The logbook, or laboratory notebook

A good software designer keeps a logbook, indeed properly maintained laboratory notebooks have significant scientific, legal, and administrative value. Keeping a logbook is good practice, and is necessary to document the progress of a piece of work. Here are the three conventional roles of the logbook:

1. As a personal journal, it is a store of your own knowledge and experience. It is a way of facilitating your thinking as you work on the project.
2. As a commentary for others, it can be used by colleagues to understand and reconstruct the work. Often, in a commercial environment, employees leave or are transferred to another part of the company. In such circumstances, a logbook is invaluable for those that must continue the work.
3. As a legal document, it can be used as evidence to support claims about dates of inventions and developments. (In this case, the logbook is usually attested to by witnesses, and more formal rules are followed about what does and does not go in the book.) In an industrial setting, good laboratory record keeping has always been necessary in order to preserve intellectual property rights, such as patents and know-how. Moreover, as universities continue to seek out more collaborations with industrial scientists, academic research laboratories will have to bring their record keeping up to industrial standards.

Clearly, only the first of those is directly applicable in the current context of an assessed course. However, you must get into the habit of keeping a logbook, as the latter two roles will be relevant later in your career. In order to be a good future employee, you must develop proper laboratory record keeping skills.

Even in an industrial setting, although the logbook will be used every day by the writer, it may never be read by anyone else, and most likely it will never act as legal evidence. While these roles cannot be neglected, in fact, they have never been more important, it is in the certain role of journal that the logbook is of most benefit.

Like any other journal, an engineering logbook is dated but otherwise unstructured, written once, not revised, and gives a personal perspective. It provides for your memory a framework for a series of "snapshots" of your work as it progresses. A designer needs their logbook to jog their memory as effectively as possible.

Amazingly, even if the wrong things are recorded, even if what then seemed important now looks trivial, even if there are many gaps, the logbook will probably still succeed. Any cues provided by the logbook at a later date are often sufficient to stimulate the mind into recovering what matters. The logbook is, then, a surprisingly robust partner to the writer's memory.

The logbook is a kind of user interface to the designer's knowledge, some of which is in the book and some of which is in the head. The most reliable way to create such an interface is to assume that the knowledge which will stay in the head, is whatever a reasonably proficient person in the field would have in their head anyway. This means, the way to write a good logbook is to write it for someone else. When deciding what should go in the book, the answer is, whatever commentary someone else would need to follow what is going on. This neatly addresses roles 2 and 3 of the logbook too.

It could be argued, fairly tritely, that in six months time the logbook author will be someone else (from the point of view of understanding what was done)! But the main reason for writing as if for someone else is that it gives the opportunity to practice "public" writing without cost. The writing need not be in prose—almost certainly it will be in point form, with plenty of diagrams—but it will be written to communicate. There is no overwhelming reason to suppose that careful composition of the logbook is going to be slower in the long run. One who practices writing carefully, will learn to write carefully quickly. And that combination is a most valuable communication skill.

So write the logbook as a commentary for someone else, providing enough information for them to understand your decisions, follow your reasoning and reconstruct your work. The "someone else" is an informed student on the same course, not an ignoramus, and not a genius.

When selecting a logbook and when making the entries themselves, you should observe the following rules.

• Use bound laboratory notebooks that open in a book format with left and right pages, and to which extra pages cannot be added.
• If successive notebooks are used, the volumes should be numbered sequentially.
• Make all entries using a pen with an ink colour that is durable and easily photocopied, such as dark blue or black (some blue ink is too light and is not recommended, and red ink can fade).
• Pencil is not allowed because data can be changed.
• Date each day's entries (you would also be required to sign them in some industrial environments).
• Subsequent entries should be made on the line immediately following the previous entry's date.
• All unused lines on the prior page must be crossed out if one starts a new page for a new entry.
• Make all entries in chronological order, by date. Exceptions to this include tables that may be added to or completed through time.
• Write in the first person singular, and name names when referring to other people's work.
• Do not erase, scribble through, or blot out any entry at any time. Make all corrections with a single line through the incorrect entry, and provide a reason or explanation for the correction if it's not obvious.
• Write the logbook once. Do not go back and carry out tweaks and adjustments afterwards.
• Separate pages such as forms or computer printouts may be attached to a blank logbook page. Attach such pages so that no data is covered under the page (that is, attached pages should not be "layered").
• Avoid skipping pages or leaving large blank areas. If needed, mark through large blank areas with a diagonal line.
• Use language that is objective, factual, and free of personal feelings or inappropriate terminology.
• Pages or portions of a page must not be removed from a logbook for any reason.
• Lapses between entries should be explained by making an entry, such as, "Working on other activities", followed by dating the explanation.

A logbook should be a real book, not soft documentation stored on a computer. Despite this, it is true that there are benefits in keeping a soft logbook: it is in the same medium as programs; most programmers type better than they handwrite; automatic date stamping is possible; programs and logbook can be effectively cross referenced—perhaps even linked with hypertext; it is easy to share materials with colleagues. But against these, there are strong counter-arguments:

1. No hardware platform yet combines the portability and economy of a book. Portability is vital because the logbook is used in the problem domain (perhaps during meetings with customers, perhaps gathering information on-site), in the library, on the bus, and even at home.
2. No software base yet is as convenient as pencil/pen and paper for multimodal input involving text, graphics and annotations. Drawing with a fine instrument on a good-sized canvas is an important facility.
3. No hardware/software combination yet gives such a powerful range of browsing, searching, multiple-view facilities as a book.
4. Logbook material stored in a computer is likely to suffer fates similar to other soft documentation. It will not be kept up to date, old material will be mistrusted, it will get accidentally deleted, then abandoned in the belief that what it contains was not important enough to justify recovery from backup. Because of its flexibility, a soft logbook is also susceptible to tinkering: some authors find it difficult not be spend time improving it and bringing it up to date after the fact.