The timing is half of the reading
A glucose number without a time attached is close to useless, which is why every line here carries one. Write fasting, before meal, 1h after meal, 2h after meal or bedtime and the readings are grouped and averaged separately. Writing just after meal is read as the two-hour check, which is the one most meters and most logbooks mean by it. A line with nothing recognisable goes into the overall mean but is left out of the target scoring, because there is no honest way to say whether an untimed number was in range. The date can be MM/DD/YYYY or YYYY-MM-DD, and the value can sit anywhere on the line — markers like 2h are stripped out before the number is picked up so they cannot be mistaken for the reading.
Units, and the factor of eighteen
US meters read in mg/dL and most of the rest of the world reads in mmol/L. The two are the same measurement in different units: one mmol/L is about 18 mg/dL, so 5.5 mmol/L is roughly 99 mg/dL and 10 mmol/L is roughly 180. If you pick mmol/L in the units box, every value you typed is multiplied by 18.0182 and the whole page then works in mg/dL, with the mmol/L figure shown alongside the headline mean. The target boxes are always mg/dL. Mixing units within one log will not work — pick the units your meter shows and stay with them.
The target range is yours, not ours
The boxes open at 80-130 mg/dL before meals and under 180 two hours after, because that is the range printed in most general adult diabetes education material. Treat it as a placeholder. Targets are set individually, and they are looser for someone older or prone to hypoglycemia, tighter in pregnancy, and different again with kidney disease or with the particular medicines someone is taking. Bedtime has no default here at all: the boxes are empty and bedtime readings are averaged but not scored unless you fill both in, because there is no single bedtime number that fits everyone. Whatever you put in those boxes is what the percentages count against. The count is arithmetic against a line you chose, and it is not a judgement about control.
Reading the estimate carefully
The A1c figure is a straight algebraic conversion of the mean of whatever readings you typed, using the ADAG relationship between average glucose and A1c. It inherits every bias in the sample. A log of nothing but morning fasting checks has a mean lower than the true 24-hour average, so its estimate comes out low; a log built around post-meal checks does the opposite. Continuous monitors produce the kind of coverage the equation assumed; a finger-stick diary rarely does. Take the number as a rough position, compare it to your actual lab result rather than instead of it, and if the two disagree, the lab is the one with the blood sample.
Questions people ask
My meter reads in mmol/L. Do I have to convert every line?
No. Set the units box to mmol/L and type the numbers exactly as the meter shows them. Everything is multiplied by 18.0182 internally and the page then works in mg/dL, with the mmol/L figure repeated next to the headline mean. The target boxes stay in mg/dL, so a 4-7 mmol/L target becomes roughly 72-126 there.
The A1c estimate does not match my last lab result.
A mismatch is expected. The equation assumes readings spread across whole days, and a diary of fasting checks or of post-meal checks is skewed in a predictable direction. Anemia, kidney disease and anything affecting red-cell turnover shift the real relationship as well. The blood test is the measurement; this is an arithmetic estimate of what your typed numbers average out to.
Is anything saved?
No. The whole page runs in your browser and a refresh clears the box. To keep a running log, copy the text output into a note or a file — pasting it back reads straight in, and new lines can be added to the bottom.