Like most of the other intelligent comments here, slashdotters know this product, and it's approach, is neither new or well targeted.
Not only will such a product be almost impossible to use because of the task of trying to render intelligence from all the HTML soup out there, it does not address the software architecture underlying OS and browsers. It's a waste of time.
All the major operating systems; Microsoft Windows , Apple Macintosh, and the major distributions of Unix, are built on architectures that support accessibility. These are inherent within the component architecture of the software development kits (SDK), and correctly built software utilises the component architectures for interfacing with various devices, including human interactive devices of screen, keyboard, mouse and sound.
When one builds software in accordance with the resources of the SDK, the fundamentals of user accessibility to software interfaces are addressed. If the developer is not aware of the component architecture of the SDK, and why it is structured that way, they can unwittingly penalised their software usability and their users accessibility, through failing to address standard user interface issues. Each resource needs to be correctly implemented; Icons, Cursor, Menus, Dialog Boxes, Bitmaps, Fonts, String-Tables, Accelerators, Name-Table
Version-Information.
The web also has a similar architecture, and the foundations are there, and improving with better user agent support. Such media can be targeted independently through correct use of CSS Media types in the latest browsers. Negligence in not being aware of the web's architecture also penalises the web developer when designing and building web services (such as this case).
If the web developer wraps their valid code in logical sections divided with DIV tags and assigned appropriate CSS classes, it is possible, using one design, to either display or hide various sections of the page. This also addresses the need for printer friendly pages.
When the support finally arrives (here's hoping for a perfect world... it is definitely improving), one will be able to enhance web pages, such as interviews and navigation with different voices so that the listener can better differentiate the content through intelligent and creative assignment of Aural CSS.
Given that Aural CSS is designed to be interactive, it would be of immense value to the whole mobile device development community to correctly implement such services.
This would not only assist those with disabilities, but it would be great to be able to listen to some web pages while cleaning the office desk, taking a break, or whatever... but maybe not driving a car.
What I am saying is that often many coders, until they have a lot of experience, just love their coding so much they are not yet experienced enough to engineer their code according to the rigours of strict business discipline.
The value in experience I bought to this project was that I was a coder with more than 10 years experience as a developer, and the rigour and scrutiny I applied to testing was much greater than the testing department was used to applying for web applications. There was almost 4000 test cases applied to this product.
This project was highly political with the CIO monitoring it. It was bound to fail given the parameters and time constraints, yet no department wanted egg on it's face, and to be only shown they had done their job thoroughly. This is the type of politics that pervade all big business (I have worked in) and why I was utilised for it. This department wanted to make sure that when shit happened, they were clean, they had done their job thoroughly. Not a case of blaming others, just survival and to be seen as utterly professional.
The problems in this project were obvious to anyone with real knowledge or experience, yet not so to these young unexperienced, but enthusiastic coders. Most of the others on this project did not have the experience to see where the real problems were arising from; no real unified architecture, and it was failing miserably in UAT.
Personally, I feel this type of scenario is almost endemic in large scale web development today, especially in big business where projects scale across departments and if you raise issues you are more likely to be standing on toes than helping the project and others out. More often than not it is best to STFU and just code exactly to the business specs, even if they are wrong. Just do your job as a coder as well as possible, utilise all your knowledge and experience, but don't add or neglect anything beyond the business and functional specs, and if they are not thorough, make sure you back yourself with the best of practice use of standards, engineering and architecture.
Actually, it wasn't these guys fault at all, it was the clowns in management who did not have enough knowledge and experience to properly manage web projects. If they were doing their job properly they would have implemented better guidelines, but the coders were not experienced enough to know what to do in these situations, when the specifications are poor, to completely exempt themselves from the inevitable blame that would rebound. Good programmers, with a lot of experience know what allies such discipline and rigour provide.
An "incredible aptitude for detail and complexity" != experienced knowledge in the importance of working to standards, library management, UI consistency, specifications, documentation.
