Well, Daniel, I can see you are not going to frame my column and put it in the place of honor in your office. Just a few comments about your points.
I rather liked it, actually. Of course, I totally disagreed with it, but that’s a different thing. I think you are a rather talented columnist and “We Are Sorry to Inform You… ” was memorable. I would frame “We Are Sorry to Inform You… “, but it was a multi-page article as I recall and those are hard to frame.
You might find reassuring to check out my pub. list… here are some titles:
– A Better Alternative to Piecewise Linear Time Series Segmentation
– One-Pass, One-Hash n-Gram Count Estimation
– Hierarchical Bin Buffering: Online Local Moments for Dynamic External Memory Arrays
– An Optimal Linear Time Algorithm for Quasi-Monotonic Segmentation
– …
So, no trace of standards there… you see? 😉 I’m mostly clean.
1. I do think that there is some hint that computing research has suffered from excessive attention to standard, (…)
Let me skip over the quality of software part, which is not interesting to me at this time. How was computing research impacted by excessive attention to standard?
2. Information Technology is not, as my quality standards go, fit material for academic research. if nothing else, for lack of internal coherence. I am not even quite sure what information technology is.
Publications like Communications of the ACM are probably 20% IT, 20% information systems, 40% software engineering and 20% CS these days. IEEE Computing is certainly not predominately CS.
I might as well be frank: most of the articles in either CACM or IEEE Computing are intensely boring to me. Yours excepted. Still, pure CS is a very tiny fraction of “computing research” these days. Many, many, many colleagues are into web services, semantic web, e-Learning, and so on. Surely, that must be true in Spain as well.
Is IT fit material for academic research? Well, I’m not an IT researcher so I’ll invite IT researchers to defend themselves. However, lots of people from prestigious schools have spent a lot of time documenting what IT is and how it ought to be taught:
I’d be generally curious about your criteria. Is software engineering fit for academic research? Is Information Systems fit for academic research? On which basis do you decide whether there is sufficient “internal coherence”?
From the point of view of computing science, an XML document is an unranked labeled tree, nothing more.
Hmmm… so the Web is nothing but a directed graph? An UML diagram is nothing but a fancy graph. An image is nothing but a two-dimensional array? We agree then there is no sense ever considering the Web or images, or videos, or any other Information Technology construct in Computer Science since these are only instances of perfectly reasonable mathematical constructs we can study without being bothered by standards (such as those governing image formats, on web protocols, and so on).
Ooops… just checked out your pub. list! Oh! My! Web Graphics Recommender System? Isn’t the Web an IT construct? A social phenomenon? (I’m teasing you… don’t answer this one…)
If it has a great social importance, that is a study matter for a sociologist, not a computing scientist.
I hope sociology is interested by XML! However, from this line of argument, we could derive the fact that human-computer interaction (HCI), since it is focused on how human beings perceived things, belongs squarely in psychology… which is too bad since many CS departments offer HCI courses and there are numerous HCI researchers who consider themselves Computer Scientists. I know many of those.
Computers are formal symbol manipulation machines, and I can’t see how their study can be other than mathematical in nature.
Software engineering is not very mathematical to me… Project management, unit testing, design specifications, business processes, skills analysis… is a software engineering researcher a computing researcher?
Simone Santinisays:
Well, Daniel, I can see you are not going to frame my column and put it in the place of honor in your office. Just a few comments about your points.
1. I do think that there is some hint that computing research has suffered from excessive attention to standard, and, maybe more importantly from your point of view, that industrial software quality has also suffered. For one thing internet software, which is by far the type that most relies on standards, is also by far the software with the poorest quality. The relation is, in any case, worth studying.
2. Information Technology is not, as my quality standards go, fit material for academic research. if nothing else, for lack of internal coherence. I am not even quite sure what information technology is. From the point of view of computing science, an XML document is an unranked labeled tree, nothing more. If it has a great social importance, that is a study matter for a sociologist, not a computing scientist. Computers are formal symbol manipulation machines, and I can’t see how their study can be other than mathematical in nature.
True, the W3C does not issue standards in the sense that ISO does. Let us call them “de facto” standards, for lack of a better word. As to the 40 years and FORTRAN, my argument was completely within the limits of research: the innovation over existing standards was necessary for the internal development of computing research. That it should have a positive effect on the industry should be welcomed, but it should not be what motivates research.
