As I work on the docs, the way Avogadro’s interface uses the word “molecule” increasingly bothers me. It is used essentially to refer to two things so far as I can tell:
The entire set of atoms and bonds visible in the view pane
An entry in the Molecules pane, which is essentially a list of the open files
First of all, there is a slight discrepancy between those two meanings as a file can contain multiple geometries (shown via the Animation Tool) and also contains other associated properties and data such as charge, orbitals, vibrations, energies.
A bigger issue than that though to my mind is that it doesn’t align with what a user thinks of as a “molecule” i.e. a discrete group of (usually covalently) bonded atoms.
There are many situations where the chemical and Avogadro definitions wouldn’t match e.g.:
An ensemble of water molecules
A simulation of a reaction between two molecules
A protein + something bound to its binding site
A crystal
A gas
A glass or other amorphous structure
In these cases referring to the overall collection as a single Molecule is to my mind very confusing.
As examples of where in the UI this could lead to confusion:
The Manipulation Tool has “center of molecule” as an option for the centre of rotation. If the user wants to rotate a methyl group in a methyl ester, say, and there are also a few solvating water molecules in the file, they would be surprised if that centre is determined globally as opposed to the centre of the covalently-bonded ester of which the current selection is a part
It would be somewhat ridiculous if the calculation of the “center of molecule” included the positions of the atoms in other frames of an animation, but that would be consistent with definition (2)
Saving geometries to non-CJSON formats is done via Export Molecule, but users might not realise that this is the appropriate command as they may think that it would only export a subset of the atoms
For the Selection Tool, it is tempting to say that double-left-click selects an entire molecule, but this would be inconsistent with Avogadro’s meaning of molecule. Hence in the docs it is necessary to talk around this and refer to selecting “all connected atoms” where connected atoms are any that are linked to the clicked atom over any number of bonds
Do others agree that this is not ideal?
If there is broad agreement that it’s an issue then we can consider other options for Avogadro to use – first things that come to my mind would be “geometry” or simply “file”.
The “Molecules” pane could certainly read “Files” or something similar.
But I think you’re reading far too much into the other aspects. Yes, there are plenty of applications in which Avogadro as a supramolecular or materials system (e.g., binding site of a molecule and a surface, generating a nanoparticle, etc.)
I’m not opposed to finding new wording, but a lot of programs refer to “this current chemical system” in a simulation as a “molecule.”
I’ve found that many, many people only work on one molecular fragment at a time
If you talk about “chemical system” or something like that, people get confused
So yes, generally in the Avo1 documentation, “molecule” refers to “this current set of atoms and bonds, potentially multiple fragments” and “fragment” or “connected component” to refer to “this set of things that are all connected.”
It gets pretty complicated in the broader chemical community. “Sodium acetate” is considered the molecule even though the sodium ion is not really connected to the acetate ion.
I don’t think I’ve seen anyone assume that the “center of molecule” somehow includes the other frames of an animation.
Similarly, I picked “Export Molecule” to have a clear distinction with “Export Graphics” - I haven’t seen anyone confused that it would only export some subset.
I’m happy to be convinced otherwise, but to me it seems as if you’re overthinking the wording here. I’ve done a lot of user testing with new students joining my group, including lots of undergrads.
Renaming the “Molecule” pane might be okay, but at the moment, it’s listing things by name / formula more than filename.
Perhaps I made it sound like I think this is a huge problem and that’s not the case, I just think it’s beneficial to have terminology that’s both consistent and intuitive as it makes it easier for new users to work out how to do things.
If you think I’m overthinking it then maybe wholesale changes or strict definitions aren’t necessary. Maybe, like in the docs, the best strategy is just to avoid or clarify the term in places where it might be ambiguous.
I guess of the situations I raised it is that of referring to files as molecules, and specifically in the Molecules Pane, that I find the most problematic. Back when I was a new user myself I didn’t realise for a little while that they are equivalent in that context. That the plus icon does the same as the New command wasn’t obvious to me.
Showing the formula is often useful, to be fair, the name maybe less so since it’s often not generated. But I regularly find it hard to distinguish between different files of the same molecule and have to find some way to remember which is which e.g. by the order I opened them in, which the filename would really help with. What might be really nice is to show both the filename and the current information on two separate lines.
With regards to the “center of molecule” thing, how would “center of all atoms” sound as a way to just avoid the question entirely? I think that’d be clear, and that way we’re not having to replace “molecule” with some alternative term like “chemical system”.
I think the model is able to handle multi-line text strings. That said, it would probably be nicer to adopt the HTML formula code from elsewhere so it gets things like \ce{H3O+} or \ce{C6H6^{•}}