I managed a testing project for the main Intranet project for a top Fortune company. The project was managed in the US, coded in India and tested in Australia. There were also a lot of Indians working in the Australian centre. I have worked with Indians in the past too, as well as other Asians. They have an incredible aptitude for detail and complexity.
But I have found they lacked the knowledge and background to come to terms with the importance of managing and unifying the overall software architecture. There was little knowledge of programming to standards, managing common libraries, UI consistency, working to business and functional specifications, documentation, etc, etc. And also the business and legal implications of straying from these guidelines. Everyone suffers from these things, but to my mind it doesn't matter what advantages one has, if one does not have the knowledge and skill to implement good SDLC procedures, it handicaps the whole project.
The other thing I noticed is that these developers are willing to work long hours, and they do not seem to get as stressed as those from western cultures, they seem to just thrive on the work.
The Web Content Accessibility Guidelines 2.0(WCAG) are in draft form and are trying to improve on WCAG 1.0. If you really what to contribute to them, or can contribute in anyway through writing parsers, etc, go ahead and join the WCAG GL
Hopefully this initiative will drive the production of better browsers (user agents) and web development that will facilitate services to users.
User agents are improving on the implementation of the HTTP_ACCEPT header for determining MIME types the user agent will accept. This great potential was missed with Netscape Navigator 2 because, in the rush to get it to market, they just defaulted to using *.* (this browser accepts everything), when it didn't. If this was implemented correctly it would allow the developer to deliver media according to the user agents capacity.
I know enough Unix to get me around (and into trouble). I haven't used much the last couple of years, but now I have to get back into it. Fine.
I installed Coherent on my PC back in uni days. In 95 I had difficulty installing Linux until a sys admin told me to skip the "check for bad blocks" step. I ran Redhat 4.x and 5.x without much problem, as well as using Solaris.
Just recently I went to install Redhat 7.3 onto a 20 gig partition, I wanted a developers install with basically everything. I had the official CDs, I could get a good base install, no problem, but with everything, it asked for CD3, but could not recognise it. There was no graceful regression. The only option was to hit the reset button and loose your bootup.
Much the same with Mandrake 9.0 if I wanted everything, after installing from CD1 and 2, it asked for the next CD? There was only 2 with the official distribution, and neither of these were what it wanted.
With some help from a friend, I installed Debian Woody, because I always got stuck in the "dselect" menu, which was completely foreign to the rest of the install gui, both in UI and logic. After getting an install, and finally starting x, I took it home. Then I find that I can only login as an account user and have to use "su" to login as root, which I must also do to shutdown. Not only that, but when I use apt-get to install php4 apacheconfig comments out the php module. I could not understand why certain important CPAN modules were missing and others installed. Why did that not come with a base Woody install, there's 7 CDs? The Gnome help could not find the TOCs.
The other thing is, up until the last few years, I could get old Unix sys admin manuals and file and directory locations would pretty much correspond to those documents (Essential Systems Administration, Unix Systems Administration Handbook, etc). Now with each distro, these locations could be anywhere. Debian Woody did not even install with cgi-bin directory in the web server directory. Even the group names are non standard.
This is what is loosing a lot of people who'd like to move to Linux. Basically, they just want a desktop that runs and installs okay, Linux is getting there, but any further, running daemons and wanting to do more, it gets difficult, because even standard Unix documents people refer to, do not apply anymore. Why did so many of the basics have to change and bring into existence all these inconsistencies?
I'm associated with a small rural Linux user group, and if it wasn't for their sense of humour, we'd have all gone mad. Why should any of us have to learn a whole different architecture for each distro? There are a lot of other things I have encountered recently. I'm willing to preserver because I need a Linux system at present, but most other users rightly are not.
I know there are answers to all these things, but it does not address the basic needs of someone whose approach to using Linux is primarily productivity. They do not want to have to piss around with all these shortfalls, they just want to get up and running, use workable install processes, without the constant need for editing configuration files with a text editor. And to have distros that do comply to standards so that knowing one Unix based system means knowing them all. The lack of that frustrates those users who want to use Linux, but don't want ot spend the hours of their days problem solving Linux OS problems and inconsistencies. I am also sensing from some areas that those who want the Unix features in a friendly, standards based and documented enviroment are seriously considering MacOSX.
