There is a high vacuum inside a CRT. The atmospheric pressure pushing on the glass face is 14.7lbs/sq in. If the face of the CRT is curved, it is like an Roman Arch, the glass is very strong in compression. If the glass is made flat, then it has to be much thicker (and heavier) so that the atmospheric pressure doesn't break the glass.

I have an old Sony KBR 36" flat-glass CRT TV which weighs ~300lbs, most of that in the glass in the CRT.

Another reason why curved is better has to do with focusing the electron beam on the face of the tube. If the glass is curved, the beam stays in focus from center to edge of the viewing area, because the distance is the same from electron gun to center or edge. If it is flat, there must be position-dependent refocusing to adj for the varying path length.

Thanks... Good Answer!
My teacher didnt answer this!

Modern tubes (though it's hard to say that as CRT's are harder and harder to find now days) are just about as perfectly flat as you can get, internally they're mechanically supported and use thicker glass that is carefully shaped to provide a flat outer screen but a curved inner one with nearly the same support as a domed outer one, they compensate for the distortion in the electronics. I've used a few monitors where a horizontal high tension wire was visible on the screen.

Absolutely, and it's still more useful in some situations, such as analog o'scopes.

Anyone that uses an LCD on a regular basis should know what happens when you scale the bit perfect fonts that are native to it's resolution by any non pixel dividable ratio. Nasty aliasing effects. Analog monitors within their useable range can reproduce any resolution they want natively.

I'm horribly well aware of compression artifacts. Flicker effects from AC run LED or fluorescent lighting and aliasing effects on LCD's as well as having names for all the dead pixels on an LCD I use =\

if you want to see a real interesting way of getting all of the magnetic and electrostatic variables to match and keep the beams focused and converged, see if you can find an old 1960's vintage color TV. the wave shaping circuits that added correction voltages for the focus and convergence were very complex LCR networks, and there were adjustments for every area of the screen (and they also interacted so much, you usually had to do the whole alignment several times to get it right). one of the reasons these convergence networks were so complex, was that instead of the electron guns being arranged in a straight line, they (and the phosphor dots on the screen) were arranged in a triangle. Sony's introduction of the inline tube (Trinitron) with phosphor stripes rather than dots, allowed the convergence to be done primarily with magnetic rings on the yoke. i definitely prefer the magnetic ring convergence adjustments, and can do them in less than 1/5 the time it would take to do the old convergence panel adjustments. i haven't had to converge a tube in about 8 years or so, and the only reason i did them after about 1998 was that i had client with a bunch of (externally identical, but not identical internally) monitors, some with bad tubes, some with bad guts, and not all of the yokes were interchangeable.