Only 15 posts behind. gotta get busy. Oh! I know! More coffee! Javajavajavajavajavajavajava…… 
What was the name of that special Estate coffee you were drinking? Do you still have some of that? 
Exactly. The Q8 has been significantly better than earlier projects.
For example, the BLF-A6 is still a best seller to this day, but the changes between the final prototypes and first production models were massive. They shipped it with a completely different host than the prototypes, much lower quality, with randomly-pitched threads so it couldn’t even lego with other A6es. Different parts had different diameters or didn’t fit right, some of the black anodizing was blue, functional components like springs were changed, lots of people received the wrong tint, the driver has been changed repeatedly since release, the clips couldn’t clip without breaking, etc.
… and those were small issues compared to the Cometa.
It seems Thorfire has been a much better partner.
Correct LEDs in proper tint on all.
Correct parts on driver including “raptor claw” 7135 chip
More then the aimed for 4000 lumenion all.
Fat copper DTP led board , something literally called “a dream that will not be possible” early on.
All but 2 with issues had an operational Q8 doing out the front what it was supposed to do.
One DOA, after Tom checked it it was the combo long cell and dry o ring and thus with a little lube and/or different cells was working
It is not clear how many have arrived yet but with 1800 pre orders it seems safe to say 1000.
With a rounded up troubled number of 30.
All worked
Only 2 drivers needed replacing, that really hindered use, but even here there was light yet no proper functional UI.
All getting a solution for issues and direct help getting there. (OK we need somebody from Orange County in a few weeks to make this true)@
0.2% not doing what a flashlight with advanced software is supposed to do.
Judging by what we can read about A brands I kind of think this is great work by Thorfire.
No way we could have asked for more.
Correct LEDs in proper tint on all.
Correct parts on driver including “raptor claw” 7135 chip
More then the aimed for 4000 lumenion all.
Fat copper DTP led board , something literally called “a dream that will not be possible” early on.
All but 2 with issues had an operational Q8 doing out the front what it was supposed to do.
One DOA, after Tom checked it it was the combo long cell and dry o ring and thus with a little lube and/or different cells was working
It is not clear how many have arrived yet but with 1800 pre orders it seems safe to say 1000.
With a rounded up troubled number of 30.
All worked
Only 2 drivers needed replacing, that really hindered use, but even here there was light yet no proper functional UI.
All getting a solution for issues and direct help getting there. (OK we need somebody from Orange County in a few weeks to make this true)@
0.2% not doing what a flashlight with advanced software is supposed to do.
Judging by what we can read about A brands I kind of think this is great work by Thorfire.
No way we could have asked for more.
It follows the tube - batch1 tube w/batch 2 head feels good - batch 1, batch 2 tubw w/batch 1 head feels like batch 2.
I measured what is relevant: the outer diameter where the o-ring lies when the o-ring is removed, and the batch 2's are wider. All makes sense - o-rings are the same size. Found one of the GITD 42x1.5mm o-rings feels easier to tighten than using the stock o-ring. But could be just this one, or all the ones from fancyflashlight.com. The KD GITD 42x1.5mm I tried still feels tight.
I ordered the FastTech 42x1.0mm ones. Couple weeks I'll get them, they shipped already.
The Miller - +1, couldn't agree more. This is pretty much little stuff for the most part, not to say 99.9% perfect though, but wayy better than other efforts at this. I'm anal in trying to find causes of issues though 
.
So i checked guys and the tube ridge where the o ring sits has a smaller diameter on the first batch. That explained a lot since swapping o rings didnt help
Haha just too late I see
I think we need more people to measure their lights around the area near the o-ring, both inside the groove and just outside of it. Especially need others who have lights from both batches to compare. Also, would be good if more people were able to test with the long cells. Maybe even try with a short cell + spacer to make the length if a long cell isn’t available. This way, maybe we can figure out what the actual issue is here. I only have short cells (but could make a spacer I guess), and only one light, from the second batch. So, I could do a little, but not much in the way of testing and comparing.
Rayman007, your top photo is a close match to my batch 1 o-ring groove diameter @ 47.56mm.
Grrr! Why the heck did they make so many changes to the battery tube which were not necessary or asked for?
It can’t be due to using different machines as the software program is the same. They had to have changed the software. I don’t understand it.
I also checked the groove of the tube on the tailcap side and on the first batch it is 47.5 and on the 2nd it is 47.84
So it can be solved by using 1mm o rings but they are much less common
Both tubes are 78.61mm Long by the way
I took measurements, though not 100% confident and didn't write them down in my notebook or take pics. But yes - batch 2 are ~0.2-0.3 mm wider, figure on, so no doubt to me it's this difference resulting in why it's harder to thread on the battery tube, and when you compound it with making it harder fighting spring compression with long cells, it gets much harder to thread on.
No doubt, plain and simple, direct cause/result to me. Small tolerances make big differences and this area of the battery tube was changed.
Toss it on the lathe and remove a little from the groove. Maybe even use a thin file. Maybe a dremel with cutting disc? That should fix it.
But then the ano is gone
Sigh will try to make some pics and look at proto, V1 and V2 closer.
Long cells / spring compression could be helped by putting spacers (washers) under the tail pcb, there seems to be room under the tailcap.
The o-ring covers most of it. Besides, you only see it when the head is removed.
Okay, if your sensitive to it, just add a little paint to the sides. Let the o-ring cover the bottom of the groove. Now it looks perfect. 
Yeah, that might work. If there is a possibility of putting a single ring of copper (wire?) between the pcb and batt tube, that might be even better, to keep as much contact as possible and reduce or eliminate adding resistance.
A 1.35 mm thick o-ring might be about perfect. Think'n the 1.0 mm I ordered will be too thin and will do nothing for sealing.