I'm not an XML expert, but my understanding is that MS or anyone can write proprietary code into the CDATA section of any XML document and therefore only their tools can accurately parse the document.
Of the XML code I have seen generated by MS applications, it's a mess, and lacks any adherence to well structured content, it's spaghetti xml. Same with the style sheets generated and the associated classes.
If this was put in the hands of the programmers who gave us.doc to.htm or those responsible for the code generation in FrontPage, what would you expect the results to be like?
There is also the huge debate (if my memory serves me right) that happened regarding the first W3C XSL recommendation where MS fought for a less strict implementation (because they saw their documents could not comply with a strict implementation).
to show up to training for their squad. I never knew they were so desperate for the services of an overweight late 40's something. It went on for months. They even rang, but I ignored them, they just couldn't afford me.
How confident can web developers and publishers be of being accessibility compliant if they are not knowledgeable in the accessibility field, but do understand and develop to the basic W3C standards (HTML, XML, CSS, XSLT, SVG, etc). Is it possible to feel confident that with good design, and valid code, that that in effect addresses most of the accessibility issues?
This leads me to my gripe that the "standards" body developing these "standards", few of these people come from the trenches, and few understand how people really need and want to use HTML, so of course we have this problem of forcing a round peg through a square hole. The standards committees are comprised mostly of academics and browser engineers, let's see some major web developers up in that hizzy!
If you read any of the more prominent W3C Working Groups in these areas you'll find this not to be true at all. These people do work in the real world, they are aware of the many issues, and they try their hardest to make the standards as all inclusive as possible. See http://lists.w3.org/Archives/Public/www-html/ and http://lists.w3.org/Archives/Public/www-style/ and http://lists.w3.org/Archives/Public/w3c-wai-gl/ or join and participate in any of the standards, it's open.
We just need developers to write a small browser rapper around either Gecko or the IE engine.
Crazy Browser uses the IE installed engine, but it's shell seems lighter, and it is smaller on this install, yet it does have features to block popups etc.
TopStyle CSS Editor uses the IE engine too.
There are many volunteer organisations out there that are installing Linux on 486s, but Mozilla is just too big. The one I know uses Opera with the banner ads, cause there isn't anything else. It would be great to see lightweight browsers that had good current fast rendering engines, support for standards, and easy to use.
W3C validator is extremely anal about "obsolete" tags. Maybe I want a FONT tag in one piece of text that I'm never going to change. Why do I need a CSS name for EVERYTHING?
Because all the current browsers now support rendering the DOCTYPE to one of the current DTDs, and the FONT tags is deprecated, so, sooner or later these tags will not be recognised.
The other point is if you are working in any situation where someone else has to come along and maintain your work, if you have not designed to a standard, how on earth do they know what type of HTML soup you have implemented?
This approach will add to the cost of site development/maintainance.
All the current browsers render according to HTML 4 and XHTML DTDs. It doesn't matter if you are using ASP, PHP, JSP, CFM, whatever.
These days, without validating your HTML to a DTD you Markup will eventually be left in the markup soup wilderness. If you don't define your HTML according to a DTD and validate it, then the browser will say, this is markup soup, all I can do is take a best guess (what else can it do).
This is what browsers have to do, and you have to take you hats of too them, they do a pretty darn good job of trying their best to render so many varieties of soup logic.
Even when one is trying to do the right thing, like here, there are many traps, the following are a few points to evaluate.
If you are trying to do the right thing by seperating content into structured markup (HTML) and display markup (CSS), always transform HTML into XHTML because CSS requires all tags to be closed to render correctly (or as bug free as possible). The HTML 4 spec does not require tags to be closed.
If you use attributes in TABLES etc in TRANSITIONAL DTDs and mix sizes and positions using both absolute and relative positioning you will get into trouble.
