The Indonesian F30T8 ended it's life test. The bulb was not EOL but had lost most of the brightness, loss of a mercury?  The Hungarian lamp in the test continues on.  I need to find another tube to put in the place of it. Right now for light, there is a F30T12  Reveal.  The Hungarian lamp shows no more aging. The GE indonesian had end banding on both ends, one end very pronounced with a brown spot that was going to make that raindrop shaped black spot, but had not yet.  Well over a year in the test.

I need to light this and get it warm and take a picture of the most aged end.

What ballast have you used?
On an RS the lamp with failed cathode would reduce it's power input and redirect most of that power to heat up the failed cathode, so there will be not much power left to actually generate the light in the anode column...

I assume preheat, since rapid start ballasts for the F30T8 don't exist.

Powell, at what schedule are you running the lamps? 24/7?

24/7.  GE F30T8s and Philips will light FINE on RS.  Sylvania and Duro Test will not.

I once read here that F30T8 lamps have different cathodes from F30T12 lamps, and running them on a rapid start ballast will lead to shortened life of either the lamp or the ballast, I forget which. (Maybe both!) Interesting nonetheless.

@TheMaritimeMan: Haven't you mixed that up with the F32T8 vs F40T12 case?

Nope. The F30T8 was developed for preheat use, whereas the F30T12 was developed for rapid start use, and they are only interchangeable on preheat systems. I've never seen a rapid start F30T8 ballast, and I've read here that F30T8 lamps and rapid start F30T12 ballasts aren't compatible. Obviously, as Powell has shown, there are some exceptions to that, but I have read here that in general, running an F30T8 on a rapid start F30T12 ballast will lead to harm on one or both components, due to the F30T8's vastly different preheat cathodes.

Then I would expect the T8 would be worse at starting: Thinner tube uses to have higher ignition voltage. And not ignited lamp tend to overheat a HPF ballast...

It's interesting that the GE ( both Indonesian and Hungarian) F30T8 light JUST as a T12.  My F30T8N  Philips light even faster. Syvlania will not light when new, but an older one very aged, lit right away. I used a a pair wires attacched to one end of a DuroTest F30T8  Optima 50 and used a 12 V battery for extra heating momentarily....so  I guess that would be preheat / rapid start!   

@Medved, you lost me on the failed cathode on RS ballast. How does it take away the lamp power? I've never observed that in the hundreds of experiments. Also, just wanted to add that I've never seen a HPF ballast overheat because of unlit lamp...and I've experimented with every kind I can think of.

When taking the HPF RS ballast design, you have a step up autotransformer and the lamp connected to the higher voltage via a ballasting capacitor. The capacitor loads the secondary by a leading phase reactive power. This apparent power there is the OCV times the arc current, so about 80VA. When recalculating the reactive component, it becomes about 70VAr. To not require the primary to handle this reactive power, the transformer incorporate an air gap, so the main inductance of the secondary is lowered so, it compensate the capacitive reactance of the load (the 70VAr), so the primary is then loaded just by the real power required to feed the lamp and cover the losses, so with 30W lamp about 36W, so 0.25A mains current, splitting by part flowing to the primary and the rest to the secondary winding.
But when the arc in the lamp is not present, the capacitive reactance load is missing. But the inductive reactance is still there. And as it is not compensated out, the complete 70VAr, so about 0.56A current flows then into the primary. Because that is more than twice the primary is supposed to carry, it will dissipate more than 4x the power loss than normally. The secondary does not dissipate any heat, so the total power losses in the ballast are not 4x higher, but still about 50% higher than usually. That mean 50% higher temperature difference. When the ballast is designed to operate at 120degC core temperature with 40degC in the fixture, with failed lamp it would be about 150degC, when assuming 30degC in the fixture (the lamp does not dissipate there).

Interesting info. I don't have a 30w HPF ballast at hand right now to test but in the past, the only time I've ever seen the primary current jump with no lamp was an HPF preheat F30T8 ballast. The reason for that was that the capacitor was on the mains side. An EOL lamp would still rectify, hence no DC blocking, yet it was high power factor.

Meanwhile, I just tested a 2x40w HPF ballast with cap on secondary.

Normal line current with lamps (as measured): .77a, P.F. 96%
Line current without lamps: .32a, P.F. 15%

All HPF ballasts I've run across seem to only "idle" without lamps and are barely warm to the touch.

I just tested my single-lamp HPF 30/40 ballast:

With lamp, 0.42A, 0.93 PF.
Without lamp, 0.44A, 0.17 PF.

So perhaps it's a different story for single-lamp ballasts?

Is it your Sylvania MB1x40 with the 255voc? BTW, have you actually measured the ocv to see if it matches the label?