Question on Atlas Tool & Its Status
Comments
-
BUMP!
Ben, just for kicks, I went out to the web version of Logos to see how the Atlas responded there. Much faster and smoother than when running it from within the desktop app version of Logos. I did encounter a few glitches, but don't want to get sidetracked with them now. The main point is this... regarding speed and smoothness, I'd have no issue if the Atlas performed from within the desktop Logos app the way it does from within the web app. Why so much difference???
OK Ben, following Sean's procedure and your instructions, here's what I came up with in my latest round with the Atlas.Hi Rick,
There's a bit of overlap on Atlas vs map loads, but I would classify them as such...
- Atlas load time = from when you click on the map link in Arnon's Factbook page to when the map titles show up in the left hand panel of Atlas
- Map load time = from when the map titles show up in the left hand panel to when the map is focused on Arnon
And in regards to the "shift + right click", I totally forgot to specify to hold the right click and move the mouse/pointer (whoops, sorry). The correct instruction is "shift + right click hold + mouse/pointer movement". This may be a bit awkward on a laptop or trackpad.
Hope that helps
On average, for the initial load it takes:
- Just 2 seconds for the Atlas to load from the Arnon Factbook page.
- Another 12-14 seconds for the map to load and the orange circles to flash on then disappear, and a majority of the map to be rendered.
- Another 6-8 seconds for the map to completely finish rendering. The map is usable before rendering is complete, but it is distracting since it's hard to know if it's only terrain that remains to be rendered or if anything of importance may still be added.
On average, once loaded it takes:
- 3-8 seconds to change from one map to another. (Most complete in 3-5 seconds but there are outliers.)
- 2-5 seconds to zoom in/out by 1 tap of the +/- keys. (Most complete in 2-3 seconds but there are outliers.)
- 2-4 seconds to pan.
- 2-3 seconds to rotate map "one tap" of the arrow key.
- 1-2 seconds to tilt map "one tap" of the arrow key.
- 2-3 seconds to return map to upright position
FWIW, these are better times than I've experienced in the past, though still not what I would hope for. Changes are very choppy and with each movement, leaves one unsure if the request is complete or if there's going to be more movement before the request is finished. Also, note that I am on the most recent release of Logos, though I don't know if that's a factor.
Let me know if you want any additional info.
Thanks!
0 -
<snip />
Ben, just for kicks, I went out to the web version of Logos to see how the Atlas responded there. Much faster and smoother than when running it from within the desktop app version of Logos. I did encounter a few glitches, but don't want to get sidetracked with them now. The main point is this... regarding speed and smoothness, I'd have no issue if the Atlas performed from within the desktop Logos app the way it does from within the web app. Why so much difference???
BUMP!
Rick: I've passed this thread along to our developers, but don't have any further insight to contribute at this point. The three of us internally who have reproduced your steps on desktop don't have loading times as slow as yours. There are numerous differences between the desktop and web code and contexts, so it's not easy to give an explanation without more information.
0 -
Hello Rick,
I was one of the developers that initially put together the Atlas tool. As far as the difference in performance between the web and desktop versions, the reason is because the Atlas tool (and a few others) is written natively for the web, and is running in an emulated environment on desktop. Because Atlas is graphical in nature, the emulation isn't quite as effective as some of the other tools running in the same environment. Performance improvements have been difficult to implement as well, as what may be an improvement in one environment, may be incompatible with or prove to worsen performance for another environment.
0 -
Thanks for the reply, Daniel. I am very disappointed in the answer, though. As with many community members, I've invested a lot in Logos over the years and find it pretty discouraging to learn that regarding the Atlas, I'll have to use the Web version of Logos in order to avoid the lag and slowness experienced in the desktop (the main?) version of Logos.Hello Rick,
I was one of the developers that initially put together the Atlas tool. As far as the difference in performance between the web and desktop versions, the reason is because the Atlas tool (and a few others) is written natively for the web, and is running in an emulated environment on desktop. Because Atlas is graphical in nature, the emulation isn't quite as effective as some of the other tools running in the same environment. Performance improvements have been difficult to implement as well, as what may be an improvement in one environment, may be incompatible with or prove to worsen performance for another environment.
0 -
Well, thank you very much, Mr Ausdahl for your bird-dog tenacity.
Who would have known that a bet a few years back would pay off! Emulation.
"If myth is ideology in narrative form, then scholarship is myth with footnotes." B. Lincolm 1999.
0 -
Thanks for the reply, Daniel. I am very disappointed in the answer, though. As with many community members, I've invested a lot in Logos over the years and find it pretty discouraging to learn that regarding the Atlas, I'll have to use the Web version of Logos in order to avoid the lag and slowness experienced in the desktop (the main?) version of Logos.Hello Rick,
I was one of the developers that initially put together the Atlas tool. As far as the difference in performance between the web and desktop versions, the reason is because the Atlas tool (and a few others) is written natively for the web, and is running in an emulated environment on desktop. Because Atlas is graphical in nature, the emulation isn't quite as effective as some of the other tools running in the same environment. Performance improvements have been difficult to implement as well, as what may be an improvement in one environment, may be incompatible with or prove to worsen performance for another environment.
Rick, that's not to say we won't try to improve performance on desktop. But have to prioritize and filter that development task along with all the others that our users request.
0 -
Thank you, Sean. I know everyone there works hard and does the best they can with the time, budget, and priority restraints placed on them. Having been in IT (both development and support) I appreciate the attention to detail (every small detail) required throughout each phase of a development project, from the feasibility study to final client usability testing. You can get 99.999% of the project completed on time and working as designed, but it's the 0.001% with problems that get all the attention.
Thanks for the reply, Daniel. I am very disappointed in the answer, though. As with many community members, I've invested a lot in Logos over the years and find it pretty discouraging to learn that regarding the Atlas, I'll have to use the Web version of Logos in order to avoid the lag and slowness experienced in the desktop (the main?) version of Logos.Hello Rick,
I was one of the developers that initially put together the Atlas tool. As far as the difference in performance between the web and desktop versions, the reason is because the Atlas tool (and a few others) is written natively for the web, and is running in an emulated environment on desktop. Because Atlas is graphical in nature, the emulation isn't quite as effective as some of the other tools running in the same environment. Performance improvements have been difficult to implement as well, as what may be an improvement in one environment, may be incompatible with or prove to worsen performance for another environment.
Rick, that's not to say we won't try to improve performance on desktop. But have to prioritize and filter that development task along with all the others that our users request.
I appreciate everything you all do. [:)]
0
