I see only a representation of data (with a lot of computations behind). I cay say that is there, because is there some app that can give me a representation of it.
In the first days at this forum, by the cookies that happend my a lot of times, yes, I never wrote that post, in that case I write other post, but you and all here will never see my first answer :) (if you whant, you can search or I will do maybe, some post where I say what happend? or that my anterior post was deleted!! or I can not post)
I can say that post not exist, in fact it exist for me, but not for the people here or the database, I take that like the action that serve for not express but let the work done (do the work or save the work that is represent the work in the actual medium... reality== actions, computers==a store), if you can not see my first answer, then I write a second answer from the "essense" of the anterior one. :D :) :S mmmm let say that it was a dream on become a generation in potence, but never happend because a x event not predicted by me, like some times happend, I do plans for yesterday, when I was in the crucial point I where hited by other circunstances that dont let me follow my dream or planifications. Then they where only dreams, but yes, they exist, when you do a dream reality!!!, yea yea ;)
Have a nice day or night.
In the first days at this forum, by the cookies that happend my a lot of times, yes, I never wrote that post, in that case I write other post, but you and all here will never see my first answer :) (if you whant, you can search or I will do maybe, some post where I say what happend? or that my anterior post was deleted!! or I can not post)
I can say that post not exist, in fact it exist for me, but not for the people here or the database, I take that like the action that serve for not express but let the work done (do the work or save the work that is represent the work in the actual medium... reality== actions, computers==a store), if you can not see my first answer, then I write a second answer from the "essense" of the anterior one. :D :) :S mmmm let say that it was a dream on become a generation in potence, but never happend because a x event not predicted by me, like some times happend, I do plans for yesterday, when I was in the crucial point I where hited by other circunstances that dont let me follow my dream or planifications. Then they where only dreams, but yes, they exist, when you do a dream reality!!!, yea yea ;)
Have a nice day or night.
All of the arguments fall to peices as soon as the browser has content filtering,which most do. This is in effect manipluating and adjusting the data for display, it is creating a new dataset based on user supplied parameters, that is no different than some program that does a calculation and displays the result. All software falls into both categories, a CAD program takes information from the user and other sources and produces a result based on the data provided. In all cases the software only accepts a set of limitations and data from the user or external sources and formats that data, manipulating it as necessary, you have decided that some manipulation is less valuable than others, that is a personal opinion and not a technical one. No CAD program creates information, it manipulates existing information and performs calculations based on that data, nothing new is created only information being filtered though however complex an algorithm.
The browser is the primary interface to the internet in the same way the GUI of autocad is the primary interface to it's functionality, if you wish to judge a broswer you must judge the internet as a whole. When I filter content through a search engine it is in effect doing the same thing as filtering data through a FFT, it is formatting existing data for human comprehension, after all I cannot read the bitstream of a JPG directly. Whether that functionality is included as part of the main executable, a DLL or a third party plugin is a moot point, the fact is that the browser is a necessary component in it's use. Also if you wish to arbitrarily exclude certain parts of a browsers functionality such as PHP etc.. you must also exclude all API and other parts as well, including from the CAD software, the external programs that it was built to use must all be included or all excluded in order to judge properly. You seem bent on limiting the scope of browsers while expanding all others for the comparison, hardly fair.
The browser is the primary interface to the internet in the same way the GUI of autocad is the primary interface to it's functionality, if you wish to judge a broswer you must judge the internet as a whole. When I filter content through a search engine it is in effect doing the same thing as filtering data through a FFT, it is formatting existing data for human comprehension, after all I cannot read the bitstream of a JPG directly. Whether that functionality is included as part of the main executable, a DLL or a third party plugin is a moot point, the fact is that the browser is a necessary component in it's use. Also if you wish to arbitrarily exclude certain parts of a browsers functionality such as PHP etc.. you must also exclude all API and other parts as well, including from the CAD software, the external programs that it was built to use must all be included or all excluded in order to judge properly. You seem bent on limiting the scope of browsers while expanding all others for the comparison, hardly fair.
All of the arguments fall to peices as soon as the browser has content filtering,which most do. This is in effect manipluating and adjusting the data for display, it is creating a new dataset based on user supplied parameters, that is no different than some program that does a calculation and displays the result.
Nonsense. There is no new data created. You are simply discarding some data, leaving a subset.
In the case of CAD programs, they can actually generate new data for you. You can get more output than input. Browsers can only give less or equal output, since they can only filter out data.