It's good to try as much as possible to try to develop according to STRICT DTDs and move all positioning into CSS. I know some browsers do not cope with it, but the V6 browsers out there, IE6, Mozilla 1.x & Opera 6, etc. Even if you can't do what you want under STRICT and have to go back to TRANSITIONAL, the learning process helps.
Validate your CSS, and also run it by Bobby (http://bobby.watchfire.com/bobby/html/en/index.js p) and WAVE (http://www.temple.edu/inst_disabilities/piat/wave/), which will give you some good feedback.
This is the man who, when head of security for MS, gave us the above quote in August 2001 when viruses such as Melissa virus were targeting MS products. If your chief security officer makes such a statement, doesn't it set you wondering about their credibility working in the field of security at all, and the attitude of the company or government that employes them?
You need a bigger vision than just documenting code. If you don't document the whole architecture of the code libraries and classes and describe their relationships and functions, you just end up with a mass of spaghetti comments. Too many comments becomes yet another cognitive overhead. It is better to write code in a clear style and just comment occasionally why you did something, which may not be immediately obvious.
If your team is not documenting the architecture and code libraries in a programmers manual they are not being professionally managed (Danger Will Robinson). Also, coding teams should either work to a given set of coding standards and guidelines (it's up to your team what they are) to ensure and enhance a style within the architecture that is easy for any maintainer to work with, and for future teams to enhance and develop, which itself should also be a formal document.
Many programmers do not like working with such disciplines, but I am sure the more experienced ones out there actually agree, that if projects start with such disciplines, the chances of project success and deliverables if much more enhanced, and the ease of development between version is also better managed. Without such disciplines, each version becomes more and more difficult to manage. You can have fun working with such disciplines, the experience of writing better, more clearly defined quality code is very rewarding in itself.
And as stated here already, say why you did something, not what the code does.
I just love it how Yoda seems crippled with arthritis and cane ridden, but put a light sabre in his hand and he makes Jackie Chan look ordinary and lacking in any sense of energy.
That seems to be the problem. Many users are not aware of the data exchange capacities of their apps and plugins. If there is data exchange the user should be made aware of it at the time of installation or first use.
There should be a configuration walk through on install. If companies would do this, they could at least gain respect for the discloser, and educate the user to the feature set of the product. And there is always the default / advanced installation for those who just blindly want to accept it or custom configure it.
Slashdot Mirror
Bertrand Meyer for his contribution to OO Software Engineering, Eiffel language and Design By Contract.
I think an honourable mention most go to Ted Nelson and Xanadu.
Like most of the other intelligent comments here, slashdotters know this product, and it's approach, is neither new or well targeted.
Not only will such a product be almost impossible to use because of the task of trying to render intelligence from all the HTML soup out there, it does not address the software architecture underlying OS and browsers. It's a waste of time.
All the major operating systems; Microsoft Windows , Apple Macintosh, and the major distributions of Unix, are built on architectures that support accessibility. These are inherent within the component architecture of the software development kits (SDK), and correctly built software utilises the component architectures for interfacing with various devices, including human interactive devices of screen, keyboard, mouse and sound.
When one builds software in accordance with the resources of the SDK, the fundamentals of user accessibility to software interfaces are addressed. If the developer is not aware of the component architecture of the SDK, and why it is structured that way, they can unwittingly penalised their software usability and their users accessibility, through failing to address standard user interface issues. Each resource needs to be correctly implemented; Icons, Cursor, Menus, Dialog Boxes, Bitmaps, Fonts, String-Tables, Accelerators, Name-Table Version-Information.
The web also has a similar architecture, and the foundations are there, and improving with better user agent support. Such media can be targeted independently through correct use of CSS Media types in the latest browsers. Negligence in not being aware of the web's architecture also penalises the web developer when designing and building web services (such as this case).
If the web developer wraps their valid code in logical sections divided with DIV tags and assigned appropriate CSS classes, it is possible, using one design, to either display or hide various sections of the page. This also addresses the need for printer friendly pages.
When the support finally arrives (here's hoping for a perfect world... it is definitely improving), one will be able to enhance web pages, such as interviews and navigation with different voices so that the listener can better differentiate the content through intelligent and creative assignment of Aural CSS.
