[toddos]

Originally Posted by Sabardeyn

True. But I figured if it was implemented it would be user customizable to offer a 1 page = "N x N" customized grid size. So... tiny to huge screens, and relative database size, could be accommodated to some degree.

Let's say you have a reasonably-sized library of 500 books. Let's also say you have a 1920x1080 screen and you run calibre full screen. Finally, let's say you display covers at 128x128, because any smaller and they don't convey much information (also assume that covers are perfect squares, which they generally aren't). For usability, let's assume that there's a 20px gap between each cover. So, each cover requires 138x138px to display. Assuming no other chrome and complete edge-to-edge layout of books, you can fit 13 covers horizontally and 7 covers vertically. 13x7 = 91 covers == 6 pages of covers for your library. And that's best-case scenario. If you're a more average user, with a 1366x768 display, with sidebars and other UI chrome enabled in calibre giving a ~800x600 display area. You're now down to a 5x4 grid. 5x4 = 20 covers = 25 pages in cover view. Maybe that sounds fine to you, but that definitely does not work well for me.

Originally Posted by Sabardeyn

I think the failure to come up with a such a better, simpler interface is because no one has thought up an appropriate design for an interface widget offering all the possible ways in which features can be accessed and/or mixed together currently. I honestly think that if anyone could come up with a new and viable interface idea that it would at least be discussed and explored.

People have tried. Multiple times. Well, "tried", anyway -- basically this exact type of discussion, just theoreticals and "Somebody should do something!" but the people who want a simpler/prettier/better/<choose your adjective> GUI never come through with anything. Kovid and crew aren't opposed to making the GUI better/easier/cleaner/whatever (they are opposed to making one-off native GUIs per platform, and for good reason), but apparently nobody has provided any patches that make things sufficiently better all around (what's good for you may not be good for someone else -- that's why the GUI is configurable) for those patches to make it into the product itself.

Originally Posted by Sabardeyn

Another, somewhat related issue, is the tech level of the user manual. Almost all good programmers are terrible writer's/instructors; they're simply too knowledgeable within their field (in general) and too close to the software (specifically) to effectively write instructions for the average user. Most users will never understand the hardware or software processes and limitations; those few that have some knowledge will have an oversimplified understanding of things. Once you get past that, you still have linguistic barriers - no two people will read and comprehend a page of text and arrive at the same understanding.

So do something about it? The documentation has what the documentation needs -- specifics about how stuff works. What you're asking for is more of a "user guide" than documentation, and there's no reason you or somebody else couldn't write a simple guide saying, "Here's how you add a book. Here's how you send it to a device." It doesn't even have to be official documentation. Write a blog post and/or record a video and post it in various places. If it's good, people will find it, Google will bubble it up to the top of the search results for calibre, etc.

Originally Posted by Sabardeyn

I also agree with your supposition (without any factual proof available to either of us) that the abandonment rate of calibre is probably quite high. Additionally, I suspect that there are a high number of "repeat new installers" within the download numbers cited. Having uninstalled calibre, I imagine the unique ID from that installation also ceases to exist (rightly so - tracking should not be necessary). So repeating the install process only shows up as a new user rather than a repeat installation - thus skewing the numbers.

According to the stats page (emphasis added by me):

Usage statistics are collected whenever a user starts calibre. Every calibre installation has a unique ID, this ID remains unchanged by upgrades and even an uninstall/re-install. This ID is used to collect usage statistics. Only this ID is stored, no other identifying information is collected.

My guess without reading through the code (which you can absolutely do) is that this is probably based off of the PC's MAC address. If you install across multiple PCs, that would look like different installations. But if you install/uninstall/reinstall on the same PC then you're not duplicating your original installation. I don't know if this is recorded only at installation time, or if calibre phones home periodically/on every start. If the latter, then there's probably a database somewhere that could be mined for "active users" (defined as "users who have started the app within some timeframe"). You'd have to take that up with Kovid, though.