Check out the link, provided by chad, to William Harris’ article---that's where I first got the idea. I started this method with the beginning of the Iliad, so those lines were my first attempt. At that point, I was relying almost exclusively on Pharr, and then discovered Perseus shortly afterward. Later, I purchased Cunliffe, and began embarking on quixotic adventures in quest of commentaries at distant libraries.
I was disheartened to find that Homer used the augment sporadically, so I found myself noting unaugmented words. I also tried to note the root if it wasn't clear to me.
My general pattern for nouns was---for example---[face=spionic]mh=nin[/face]
, acc. sg. f. And for verbs---[face=spionic]a)/eide[/face]
, pres. act. imperat. 2d sg, ([face=spionic]a)ei/dw[/face]
). But I never really standardized my notations, which I regret slightly. It was really only limited by what information I could find; for example, I have a note for [face=spionic]ou)lome/nhn[/face]
: acc. sg. f, ([face=spionic]ou)lo/menoj[/face]
, 2aor. med. part., [face=spionic]o)/llumi[/face]
, used as adj., in acc.)---I must have found this somewhere else and added it later, because I don’t think it’s from Pharr. I also made a note for [face=spionic]dai=ta[/face]
, when I discovered Perseus, that some read [face=spionic]pasi[/face]
. [Does anyone know about that debate? is it ongoing?]
I also found it helpful to draw lines connecting words. Early on, I would note things like a connection in Line 2 of [face=spionic]muri/'[/face]
, which is clear already from the use of the accusative, but I felt that a connecting line allowed me to keep my eye on the Greek rather than look my notes (I felt that the elision hampered recognition). Usually, though, it was to indicate connected words between lines. But Homer's word order is more managable than others, like chad says, and you probably already know. I haven’t read Pindar yet, but I experienced a reality-check when I tried to decipher some of the tragedians. [I find the sentence diagrams on Rodney Decker's website helpful, even though NT syntax is relatively simple. I think they are computer generated. Is there something like this for the tragedians, Pindar, Homer, et al?]
Whether or not to use English is an important question. I like to use flashcards for vocabulary, so I don’t translate words for which I have made cards. But this method was, for me, a way to record language notes for quick reference; so, when I found myself looking at a dictionary, I usually either made notes on these pages, or created a new flashcard. But until I found a used copy of the OCT Iliad, I used to write out Greek-only pages to read, without any notes, so that I would focus on the Greek syntax (and because I secretly enjoyed writing Greek *blush*). I think I recall William Harris stressing the importance of understanding how the meaning develops from the Greek word order.
The style and content of your notes should suit your purposes. Now I find myself browsing and searching Textkit as a large collective notebook.