Given that Aural CSS is designed to be interactive, it would be of immense value to the whole mobile device development community to correctly implement such services.
This would not only assist those with disabilities, but it would be great to be able to listen to some web pages while cleaning the office desk, taking a break, or whatever... but maybe not driving a car.
What I am saying is that often many coders, until they have a lot of experience, just love their coding so much they are not yet experienced enough to engineer their code according to the rigours of strict business discipline.
The value in experience I bought to this project was that I was a coder with more than 10 years experience as a developer, and the rigour and scrutiny I applied to testing was much greater than the testing department was used to applying for web applications. There was almost 4000 test cases applied to this product.
This project was highly political with the CIO monitoring it. It was bound to fail given the parameters and time constraints, yet no department wanted egg on it's face, and to be only shown they had done their job thoroughly. This is the type of politics that pervade all big business (I have worked in) and why I was utilised for it. This department wanted to make sure that when shit happened, they were clean, they had done their job thoroughly. Not a case of blaming others, just survival and to be seen as utterly professional.
The problems in this project were obvious to anyone with real knowledge or experience, yet not so to these young unexperienced, but enthusiastic coders. Most of the others on this project did not have the experience to see where the real problems were arising from; no real unified architecture, and it was failing miserably in UAT.
Personally, I feel this type of scenario is almost endemic in large scale web development today, especially in big business where projects scale across departments and if you raise issues you are more likely to be standing on toes than helping the project and others out. More often than not it is best to STFU and just code exactly to the business specs, even if they are wrong. Just do your job as a coder as well as possible, utilise all your knowledge and experience, but don't add or neglect anything beyond the business and functional specs, and if they are not thorough, make sure you back yourself with the best of practice use of standards, engineering and architecture.
Actually, it wasn't these guys fault at all, it was the clowns in management who did not have enough knowledge and experience to properly manage web projects. If they were doing their job properly they would have implemented better guidelines, but the coders were not experienced enough to know what to do in these situations, when the specifications are poor, to completely exempt themselves from the inevitable blame that would rebound. Good programmers, with a lot of experience know what allies such discipline and rigour provide.
An "incredible aptitude for detail and complexity" != experienced knowledge in the importance of working to standards, library management, UI consistency, specifications, documentation.
I managed a testing project for the main Intranet project for a top Fortune company. The project was managed in the US, coded in India and tested in Australia. There were also a lot of Indians working in the Australian centre. I have worked with Indians in the past too, as well as other Asians. They have an incredible aptitude for detail and complexity.
But I have found they lacked the knowledge and background to come to terms with the importance of managing and unifying the overall software architecture. There was little knowledge of programming to standards, managing common libraries, UI consistency, working to business and functional specifications, documentation, etc, etc. And also the business and legal implications of straying from these guidelines. Everyone suffers from these things, but to my mind it doesn't matter what advantages one has, if one does not have the knowledge and skill to implement good SDLC procedures, it handicaps the whole project.
The other thing I noticed is that these developers are willing to work long hours, and they do not seem to get as stressed as those from western cultures, they seem to just thrive on the work.
Also of interest in the same area are;
Hopefully this initiative will drive the production of better browsers (user agents) and web development that will facilitate services to users.
User agents are improving on the implementation of the HTTP_ACCEPT header for determining MIME types the user agent will accept. This great potential was missed with Netscape Navigator 2 because, in the rush to get it to market, they just defaulted to using *.* (this browser accepts everything), when it didn't. If this was implemented correctly it would allow the developer to deliver media according to the user agents capacity.
Also see User Agent Accessibility Guidelines 1.0 and the UAAG discussion list.
nuts.
I know enough Unix to get me around (and into trouble). I haven't used much the last couple of years, but now I have to get back into it. Fine.
I installed Coherent on my PC back in uni days. In 95 I had difficulty installing Linux until a sys admin told me to skip the "check for bad blocks" step. I ran Redhat 4.x and 5.x without much problem, as well as using Solaris.
