as well. The only ads thing that really hurts a page in my eyes is being nasty about it. If I visit a travel agent's site, and when I've finished I close the window to find seventeen different ad windows for holidays I haven't even asked about, I'm never going back.
I guess you could argue that even this last case is really usability, but there comes a point where a site is sufficiently hard to use, or disregards the feelings of its visitors to such an extent, that it becomes a credibility problem as well.
Nah. Somehow I just didn't like the look of it.