No CAD program creates information, it manipulates existing information and performs calculations based on that data, nothing new is created only information being filtered though however complex an algorithm.
What do you call it then when a CAD program distills a grid for Finite Element Analysis, based on a model that the user entered? To give just one example of where new information is generated, which could in no way be filtered out from the original data (it is not a subset of the original data).
Or what if you then use that grid to do simulations of fluid flow? Where does the flow come from? The user doesn't input that, does he? And he doesn't input the resulting data of the analysis either, does he? (How could he, he didn't know the data, which is why he had to do the analysis in the first place).
Not nonsense at all, what does a database application do other than that, or a CAD program other than present data based on user supplied parameters filtered though a set of formulas, the data was already there obviously, the program didn't invent anything. Again you have decided to limit your definitions due to a personal bias. No software does anything other than format and present data or manipulate it in order to display what the user has requested based on his input and parameters, whether you accept this or not it is reality. You arbitrarily say that some manipulation is valid and anything that does not agree with your assessment is not, that makes any argument pointless, I remember several times when you accused others of the same limited vision.
present data based on user supplied parameters filtered though a set of formulas, the data was already there obviously, the program didn't invent anything.
Not everything can be expressed in formulas (that can be solved directly by a computer), and not everything is a one-to-one or many-to-one relation. There are also one-to-many or many-to-many relations.
Sometimes you need simulations or iterative systems to distill a new usable dataset. Which has nothing to do with the user input, and everything to do with the CAD application at hand. Different CAD applications will have different methods of solving complex problems, which means they are not interchangable. This is the thing that you do not seem to understand. CAD programs are infinitely more complex than browsers.
No software does anything other than format and present data or manipulate it in order to display what the user has requested based on his input and parameters, whether you accept this or not it is reality.
This is wrong, try to solve the cone-cone problem I mentioned earlier, and see what I mean.
There is not a single solution, and there is no 'right' solution. You will have to approximate it in one way or the other. Different software will approximate it in different ways (same goes for the aforementioned grid by the way, and there are many other examples).
You arbitrarily say that some manipulation is valid and anything that does not agree with your assessment is not
Not at all, I'm saying that if applications do not generate new data, but only visualize existing data, generated elsewhere, then these applications are interchangeable, while the application that generated the data probably is not.
Is this an argument over standardization of formulas in CAD packages ??
The fact is that no new data is created, the formulas "HOWEVER COMPLEX" just manipulate data that is already supplied by the user or available in tables or calculation matrices. The CAD program does not create new data only presents existing data in a new light, in the same way as you cannot say that CALC.EXE creates data when you enter 1+1, the complexity of the calculation filtering the information is not an issue here. Only the fact that neither pull data out of their asses.
The fact is that no new data is created, the formulas "HOWEVER COMPLEX" just manipulate data that is already supplied by the user or available in tables or calculation matrices. The CAD program does not create new data only presents existing data in a new light, in the same way as you cannot say that CALC.EXE creates data when you enter 1+1, the complexity of the calculation filtering the information is not an issue here. Only the fact that neither pull data out of their asses.
As a matter of fact, CAD programs DO 'pull data out of their asses' in some cases, because that is the only way to approximate a solution to a problem. You do not seem to understand this.
I disagree that eg generating a grid is manipulating data, because the grid has no direct relation to the data. The data is just used as a seed to generate the grid. Or more clearly, to generate A grid. There are infinitely many grids that could be generated, since the optimal grid cannot be found in finite time.
Therefore, heuristics and other shortcuts have to be applied, and this is pretty much 'pulling data out of their asses'.
Also, the grid is an addition to the input, not a replacement of it. So your dataset has grown, new data has been added. The user didn't enter anything for that grid, he just loaded his model and pressed the 'Generate grid' button. It is in no way a filtered form of the original model.
Let's say there is no exact solution, and many ways to approximate 1 + 1. Now if let's say if the calculator would answer 1 + 1 = 2.1232 on some OSes, and 1 + 1 = 2.0723 on others, perhaps you see what I mean?
I disagree that eg generating a grid is manipulating data, because the grid has no direct relation to the data. The data is just used as a seed to generate the grid. Or more clearly, to generate A grid. There are infinitely many grids that could be generated, since the optimal grid cannot be found in finite time.
Therefore, heuristics and other shortcuts have to be applied, and this is pretty much 'pulling data out of their asses'.
