How To Criticize Computer Scientists

link:
http://www.cs.purdue.edu/homes/dec/essay.criticize.html

How To Criticize Computer Scientists
                or
                Avoiding Ineffective Deprecation And
                Making Insults More Pointed
               
In recent exchanges, members of the faculty have tried in vain to attackother Computer Scientists and disparage their work.Quite frankly, I find the results embarrassing -- instead of cutting theopponent down, many of the remarks have been laughably innocuous.Something must be done about it because any outsider who hears such blatherwill think less of our department: no group can hold the respect of othersunless its members can deal a devastating verbal blow at will.
This short essay is an effort to help faculty make their remarks morepointed, and help avoid wimpy vindictives.It explains how to insult CS research, shows where to find the Achilles'heel in any project, and illustrates how one can attack a researcher.


The Two Basic Types Of Research
Most lousy insults arise from a simple misimpression that all researchersagree on the overall aims of CS research.They do not.In particular, CS has inherited two, quite opposite approaches fromroots in mathematics and engineering.


Researchers who follow the mathematical paradigm are calledtheorists, and include anyone working in an area thathas the terms ``analysis'', ``evaluation'', ``algorithms'', or ``theory''in the title.

Researchers who follow the engineering paradigm are calledexperimentalists, and include most people working in areas thathave the terms ``experimental'', ``systems'', ``compiler'', ``network'',or ``database'' in the title.


Complex Theory And Simple Systems
Knowing the tradition from which a researcher comes provides the basisfor a well-aimed insult.

Theorists Favor Sophistication   
Like mathematicians, theorists in Computer Science take the greatestpride in knowing and using the most sophisticated mathematics to solve problems.For example, theorists will light up when telling you that they havediscovered how an obscure theorem from geometry can be used in theanalysis of a computer algorithm.Theorists focus on mathematical analysis and the asymptotic behavior ofcomputation; they take pride in the beauty of equations and don't worryabout constants.Although they usually imply that their results are relevant to realcomputers, they secretly dream about impressing mathematicians.

Experimentalists Favor Simplicity   
Like engineers, systems researchers take pride in being able to inventthe simplest system that offers a given level of functionality.For example, systems researchers will light up when telling you that theyhave constructed a system that is twice as fast, half the size, andmore powerful than its predecessor.Experimentalists focus on the performance of real computer systems;they take pride in the beauty of their code and worry about constants.Although they usually imply that their results can extend beyond realcomputers, they secretly dream of filing patents that apply to extanthardware.

.--

The Insult
Knowing that CS can be divided into two basic groups helps immenselywhen criticizing someone.There are two basic rules:identify the type of the researcher andissue an insult for that type.Avoid saying anything that inadvertently compliments them.If performed well, an insult will not only stun the researcher (who willbe shocked to learn that not everyone agrees with his or her basic valuesystem), but will also intimidate others in the audience.


Identifying A Type
. 1point 3acres Identifying the type of a researcher is usually easy and does not requirea strong technical background or real thinking.It can be done using keyword matching according to the following lists.

Detecting Theory
You can tell someone is a theorist because they slip one or more of thefollowing keywords and phrases into lectures and technical conversations:``theorem'',``lemma'',``proof'',``axiom'',``polynomial time'',``logarithmic'',``semantics'',``numerical'',``complexity'',``nondeterministic'' or ``nondeterminism'',and``for large enough N''.They write lots of equations,brag about knocking off the ``extra log factor'',and often end their lecture with an uppercase ``O''followed by a mathematical expression enclosed in parentheses.You can also recognize a theorist because they take forever toprove something that may seem quite obvious. (I once sat through anhour lecture where someone proved that after a computer executed anassignment statement that put the integer 1 into variable x,the value in x was 1.)

