A Short Guide to Good Coding describes the very minimal good habits to make software understandable by yourself, for debugging, and by others, for possible reuse. Read it before writing code for a semester or master project at LAP.
A Short Guide to Presentations gives you some hints to give a good public presentation, including how to plan it and how to prepare the slides. Make sure you read it carefully if you have to give a semester or master project presentation at LAP. The references at the end are in the LAP Library or can be obtained at the LAP Secretariat (the HBR article). You may also want to check this oral presentation advice by Mark H. Hill (and David A. Patterson).
If you are looking for writing style references or for other books which can improve your written communication skills, the following may interest you (all books are in the LAP library under the XXX99 tags shown below):
- The Elements of Style by Strunk and White [STR99] is one of the great classics. In less than hundred pages, it spans several fundamental topics in English writing: rules of usage (“shall I use a comma or a dash?”), principles of composition (“why should I use the active voice?”), a few matters of form (“how do I introduce formal quotations?”), and reviews classic misused expressions (“that or which?”). Not only: 20 pages contain a couple of dozens “subjective” recommendations on how to make your prose more distinguished and effective. A must-read!
- 100 Ways to Improve Your Writing by Provost [PRO72] is also a classic reference, very complementary to the previous. Most of the “ways” deal with the very process of writing: where to start, how to catch the reader’s attention, how to be efficient and effective, how to avoid negative reactions from the reader, etc. It is as good for writing a scientific paper as for writing a novel—in case you get bored with your PhD work….
- The Chicago Manual of Style [online from EPFL and with VPN] is probably the most authoritative reference in professional editing of English prose. It describes in great details what makes a book, how many pages should come before the main matter, how they should be numbered, what should they contain, etc. But the largest part of the book deals with editing style topics, down to the smallest detail. Some people may find it fun to read as a compelling novel (as indeed this writer does), but most people can just make use of the extremely complete index at the end (more than 40 dense pages in the paper version) to locate quickly the answers to their worries. Please note that the index is available also in the online version and is often a much more efficient way to find an answer than random web searches. By default, you can assume that any answer in matters of editing is in there and can be blindly relied upon!
- A quite different book is The Visual Display of Quantitative Information by Tufte [TUF01]. It is a gorgeous book which helps you understand what makes a good graph very different from a bad graph (the latter often be the automated product of a common software tool—chartjunk, as Tufte calls it). It addresses both content (“which data for a graph?”) and form (“which graph for your data?”). About content, many pages are devoted to the fundamental topic of graphical integrity and sophistication. The second part of the book illustrates fundamental principles such as the maximisation of data-ink, the ink which is used to represent information as opposed to the ink invested in grids, shadows, decoration, etc. The author proves fundamentally wrong the popular idea that good graphs are simple and based on limited datasets: if you have five numbers to show, you should rather make a table; if you have millions of numbers to show, you can only try to think of an effective graphical representation. Tufte proves by examples that the best graphical representations can have data densities of many thousands numbers per square centimeter! The book contains one of the most impressive collection of wonderfully reproduced historic and modern examples of best and worst real graphs. Read this book and your graphical output will be changed forever….
A nice summary of common errors in papers and bibliographies has been compiled by John Owens of UC Davis. This writer finds it quite matching his experience of the usual horrors encountered when reviewing papers—in fact, what is utterly depressing is that they are always the same and that it would be so incredibly easy to avoid them if anyone just bothered….
Bibliographies are usually a particularly painful topic. Since almost everybody uses BibTeX, it seems that bibliographies should be spic and span with almost no effort. In fact, they are usually horrific. This is especially due to the fact that most websites these days offer some sort of BibTeX output which is invariably of abysmal quality and violates (often in many respects) the BibTeX specifications. Here is a copy of the BibTeX specification reproduced from the original 1986 book on LaTeX by Leslie Lamport [LAM86]. Please, (1) avoid downloading BibTeX entries from any website and instead (2) read the above nine pages and enter the entries manually. The above-mentioned summary of common errors by John Owens of UC Davis contains many good suggestions and appears to share this writer’s astonishment at the total disregard for the BibTeX specification by major publishers in our areas. Finally, also note that in the LAP CVS repository, there is a file
bibl.bib which can serve as a reference, with many constants defined for conferences and journals of interests, a memento of the Chicago Manual of Style rules on continuation numbers (for page ranges), etc. Please do not update the file
bibl.bib directly (send updates and additions to Paolo instead) but use it as a source of templates.