Also, the grid is an addition to the input, not a replacement of it. So your dataset has grown, new data has been added. The user didn't enter anything for that grid, he just loaded his model and pressed the 'Generate grid' button. It is in no way a filtered form of the original model.
Let's say there is no exact solution, and many ways to approximate 1 + 1. Now if let's say if the calculator would answer 1 + 1 = 2.1232 on some OSes, and 1 + 1 = 2.0723 on others, perhaps you see what I mean?
Kind of sounds like a good filtering search engine to me. You should try vivisimo, it does a good job of searching with all possible engines then orgainzing and limiting the data presented as surely a search for IF in the internet would produce a number of results apporaching infinite for practical purposes. Also most good search engines can also provide related data that is not in direct correlation to the original request but might also be of interest to the searcher.
Of course manipuliting a grid is just data manipulation, whether the user has to input all of the parameters or not is a measure of the softwares extensiveness but just because it includes part of the dataset as an integrated peice of the package doesn't mean it's inventing data, just that the software engineer has input some preset data for the user, nothing more.
Of course manipuliting a grid is just data manipulation, whether the user has to input all of the parameters or not is a measure of the softwares extensiveness but just because it includes part of the dataset as an integrated peice of the package doesn't mean it's inventing data, just that the software engineer has input some preset data for the user, nothing more.
Okay, you don't get it. I guess you never programmed at that end of the spectrum. I don't know any other way to explain it.
I didn't say manipulating, I said GENERATING. You want a grid where each cell loosely fits your model, and loosely complies with your ideal cell volume and shape. And you want this to work for any model. You cannot manipulate the model into becoming a grid. You cannot try every possible cell subdivision because it will take forever.
So the CAD package 'invents' a possible solution, and tries to refine it (perhaps with interaction from the user to solve problem-areas). But in no way does the user specify anything about the actual grid-shape. The software does this all by itself.
Just like it would invent a possible result for 1 + 1, assuming we don't know the exact answer (the answer would not follow directly from the inputs 1 and 1, but extra information would be required, which the user cannot provide, so the software would have to invent this information in some way. Perhaps by using a random number generator).
Of course manipuliting a grid is just data manipulation
I didn't say manipulating, I said GENERATING. You want a grid where each cell loosely fits your model, and loosely complies with your ideal cell volume and shape. And you want this to work for any model. You cannot manipulate the model into becoming a grid. You cannot try every possible cell subdivision because it will take forever.
So the CAD package 'invents' a possible solution, and tries to refine it (perhaps with interaction from the user to solve problem-areas). But in no way does the user specify anything about the actual grid-shape. The software does this all by itself.
Just like it would invent a possible result for 1 + 1, assuming we don't know the exact answer (the answer would not follow directly from the inputs 1 and 1, but extra information would be required, which the user cannot provide, so the software would have to invent this information in some way. Perhaps by using a random number generator).
Nope, never really interested me, I see that sort of programming as fundamentally boring. Just inputting datasets and formulas to the end of time to manipulate data in prearranged ways. Nothing much of interest there though I imagine there are technical challenges. However I am interested in OCR and other neural net applications that actually fit the description you are trying to pass off for CAD.
But from the look of the poll (about 50/50 right now) I think it is safe to say that I was at least partially wrong and gave too much credit to browsers as information manipulators. But then it is my opinion, and it can only ever be that as the lines are necessarily murky and unclear. I think bitRAKE and Iblis had it right it is neither and both. But whether you believe it or not I do agree with some of what you say but your opinion is also colored by prejudice and little of it is based on any techinical merit, as is normal in an argument of this type, there is no technical grounds for either side.
But from the look of the poll (about 50/50 right now) I think it is safe to say that I was at least partially wrong and gave too much credit to browsers as information manipulators. But then it is my opinion, and it can only ever be that as the lines are necessarily murky and unclear. I think bitRAKE and Iblis had it right it is neither and both. But whether you believe it or not I do agree with some of what you say but your opinion is also colored by prejudice and little of it is based on any techinical merit, as is normal in an argument of this type, there is no technical grounds for either side.
Just inputting datasets and formulas to the end of time to manipulate data in prearranged ways.
Wrong, as I said, solve cone-cone distance and see what I mean.
There's no prearranged way or formula to do that.
However I am interested in OCR and other neural net applications that actually fit the description you are trying to pass off for CAD.