Just recently I went to install Redhat 7.3 onto a 20 gig partition, I wanted a developers install with basically everything. I had the official CDs, I could get a good base install, no problem, but with everything, it asked for CD3, but could not recognise it. There was no graceful regression. The only option was to hit the reset button and loose your bootup.
Much the same with Mandrake 9.0 if I wanted everything, after installing from CD1 and 2, it asked for the next CD? There was only 2 with the official distribution, and neither of these were what it wanted.
With some help from a friend, I installed Debian Woody, because I always got stuck in the "dselect" menu, which was completely foreign to the rest of the install gui, both in UI and logic. After getting an install, and finally starting x, I took it home. Then I find that I can only login as an account user and have to use "su" to login as root, which I must also do to shutdown. Not only that, but when I use apt-get to install php4 apacheconfig comments out the php module. I could not understand why certain important CPAN modules were missing and others installed. Why did that not come with a base Woody install, there's 7 CDs? The Gnome help could not find the TOCs.
The other thing is, up until the last few years, I could get old Unix sys admin manuals and file and directory locations would pretty much correspond to those documents (Essential Systems Administration, Unix Systems Administration Handbook, etc). Now with each distro, these locations could be anywhere. Debian Woody did not even install with cgi-bin directory in the web server directory. Even the group names are non standard.
This is what is loosing a lot of people who'd like to move to Linux. Basically, they just want a desktop that runs and installs okay, Linux is getting there, but any further, running daemons and wanting to do more, it gets difficult, because even standard Unix documents people refer to, do not apply anymore. Why did so many of the basics have to change and bring into existence all these inconsistencies?
I'm associated with a small rural Linux user group, and if it wasn't for their sense of humour, we'd have all gone mad. Why should any of us have to learn a whole different architecture for each distro? There are a lot of other things I have encountered recently. I'm willing to preserver because I need a Linux system at present, but most other users rightly are not.
I know there are answers to all these things, but it does not address the basic needs of someone whose approach to using Linux is primarily productivity. They do not want to have to piss around with all these shortfalls, they just want to get up and running, use workable install processes, without the constant need for editing configuration files with a text editor. And to have distros that do comply to standards so that knowing one Unix based system means knowing them all. The lack of that frustrates those users who want to use Linux, but don't want ot spend the hours of their days problem solving Linux OS problems and inconsistencies. I am also sensing from some areas that those who want the Unix features in a friendly, standards based and documented enviroment are seriously considering MacOSX.
I'm not an XML expert, but my understanding is that MS or anyone can write proprietary code into the CDATA section of any XML document and therefore only their tools can accurately parse the document.
Of the XML code I have seen generated by MS applications, it's a mess, and lacks any adherence to well structured content, it's spaghetti xml. Same with the style sheets generated and the associated classes.
If this was put in the hands of the programmers who gave us .doc to .htm or those responsible for the code generation in FrontPage, what would you expect the results to be like?
There is also the huge debate (if my memory serves me right) that happened regarding the first W3C XSL recommendation where MS fought for a less strict implementation (because they saw their documents could not comply with a strict implementation).
Also, didn't Microsoft Corporation Selects SoftQuad XMetaL To Create XML Content ?
to show up to training for their squad. I never knew they were so desperate for the services of an overweight late 40's something. It went on for months. They even rang, but I ignored them, they just couldn't afford me.
Win2000 loads this app on setup. I really want to know what it does and how it works.
How confident can web developers and publishers be of being accessibility compliant if they are not knowledgeable in the accessibility field, but do understand and develop to the basic W3C standards (HTML, XML, CSS, XSLT, SVG, etc). Is it possible to feel confident that with good design, and valid code, that that in effect addresses most of the accessibility issues?
It is possible to design good sites and get passed by Bobby, E-bility.com
If you read any of the more prominent W3C Working Groups in these areas you'll find this not to be true at all. These people do work in the real world, they are aware of the many issues, and they try their hardest to make the standards as all inclusive as possible. See http://lists.w3.org/Archives/Public/www-html/ and http://lists.w3.org/Archives/Public/www-style/ and http://lists.w3.org/Archives/Public/w3c-wai-gl/ or join and participate in any of the standards, it's open.
