Tom Forsyth listed one of the reasons--when a storage is explicitly expressed. The other two that I can think of are for external linkage and non-constant expressions on the rhs. In the end, you really don't know without some introspection. It's up to the compiler. External linkage is one that will force storage, but in general const will require internal linkage and fold expressions accordingly (at least for primitives).
Edit: I would never expect anyone to actually remember or need to know that for an interview. It would be definitely something you would re-discover/discover because you had to run into it.
I would love to take feature requests, but it's honestly not maintained. Don't get me wrong, it wouldn't be hard to reintegrate it with the current release, but I believe the idea should be at least approximated with vimscripts rather than making someone build a custom vim release to support this feature alone. Plus, now :rnu sort-of mimics this functionality by showing the current line number on the 0th row, instead of '0'.
On a side note, to your second request of left-side/right-side splits for line numbers: my first experimentation was to do just that. Left side was rnu and right side was nu, but the way vim's buffers work out, it wasn't very practical. It also caused headaches for rtl text and so on if I remember correctly.
I was at one point very interested in the same idea. I wanted relative line numbers and absolute line numbers integrated together. So, I created a patch, called it `dnu` for "dual line numbering", and sent the suggestion to Bram Moolenaar.
However, in 7.3.787 (Jan 2013) a patch was introduced which showed the current line number in :rnu instead of 0. That's definitely in the direction that this patch was going in.
So, I didn't get a patch in, but on the plus side, I did learn a bunch about vim's source code.
Of course! There are multiple birds/sensors with SAR, IR, and multispectral capabilities always surveying various parts of the world. Organizations like NGA (formerly National Imagery and Mapping Agency to give some context) are very keen to using such geospatial data and provide resources in acquiring certain types of unclassified data sets:
I thought I would chime in on my own experience with related imagery. It's not uncommon to use a drone similiar to ARGUS that is capable of NIIRS 5-7 (see http://www.fas.org/irp/imint/niirs.htm) resolution for surveying. The guys at the ground station will find some region of interest and then use another bird/sensor--whether that is SAR, IR, multispectral, etc--with NIIRS 8-9 resolution and have it give a better look at the finer details of the area.
In terms of detecting objects: anybody with any sort of computer vision background will probably be scoffing at drawing rectangles around objects that are moving within a stable image. However, at a broader scope, change detection and broad area detection are heavily research fields and are always of interest to whatever 3-letter agency that deals with the imagery.
So, not only does ARGUS solve problems, it also provides data for creating new problems. Cool little piece of hardware.
Hm, I think an easy way to circumvent this style of captcha with a small amount of work would be to just send every iteration to Google as a reverse image search and pick the iteration with the most results.
- Non-distorted image: http://goo.gl/R3WnQ This is the "perfect" choice and also returns a good amount of search results.
- Slightly distorted image: http://goo.gl/G0xQN This results in a "human" choice on the picker and picks up a fair number of search-results.
I imagine 15-20 searches per captcha, but if you just pick the best per set you're probably going to end up with adequate results in circumventing the system.
"I don't know why more people don't just use tiling window managers. It's a bit confusing at first [...]"
That pretty much sums it up. I believe it's fair to say that casual users are (often) adverse to dramatical change--and switching from a point-click-and-move to a system that manages it all for you can be a bit crazy.
Where does PostGIS stand in comparison?