'Trying to pass off' sounds like you still don't understand it. Perhaps it's easier to see that for OCR there is no direct solution, since your scans are never 100% perfect, so you won't get 100% accurate alignment and 100% perfect bitmaps for fonts.
Sometimes you have to add noise to the input to make it better. Does the user input that noise?
And closer to the original point... Does every OCR package solve non-perfect input the same way? Do you get the exact same results, or does your input work better with one package than with another?
I think this is the difference with browsers. Any half-decent browser will allow you to use this forum, for example. It makes no difference which one you use.
A web browser consumes data and produces visual data or output based on what it has consumed. The visuals are a fundamental aspect that you're overlooking. Information spans far more territory than just textual.
When I write some HTML, for example "<p>hello world!</p>" and I view it through different web browsers, the textual information will generally stay the same, however, the visual information may be different. The text might be in different size or font, it might fill the entire client area, or it might not fit and wrap around the edges. One web browser might even display it with a font that has no distinguishable lowercase letters. The viewer of the page might think that I'm yelling at them.
The thing is, when I wrote that HTML, I never told the browser about which positioning, format, size or font I wanted it to be displayed. The web browser had to decide for itself how to display it, and then produced output above and beyond what the input instructed it to produce. Now, you might argue that the way in which information is presented is trivial to the issue. But anybody with any experience in print or graphic design will quickly tell you otherwise. The proverb, "A picture is worth a thousand words" comes to mind. Sure, lately there are standards for formatting, like CSS and such, but there's no law that says a web browser must conform to them. And many web browsers do not. Others simply cannot.
So, the answer to your poll is 'both'.
When I write some HTML, for example "<p>hello world!</p>" and I view it through different web browsers, the textual information will generally stay the same, however, the visual information may be different. The text might be in different size or font, it might fill the entire client area, or it might not fit and wrap around the edges. One web browser might even display it with a font that has no distinguishable lowercase letters. The viewer of the page might think that I'm yelling at them.
The thing is, when I wrote that HTML, I never told the browser about which positioning, format, size or font I wanted it to be displayed. The web browser had to decide for itself how to display it, and then produced output above and beyond what the input instructed it to produce. Now, you might argue that the way in which information is presented is trivial to the issue. But anybody with any experience in print or graphic design will quickly tell you otherwise. The proverb, "A picture is worth a thousand words" comes to mind. Sure, lately there are standards for formatting, like CSS and such, but there's no law that says a web browser must conform to them. And many web browsers do not. Others simply cannot.
So, the answer to your poll is 'both'.
The ability of diferent browser to use this forum is not related with the choice of content generator- or displayer, is more because the nature of internet, in fact there is not necesity of not allow the broswer modify directly the data in a webserver, but like posted before there are some security reasons for follow this and use this ways.
I think you dont need include complex generations, manipulations of data and user inout, if an application have the 3 points that was stated before, is a content-generator.
A explorer is a polimorfic application, in the way that it can be treated as a only-reader(displayer) when it reads pages of plain html, when it gets pages with forms that are interpreted by the browser it is converted in a content-generator, because it comply with the 3 points.
Also see that the 3 points are more general that the user enter a word or draw with the mouse then process or call a function for blur the data, or search (sort) results.
Also sure, the complexity of a photo is more than a webbrowser, and the manipulations, calculations over the data are far more complex that only save tect with the broswer, but like I say, it follow the 3 points. Maybe not complex data, but is generated starting with the browser.
Also if you whant to see a dynamic way that a simple text can take, only follow this thread and others, when you are drawing, you see the result of anterior changes in the moment, and you can choice continue drawing or aply a effect or other thing, with this simple generation of text you get the result of anterior reading of other people and you can redact a new text, based or not based in what other says before.
Also see that the use of internet of real use not (if I know well) was some time a go, I think maybe 7 years or less, then think that the techniques used actually are in a child face, when they can only manipulate text.
The separation levels that have the browser in the 3 points is for the configuration of a web, centralized data is best that have copies of local databases of the board, the use of php for interpret the commands sended by the browser allow yes diferent browser read the same data, instead that it interpretate only commands sended by a WindowsXP OS, or a freeBSD commands or calls.
For what you will allow an application like photoshop use remote calls, if you have locals that are more fast. Also for what intead of use a direct call, using the stack and such, you will allow ohitishio for call via a text or say a data stream, and let the call be doit interpretating that in other computer. That will have no sense, because a photoshop is designed for be a content generator locally, where the browser can generate data in a "standarized way" a web, say me if not, is for share things, then the capacity of diferent broswer for read/write is because they are designed for that in the web medium.