Nothing to add to this... it's spot on advice...
Crazy Browser uses the IE installed engine, but it's shell seems lighter, and it is smaller on this install, yet it does have features to block popups etc.
TopStyle CSS Editor uses the IE engine too.
There are many volunteer organisations out there that are installing Linux on 486s, but Mozilla is just too big. The one I know uses Opera with the banner ads, cause there isn't anything else. It would be great to see lightweight browsers that had good current fast rendering engines, support for standards, and easy to use.
W3C validator is extremely anal about "obsolete" tags. Maybe I want a FONT tag in one piece of text that I'm never going to change. Why do I need a CSS name for EVERYTHING?
Because all the current browsers now support rendering the DOCTYPE to one of the current DTDs, and the FONT tags is deprecated, so, sooner or later these tags will not be recognised.
The other point is if you are working in any situation where someone else has to come along and maintain your work, if you have not designed to a standard, how on earth do they know what type of HTML soup you have implemented?
This approach will add to the cost of site development/maintainance.
All the current browsers render according to HTML 4 and XHTML DTDs. It doesn't matter if you are using ASP, PHP, JSP, CFM, whatever.
These days, without validating your HTML to a DTD you Markup will eventually be left in the markup soup wilderness. If you don't define your HTML according to a DTD and validate it, then the browser will say, this is markup soup, all I can do is take a best guess (what else can it do).
This is what browsers have to do, and you have to take you hats of too them, they do a pretty darn good job of trying their best to render so many varieties of soup logic.
Even when one is trying to do the right thing, like here, there are many traps, the following are a few points to evaluate. If you are trying to do the right thing by seperating content into structured markup (HTML) and display markup (CSS), always transform HTML into XHTML because CSS requires all tags to be closed to render correctly (or as bug free as possible). The HTML 4 spec does not require tags to be closed. If you use attributes in TABLES etc in TRANSITIONAL DTDs and mix sizes and positions using both absolute and relative positioning you will get into trouble. It's good to try as much as possible to try to develop according to STRICT DTDs and move all positioning into CSS. I know some browsers do not cope with it, but the V6 browsers out there, IE6, Mozilla 1.x & Opera 6, etc. Even if you can't do what you want under STRICT and have to go back to TRANSITIONAL, the learning process helps. Validate your CSS, and also run it by Bobby (http://bobby.watchfire.com/bobby/html/en/index.js p) and WAVE (http://www.temple.edu/inst_disabilities/piat/wave /), which will give you some good feedback.
test(this);
... and it will also transform documents into XHTML... http://tidy.sourceforge.net/
This is the man who, when head of security for MS, gave us the above quote in August 2001 when viruses such as Melissa virus were targeting MS products. If your chief security officer makes such a statement, doesn't it set you wondering about their credibility working in the field of security at all, and the attitude of the company or government that employes them?
If your team is not documenting the architecture and code libraries in a programmers manual they are not being professionally managed (Danger Will Robinson). Also, coding teams should either work to a given set of coding standards and guidelines (it's up to your team what they are) to ensure and enhance a style within the architecture that is easy for any maintainer to work with, and for future teams to enhance and develop, which itself should also be a formal document.
Many programmers do not like working with such disciplines, but I am sure the more experienced ones out there actually agree, that if projects start with such disciplines, the chances of project success and deliverables if much more enhanced, and the ease of development between version is also better managed. Without such disciplines, each version becomes more and more difficult to manage. You can have fun working with such disciplines, the experience of writing better, more clearly defined quality code is very rewarding in itself.
And as stated here already, say why you did something, not what the code does.
I just love it how Yoda seems crippled with arthritis and cane ridden, but put a light sabre in his hand and he makes Jackie Chan look ordinary and lacking in any sense of energy.
There should be a configuration walk through on install. If companies would do this, they could at least gain respect for the discloser, and educate the user to the feature set of the product. And there is always the default / advanced installation for those who just blindly want to accept it or custom configure it.