Detecting Systems
An experimentalist will slip one or more of the following keywords andphrases into lectures and technical conversations:``architecture,''``memory,''``cpu''(sometimes abbreviated``CISC'' or ``RISC''),``I/O'' or ``bus'',``network'',``interface'',``virtual'',``compile'' or ``compiler'',``OS'' or ``system'',``distributed'',``program'' or ``code'',and``binary''.They talk about building programs and running the resulting system onreal computer systems.They refer to companies and products, and use acronyms liberally.Their lectures often end with a graph or chart of measured system performance.You can also recognize an experimentalist because they describe inexcruciating detail how they set up an experiment to measure a certainvalue even if the measurement produced exactly the expected results.(I once sat through an hour lecture where someone carefully explained howthey used three computer systems to measure network traffic, whentheir whole point was simply to show that the network was notthe cause of the problem they were investigating.)

Forming An Insult
The key to a good insult lies in attacking whatever the researcher holdsmost dear and avoiding whatever the researcher does not care about.Thus, an insult lobbed at a theorist should focus on lack of sophisticatedmathematics such as the following:

•         Despite all the equations, it seems to me that your work didn't        require any real mathematical sophistication.  Did I miss something?        (This is an especially good ploy if you observe others struggling to        understand the talk because they will not want to admit to that        after you imply it was easy.)
•         Isn't this just a straightforward extension of an old result by        Hartmanis? (Not even Hartmanis remembers all the theorems Hartmanis        proved, but everyone else will assume you remember something they        have forgotten.)
•         Am I missing something here?  Can you identify any deep mathematical        content in this work? (Once again, audience members who found the        talk difficult to understand will be unwilling to admit it.)

. In contrast, an insult lobbed at an experimentalist should imply that thetechniques were used in previous systems or that the work isn'tpractical such as:

•         Wasn't all this done years ago at Xerox PARC?  (No one remembers        what was really done at PARC, but everyone else will assume you        remember something they don't.)
•         Have you tested this on the chip Intel got running last week in        their lab?  (No one knows what chip Intel got running last week,        but everyone will assume you do.)
•         Am I missing something?  Isn't it obvious that there's a bottleneck        in the system that prevents scaling to arbitrary size?  (This is safe        because there's a bottleneck in every system that prevents arbitrary        scaling.)

How To Avoid Having An Insult Backfire On You
A misplaced insult can backfire, turning into an embarrassment forthe attacker and a victory for the intended attackee.To avoid such occurrences, remember the following:

• Never attempt to attack theoretical work as not considering constants,as unrelated to real computer systems, or as requiring too muchsophisticated mathematics.(The intended victim is likely to smile and thank you for the flattery.)
• Never attempt to attack a system as too small, too simple, or as lackingsophisticated mathematics(Again, the intended victim is likely to smile and thank you for theflattery.)
• Never attempt to attack systems work simply by saying that it's sosimple and obvious that you could have done it.(For years, people said that about UNIX and the TCP/IP protocols.)In fact, this is merely an extension of a ploy used by children on aplayground: ``Oh yeah? I could have done that if I wanted to.''Don't try using it or someone will tell you to grow up.

Attacking Crossover Work
Although rare, a few researchers include both theoretical and experimentalwork in the same project.Insulting such combinations can be tricky because a researcher can escapeunscathed by pointing to one part of their work or the other as the answer.You can try to attack both parts simultaneously:

• I note that the systems aspect of this project seems quite complex.Do you think the cause of the convoluted implementation can be attributedto the more-or-less ``simplistic'' mathematical analysis you used?

However, a clever insult can avoid talking about the work by suggestingsinister reasons for the paradigm shift:

• I notice that you did something unusual by combining both theory and experiment.Did you decide to try a second approach because you had insufficientresults from the first?
• You seem to have a little theory and a little experimental work combinedinto one project.Isn't it true that if you had a sufficiently strong contribution in oneor the other you would have lectured about them separately?

A Final Plea
I certainly hope faculty will take this essay to heart and sharpen theirinsult skills.In the future please make all your thrusts count.

interesting~

哈哈哈好有意思～～
前半段说的还有板有眼，后面就开始放肆调侃～～