Also yes, make use of diferent browser impact on the representation of data, if not, you will find some pages that say that the autor recibe info that with IE they can not whatch the page properly, and they say that ms dont follow certain rules.
Have a nice day or night.
Follow what say iblis, I remember time a go, pascal, that a thing confuses me a lot, the diference between procedure and function, in math is very easy to follow, but in that programming language was not, because, in fact, I not return a value on a procedure, but instead I process data, but what happend to the data?, was normally used for display some, for example a procedure for write a text in a formated way, it not return a value to other functions, but instead it returns a direct value to the user.
I think you dont need include complex generations, manipulations of data and user inout, if an application have the 3 points that was stated before, is a content-generator.
A explorer is a polimorfic application, in the way that it can be treated as a only-reader(displayer) when it reads pages of plain html, when it gets pages with forms that are interpreted by the browser it is converted in a content-generator, because it comply with the 3 points.
Also see that the 3 points are more general that the user enter a word or draw with the mouse then process or call a function for blur the data, or search (sort) results.
Also sure, the complexity of a photo is more than a webbrowser, and the manipulations, calculations over the data are far more complex that only save tect with the broswer, but like I say, it follow the 3 points. Maybe not complex data, but is generated starting with the browser.
Also if you whant to see a dynamic way that a simple text can take, only follow this thread and others, when you are drawing, you see the result of anterior changes in the moment, and you can choice continue drawing or aply a effect or other thing, with this simple generation of text you get the result of anterior reading of other people and you can redact a new text, based or not based in what other says before.
Also see that the use of internet of real use not (if I know well) was some time a go, I think maybe 7 years or less, then think that the techniques used actually are in a child face, when they can only manipulate text.
The separation levels that have the browser in the 3 points is for the configuration of a web, centralized data is best that have copies of local databases of the board, the use of php for interpret the commands sended by the browser allow yes diferent browser read the same data, instead that it interpretate only commands sended by a WindowsXP OS, or a freeBSD commands or calls.
For what you will allow an application like photoshop use remote calls, if you have locals that are more fast. Also for what intead of use a direct call, using the stack and such, you will allow ohitishio for call via a text or say a data stream, and let the call be doit interpretating that in other computer. That will have no sense, because a photoshop is designed for be a content generator locally, where the browser can generate data in a "standarized way" a web, say me if not, is for share things, then the capacity of diferent broswer for read/write is because they are designed for that in the web medium.
Also yes, make use of diferent browser impact on the representation of data, if not, you will find some pages that say that the autor recibe info that with IE they can not whatch the page properly, and they say that ms dont follow certain rules.
Have a nice day or night.
Follow what say iblis, I remember time a go, pascal, that a thing confuses me a lot, the diference between procedure and function, in math is very easy to follow, but in that programming language was not, because, in fact, I not return a value on a procedure, but instead I process data, but what happend to the data?, was normally used for display some, for example a procedure for write a text in a formated way, it not return a value to other functions, but instead it returns a direct value to the user.
I agree Iblis,
There is also the fact that the browser can also adjust the image information based on criteria supplied by the user, for example with FireFox I can specify that all images must be resized to fit on the screen, obviously the data displayed does not statically match the data being transmitted. And anyone who has done a dynamic resizing routine knows that it is not a simple matter of snipping out a few lines. There is also the color reduction algorithm for images that are created in more colors than your machine can display, the closest match is sometimes more voodoo science based on perceptual qualities than picking the closest numerical constant. But I geuss it's just too easy to ignore these types of data manipulation, there are more if you need a list.
There is also the fact that the browser can also adjust the image information based on criteria supplied by the user, for example with FireFox I can specify that all images must be resized to fit on the screen, obviously the data displayed does not statically match the data being transmitted. And anyone who has done a dynamic resizing routine knows that it is not a simple matter of snipping out a few lines. There is also the color reduction algorithm for images that are created in more colors than your machine can display, the closest match is sometimes more voodoo science based on perceptual qualities than picking the closest numerical constant. But I geuss it's just too easy to ignore these types of data manipulation, there are more if you need a list.
So, the answer to your poll is 'both'.
Not at all. It doesn't produce any of the information it is visualizing (the content). And that is what we were talking about.
And as you said yourself, there are rules for how to render HTML, and all browsers render pages in pretty much the same way. This is a very different situation from programs that create content and are unique in some way.
