Some notes in tutorial about assignment three

The answers to parts 1 and 2 of the assignment are spreadsheets. We advise against starting in some other program and then attempting to convert your file to gnumeric; this is tricky and students end up with zeroes for trying this and submitting something unreadable. If you want to work at home, you could play with your formulas in some other spreadsheet program, but you really have to come in to the lab to produce your final file, and you should be willing to re-type the formulas as needed.

Note that you are producing a model. Someone using the spreadsheet should not have to write new formulas or modify the supplied formulas.

Some notes on part 1:

You submit a Gnumeric spreadsheet (only). You can run gnumeric by typing "gnumeric" at the shell prompt, or "gnumeric" and a space and a file name. Gnumeric files are usually named ending with the unwieldy extension ".gnumeric".

The lab 4 graded bit is substantial background. In that lab you determined, for a given date, the day number within a year. In this assignment you will do this twice, for two dates, and then subtract them.

The use of abs() was discussed in the tutorial at this point.

Some notes on part 2:

Again, what you submit is a gnumeric spreadsheet.

Example calculations were discussed here, and how CEILING applies.

There is a table of parameters at the top which control the calculations through the rest of the spreadsheet. The user of your resulting spreadsheet can enter any values in this table, and this changes what happens below, without changing any of the formulas below. For example, the formulas must not contain numbers such as 20 (for the size of the area) or 0.4 (for the size of each tile). Those are just examples, and the user of the spreadsheet might type other things.

For the width column, you just say "0.5 more than [the previous value]" on each line. You are required to write a formula for this too. If the top number is changed, the column should change.

Remember to use absolute cell references where necessary (with the dollar signs) to refer to the cells in the table at the top, so that the references are NOT adjusted when the formulas are copied around. When writing a formula, some references we want to be adjusted when the formulas are copied around, and some we want not to be adjusted, and we select this by whether or not we use the dollar signs.

You don't need hlookup() in tiles.gnumeric.
The assignment requests the two tables, that's all. You don't have to attempt to find the minimum number of tiles, etc.
It is a human judgement which size is best. The spreadsheet should give the human all of the data so that they can decide.

Part 3 notes

For part 3, all I asked the TAs to discuss in this tutorial is the issue of using only structural tags. If you stick to structural tags, then the web browser is able to implement everything its author knows about typography and screen layout, which should, ideally, be more than the average writer knows. Users can change sizes, fonts, etc; they can resize the web browser window; and everything gets laid out again nicely by the web browser if and only if the HTML code describes the structure of the document rather than its exact desired appearance.

See http://www.dgp.toronto.edu/~ajr/104/diary/03/tags.html for a list of tags we've discussed, but note that not all of those are structural tags.

Have you ever had the experience of e-mailing someone, say, a Microsoft Word file and they print it and it comes out wrong? Maybe you formatted it to fit exactly on two pages on your computer, but they have different fonts or a different kind of printer and it is two pages plus one line on theirs. Or other fine-tuning you do of the appearance on your computer screen might go awry when some of the parameters of the formatting situation on your correspondent's computer are different. HTML doesn't (and can't) prevent this. What it can do is allow the web browser which is formatting the document for a specific computer to have the information it needs to make good layout decisions. That is, we say it's good that people's computers are different, that their web browsers let them set different preferences, and then the HTML describes the data's structure at a higher level than this so that it can all be made to work out. In writing a web page, I don't know what fonts and sizes you prefer and how big your web browser window is, and nothing can change that. What we can change is that we can provide a situation where I don't have to know, or make dubious assumptions.

As for creating these HTML files, as you know we recommend the editor "nedit" for use at CDF, and the editor "pico" for use in a terminal window (e.g. from home), but any text editor will do. You should make sure your HTML file looks appropriate with "cat" on a CDF linux machine, and you should make sure it looks right when formatted by opening it in a web browser. Note that you can open files in a web browser with the 'open' menu item in the web browser; you don't need to publish it on the web to be able to view it in a web browser. (Although you also need to publish it on the web to comply with assignment three part four.) You can also type "firefox filename".


Time permitting, there was also some material on data representation. This was mostly a recap of some lecture material, with more working through examples.