The presentation of this text with aids was done using open-source software called Ransom. The software can generate both screen-reading output like this and printer-friendly output.
Each page or section of the text has its own web page. The web page is split into panes, which are meant to be visualized spatially as if they were side by side. The idea is similar to the way the Loeb Classical Library books present Greek or Latin text on the left and English on the right facing page. However, the number of pages "side by side" is three rather than two. You can access the different panes by clicking on the triangles.
The main pane, which is the one you landed on, shows the foreign-langage text. In that pane, if you hover the mouse over a word, an interlinear gloss will pop up after a one-second delay. If you click on the word, a more detailed gloss pops up instead. Double clicking brings up a full-length dictionary entry from LSJ.
There is also a separate vocabulary pane in which lemmas are listed in alphabetical order. Only uncommon words are listed. If you like, you can study the vocabulary before reading the text.
The description above is for laptop and desktop computers, which I think of as the norm for reading long texts. If you're using a device such as a tablet, and your screen is below a certain size (currently set to a width of 1024 pixels), the page will be formatted a little differently in order to fit a decent amount of text on the screen. This is accomplished partly by making the line spacing smaller, which has the side-effect that a gloss will at least partially obscure the following line of text. I hope that the set of tweaks I chose makes this presentation fairly usable for people on tablets. You can also make the font size bigger or smaller, according to your preference, using your browser's normal controls for that. On touch-screen devices, there is no such thing as hovering the cursor over a word, but you can still click or double click to get the longer glosses. I doubt that this system will be at all usable on a smartphone, and I don't think there is a solution for that.
The Greek text of the Anabasis is based on the edition by Marchant, 1904. The English translation is by Dakyns, 1901.
Text I've written, such as notes and glosses, is CC-BY-SA 4.0. Glosses taken verbatim from Cunliffe or White and Morgan are in the public domaion, as are the LSJ dictionary entries. Glosses taken verbatim from Wiktionary are CC-BY-SA 3.0. US law does not allow copyrighting of grammatical facts such as part-of-speech tagging and does not give any additional protection to databases beyond the normal protection of copyright.