Well, Daniel, I can see you are not going to frame my column and put it in the place of honor in your office. Just a few comments about your points.
I rather liked it, actually. Of course, I totally disagreed with it, but that’s a different thing. I think you are a rather talented columnist and “We Are Sorry to Inform You… ” was memorable. I would frame “We Are Sorry to Inform You… “, but it was a multi-page article as I recall and those are hard to frame.
You might find reassuring to check out my pub. list… here are some titles:
– A Better Alternative to Piecewise Linear Time Series Segmentation
– One-Pass, One-Hash n-Gram Count Estimation
– Hierarchical Bin Buffering: Online Local Moments for Dynamic External Memory Arrays
– An Optimal Linear Time Algorithm for Quasi-Monotonic Segmentation
– …
So, no trace of standards there… you see? 😉 I’m mostly clean.
1. I do think that there is some hint that computing research has suffered from excessive attention to standard, (…)
Let me skip over the quality of software part, which is not interesting to me at this time. How was computing research impacted by excessive attention to standard?
2. Information Technology is not, as my quality standards go, fit material for academic research. if nothing else, for lack of internal coherence. I am not even quite sure what information technology is.
Publications like Communications of the ACM are probably 20% IT, 20% information systems, 40% software engineering and 20% CS these days. IEEE Computing is certainly not predominately CS.
I might as well be frank: most of the articles in either CACM or IEEE Computing are intensely boring to me. Yours excepted. Still, pure CS is a very tiny fraction of “computing research” these days. Many, many, many colleagues are into web services, semantic web, e-Learning, and so on. Surely, that must be true in Spain as well.
Is IT fit material for academic research? Well, I’m not an IT researcher so I’ll invite IT researchers to defend themselves. However, lots of people from prestigious schools have spent a lot of time documenting what IT is and how it ought to be taught:
http://www.acm.org/education/curric_vols/IT_October_2005.pdf
I’d be generally curious about your criteria. Is software engineering fit for academic research? Is Information Systems fit for academic research? On which basis do you decide whether there is sufficient “internal coherence”?
From the point of view of computing science, an XML document is an unranked labeled tree, nothing more.
Hmmm… so the Web is nothing but a directed graph? An UML diagram is nothing but a fancy graph. An image is nothing but a two-dimensional array? We agree then there is no sense ever considering the Web or images, or videos, or any other Information Technology construct in Computer Science since these are only instances of perfectly reasonable mathematical constructs we can study without being bothered by standards (such as those governing image formats, on web protocols, and so on).
Ooops… just checked out your pub. list! Oh! My! Web Graphics Recommender System? Isn’t the Web an IT construct? A social phenomenon? (I’m teasing you… don’t answer this one…)
If it has a great social importance, that is a study matter for a sociologist, not a computing scientist.
I hope sociology is interested by XML! However, from this line of argument, we could derive the fact that human-computer interaction (HCI), since it is focused on how human beings perceived things, belongs squarely in psychology… which is too bad since many CS departments offer HCI courses and there are numerous HCI researchers who consider themselves Computer Scientists. I know many of those.
Computers are formal symbol manipulation machines, and I can’t see how their study can be other than mathematical in nature.
Software engineering is not very mathematical to me… Project management, unit testing, design specifications, business processes, skills analysis… is a software engineering researcher a computing researcher?
Well, Daniel, I can see you are not going to frame my column and put it in the place of honor in your office. Just a few comments about your points.
1. I do think that there is some hint that computing research has suffered from excessive attention to standard, and, maybe more importantly from your point of view, that industrial software quality has also suffered. For one thing internet software, which is by far the type that most relies on standards, is also by far the software with the poorest quality. The relation is, in any case, worth studying.
2. Information Technology is not, as my quality standards go, fit material for academic research. if nothing else, for lack of internal coherence. I am not even quite sure what information technology is. From the point of view of computing science, an XML document is an unranked labeled tree, nothing more. If it has a great social importance, that is a study matter for a sociologist, not a computing scientist. Computers are formal symbol manipulation machines, and I can’t see how their study can be other than mathematical in nature.
True, the W3C does not issue standards in the sense that ISO does. Let us call them “de facto” standards, for lack of a better word. As to the 40 years and FORTRAN, my argument was completely within the limits of research: the innovation over existing standards was necessary for the internal development of computing research. That it should have a positive effect on the industry should be welcomed, but it should not be what motivates research.