The point is, just like with a TV set, all browsers are more or less equal. Your argument would be that some TVs have a slightly larger screen, or other TVs have a bit less contrast, or the aspect ratio may be different... Or even more technical, their internal circuitry is not entirely equivalent. That doesn't change anything about the information you can receive though.
But I geuss it's just too easy to ignore these types of data manipulation
Resizing may change the appearance of the information a bit, but the information itself is preserved. The images still display the same pixels.
And colour reduction is more a case of the image library used than the browser itself, if you ask me (I suppose the OS will handle that?). In any case, I think colour reduction is too trivial to mention here. It changes the appearance a bit, but not the information itself. I mean, a picture of a cow doesn't change into a picture of a monkey by applying resizing or colour reduction.
If you go that far, you can argue that no two browsers are alike, because they are run in different resolutions, on different desktops, different computers etc... But these differences are all trivial, and the informational value of the content is preserved.
It does not work the same as a TV works BTW. A television takes a set amount of lines and draws them, if there are too many lines the remainder are not displayed or the TV is damaged. A browser may adjust the number of lines without substantially altering the image. A television can only display an image in B&W because the image is transmitted in both color and B&W (ie B&W image + Color burst), it cannot modify a color image and make it B&W. In other words a television has no power to modify the information even in the most rudemantary fashion while a browser can modify any part of the image that is necessesary. There is a great difference between changing the voltage supplied to the red gun to get less red than replacing all the red with blue, something a static information display device cannot ever do.
My TV can view anything in B/W, I just have to turn down the colour. Dunno what kind of weird TV sets you have over there.
Isn't that still some kind of filtering though?
You only take information away, you do not generate new stuff.
At any rate, the discussion is now getting far too stupid and annoying.
My point should be clear, and if it's not, then either you do not WANT to agree with me, or you are not capable of understanding what I mean.
Isn't that still some kind of filtering though?
You only take information away, you do not generate new stuff.
At any rate, the discussion is now getting far too stupid and annoying.
My point should be clear, and if it's not, then either you do not WANT to agree with me, or you are not capable of understanding what I mean.
No, it is not filtering. A television image is transmitted in black and white with the color information sent just before the vertical blanking interval in something called the color burst. A television can no more convert an image to black and white than I can change a cat into a dog. It just does not fire the color guns. As you well know the act of changing a color image to black & white involves taking the relative luma of each pixel and converting it to a grayscale value, your browser is capable of doing this, a television by it's very nature as a static displayer of information is not. A television does not have the capacity to alter an image in any way other than to adjust the voltage on the guns. Maybe over there you have some new kind of television but last I heard you were still using PAL which sends color data separately (on a separate channel) from the B&W image.
The discussion is stupid because the question is too vague. "A browser" connotes just that -- a browser. Most of us assumed web browser but even under that context there are many many other web browsers out there. And many produce different output from the same input, or have additional functionality beyond simply processing HTML content.
It doesn't produce any of the information it is visualizing (the content). And that is what we were talking about.
And yet I don't see that qualification in your poll question. Visual information is information, like it or not.
But, if you want to get really technical, then even disregarding visual information, we can still see where a web browser is an information producer by the other facilities it might have. For example, some browsers might keep a list of 'bookmarks' to the users favorite web site. In this manner the web browser functions as an application that produces and maintains a configurable database of URLs. The web browser might also inform the user that it has completed a file transfer, or it might generate a report based on average transfer speed over n time, or it might even have native support for web page creation without any third party involvement.
If you would not like to consider these other facilities of web browsers than perhaps you need to further qualify your poll question by focusing on the just the module that processes HTTP data, and not the entire web browser.
It doesn't produce any of the information it is visualizing (the content). And that is what we were talking about.
And yet I don't see that qualification in your poll question. Visual information is information, like it or not.
But, if you want to get really technical, then even disregarding visual information, we can still see where a web browser is an information producer by the other facilities it might have. For example, some browsers might keep a list of 'bookmarks' to the users favorite web site. In this manner the web browser functions as an application that produces and maintains a configurable database of URLs. The web browser might also inform the user that it has completed a file transfer, or it might generate a report based on average transfer speed over n time, or it might even have native support for web page creation without any third party involvement.
If you would not like to consider these other facilities of web browsers than perhaps you need to further qualify your poll question by focusing on the just the module that processes HTTP data, and not the entire web browser.