Information Technology Reference
In-Depth Information
appendix is to hold bulky material that would otherwise interfere with the narrative
flow of the paper, or material that even interested readers do not need to refer to.
Appendices are only occasionally necessary for a paper, in cases where there is
material such as a proof whose length would interrupt the flow. But they often have a
useful role in a thesis, where they can be used for supporting material such as ethics
approvals, extended tables of data, and transcripts of interviews.
The First Draft
For the first draft of a write-up you may find it helpful to write freely—without
particular regard to style, layout, or even punctuation—so that you can concentrate
on presenting a smooth flow of ideas in a logical structure. Worrying about how to
phrase each sentence tends to result in text that is clear but doesn't form a continuous
whole, and authors who are too critical on the first draft are often unable to write
anything at all. If you tend to get stuck, just write anything, no matter how awful;
but be sure to delete any ravings later.
Some people, when told to just say anything, find they can write freely—if any-
thing is acceptable, then nothing is wrong. For others, finding words is a struggle. A
last resort is to write in brief sentences making the simplest possible statements.
In-memory sorting algorithms require random access to records. For large
files stored on disk, random access is impractically slow. These files must be
sorted in blocks. Each block is loaded into memory and sorted in turn. Sorted
blocks are written to temporary files. These temporary files are then merged.
There may be many files but in practice the merge can be completed in one
pass. Thus each record is read twice and written twice. Temporary space is
required for a complete copy of the original file.
This text certainly isn't elegant—it is annoying to read and should be thoroughly
edited before the paper is submitted. But it is capturing the ideas, and the writing is
proceeding.
A consequence of having a sloppy first draft is that you must edit and revise
carefully; initial drafts are often awkwardly written and full of mistakes. But few
authors write well on the first draft anyway. The best writing is the result of frequent,
thorough revision.
Mathematical content, definitions, and the problem statement should be made
precise as early as possible. The hypothesis and the results flow from a clear state-
ment of the problem being tackled. Describing the problem forces you to consider
in depth the scope and nature of the research. If you find that you cannot describe
the problem precisely, then perhaps your understanding is lacking or the ideas are
insufficiently developed.
It was said earlier, but is worth repeating: the writing should begin as soon as the
research is started. Right from the start, expect to accumulate useful fragments of text
that will later be drawn into the finished write-up. The later the writing is begun, the
