Thanks a lot, @enkei_design ! Both for encouraging words about the project and sharing your experience with larger-weight watch faces.
I did run the “history clock” watch face today for an hour, occasionally (more often than usual) checking the time. Didn’t notice drastic impact on battery. With your feedback, I will continue adding the “years” and let’s see, where it takes then. I will post an update here.
My logic tells me that the watch is so or so looking for changes at least every second, like “hey, I should look if there’s anything I should do, like sound an alarm, change the position of second hand, minute hand, or something else. … oh, I need to display another png, all right, done”. But as you could tell from this description, it’s a newbie logic. But let’s see.
I believe it’s fixed at 15 FPS by default and can’t be changed or increased. I think it was possible to set the FPS to 15, 10 or 5 back in GWS, but I couldn’t find this option in WFS now.
You can only slow the animation down by choosing to display individual frames multiple times, but I wouldn’t recommend this if you want the animation to be as smooth as possible.
For good results it’s probably best to create the animation with 15 FPS in mind from the start.
Thanks @enkei_design . We have to work within the limitations of all these platforms . I dropped some frames for my testing at the moment . Do you know about this ? I am using the UNIX timer because it is a thing of mine .
(((([UTC_TS])/4000)%1)*360)
Works fine on WFS but not on my watch I will sync it again . I am using it to time my Geneva Animation . and rotate the PIN DISC . I would not use this If I had not go it working elsewhere .
@OttS Watch displays are extraordinarily busy . Most refresh the whole screen every 1/30 sec and the movement of all the Hands an other stuff id calculated from the UNIX EPOCH COUNTER mentioned above . Which is a 11 digit number counting in milli seconds . I am sure the switching of your Images is not to much of problem . It will only switch on the one that meets the condition .
@enkei_design . I have just done the maths !!! What was you 1 minute Animation ???
I’m a complete noob when it comes to tags in WFS, so I really can’t help you there, sorry.
My animation was for full Earth rotation, if that’s what you’re asking. Exactly 900 frames, 25-35 KB each, and that’s heavily optimized and compressed.
Without optimization it was well over 100 MB, which would have been absolutely ridiculous for a watch face! 
Oh wow . I would love to see that . Sounds fantastic . I have a few Things with the Earth on one of the other Platforms . Would you be prepared to share it here as a Zipped .wfs or link us to the store where it is .
I don’t feel comfortable sharing a .wfs file of an already published paid watch face, but here is the Play Store link to it if you want to check it out:
https://play.google.com/store/apps/details?id=com.watchfacestudio.sol_3
You can see the animation in this reel:
https://www.instagram.com/reel/Ce_pTCtDquZ/?utm_source=ig_web_copy_link
Getting back on-topic, @OttS let us know how you progress with building/testing out the battery life etc.
It would be really interesting to know how much (if any) battery/performance drain there is. 
I think you are approaching this from a Web app view and not a Watch Face view,
Create your 1440 text boxes
Show or hide them by setting the opacity based on the hour Minute with tags.
write the tags out in a text box and
Once you create one hour for 60 minutes you only have to change the hour after that.
something like
0 start… + ([HOUR_0_23] == 16) * ([MIN]==23) ? 100: 0
It is only visible at 16:23
The other option with a text box is to move it in and out of view using the same type of tag
so X position is
125 Start +([HOUR_0_23] == 16) * ([MIN]==23) ? 0: 350
That way you can use text boxes as you said (once we have Word Wrap for text options)
anyway there are lots of ways to solve this,
Ron
No . Ha Ha . I understand you not wanting to post your .wfs here . Stunning work . I like your Cyber Cargo . I also like you Hud’s . Thank you, so much for showing me those.
Hi I do not think it would be much less when combined, but might be, PNG format is capable of compression and actually it should benefit from 2 neighboring areas of black or transparent pixels being described and stored as one larger area and save some bites (I could not prove this, when I stitched together the two examples below, the resulting file was actually slightly bigger than simple sum of sizes). But in general, your files seem little too big to me. I tried to remake them in 450x450 and ended with about half the kB size.
Thanks a ton, @Peter !
I’m not too enthusiastic about stitching the png’s and creating one large file. I prefer to keep them individual, so that it’d be easi(er) to change the text if needed. Also, I’m not familiar with the image stitching techniques (or any other stitching techniques, TBH), and tags in WFS.
But you gave me the idea to look if I can make my png’s still smaller in file size.
I’m working in GIMP. I converted the png into indexed colours, GIMP automatically suggested optimum palette of max 3 colours. And the resulting png is 2.97KB, instead of original 18.7KB. Yes, the indexed one is a bit pixelated, but I did a test run on the watch and only with full brightness and knowing what to look for very carefully did I notice some slight pixelization. But for all practical purposes that quality is fine. (I have attached both the original and indexed png’s here.)
So your idea gave me about 85% size reduction. Marvelous. Big thanks. A great forum we have here! 
@Peter, your image
is less pixelated than mine. How did you achieve that, if I may ask? Your file being about 8KB is still a huge win over my original 20KB, so maybe I should aim for less pixelization still …
Thanks, @r.liechty_SDR !
(I’m not sure what you mean by web app view. I’m using the Windows WFS desktop program …? My this question probably reveals how newbie I am, but, yeah, what can I do.)
I’m sure you’d figure out a lots of clever ways to achieve what I’m looking for. Unfortunately I’m a total newbie with WFS and do not know my way around the tags at all … 
I’m sure I will learn them in the process, but right now I’ll probably keep trying with individual png’s. Which is also stemming from your suggestions.
There’s another reason why I’m currently not looking into text boxes. The text in them doesn’t wrap, and you can’t make line breaks. When I tried with text boxes, in order to utilize the curved screen real estate to the fullest, I regularly had to create several boxes for each minute/display. Basically 1 box for each line, to maximally utilize the available screen width. And, BTW, I noticed that that (along with the conditional lines for every minute) made WFS (I mean the desktop program) slower and slower …
A little update. After your great suggestions I have discovered webp format, which gives files 1/3 the size of png’s. And that’s lossless. So currently that seems the way to go with my project.
Thanks a lot, everybody.
I guess you made it acceptable well, no need to make much around it, but I can describe my steps (I would not want to repeat it 1440 times). I put the text in excel cells and adjusted their sizes to be somewhere around yours (450px wide and font SF compact with size 42), then used windows snipping tool to cut them out, pasted in gimp, set layer transparency to the black colored background and exported it as grayscale PNG file. Maybe it can be done even smaller with different settings, I did not experiment. Good find btw., I was not aware of the WEBP format being supported, nor do I have any experience with it.
Try running the PNGs through TinyPNG and I think you will find it reduces the size a lot. I use this for a lot of theme PNGs to make sure my APK isn’t too large.
I can’t believe that. It probably caches them and starts to have to page memory especially if you have a Hard Drive and not a SATA ssd.
String formatting for line breaks and word wrapping, etc. is a very highly requested feature especially for some Complications like next event.
I’m learning a lot from this topic I’m really glad you are asking good questions.
Ron
Samsung Developer Relations
I have a SATA 256GB SSD, 8GB RAM Latitude 5490.
I’m not sure about page file and who or what is to blame, but I am very sure that when I added and added text boxes and conditional lines in WFS, at some point the whole program started to become noticeably slower and slower. Menus opening slower, etc. Again, may really be the fault of too little memory or whatever, but so it was. Really bad that I deleted those old .wfs files now that I switched to images’ solution, otherwise I could have uploaded some for you. I think I managed to create about 185+ conditional lines, each line covering in average 3 text boxes. Which may well be such a stupid scenario the program was never intended for.
Any predictions when the string formatting might actually happen … ? (But now that I’m doing it via images, pls do not rush it too much, otherwise I have to do first 1440 images and then 1440 strings … 
)
Just out of curiosity, what about adding just the text boxes? Could you isolate what is bigger problem, whether it was the conditions, or even layers count alone? Are there no problems with many images? What is your current layers count?
I have now switched to images’ approach and that’s a lot of work, so I’m not too enthusiastic about replicating the situation. As I read my earlier posts, I had about 550 conditional lines, so probably each text post was “governed” by its own cond. line. My muscle memory says the problem seemed to be the conditional lines, and this is also supported by the fact that I named my initial question/thread as I did.
No similar problems with images so far, so my best guess is it was somehow a combo of c/lines and t/boxes.
I almost think this would be easier to do with Android Studio and Jetpack. I think you have the programming knowledge to do it.
Still personally it has been quite interesting to try.
Ron
Samsung Developer Relations