So what's in 6.4.112?
-
I too think that Modelhead is on to something. There is this thing that SU is very processor-intensive, and using only the single one, that makes it vulnerable to slowdowns caused by other processes. My home computers are very neat his "weenie" one. They are about 3 and 4 years old, a P IV 2.8 GHz, Non HT, luggable , 1.5Gb memorywith a Nvidia laptop card, and a P IV 2.66 HT small-form factor PC, 2Gb memory with an ATI All-In-Wonder card with ónly 64 Mb of memory.
Both these run SU quite well, and the difference to my work machine (3.6 GHz PIV HT, 4 Gb memory, Nvidia Quadro 1400 card) is surprisingly small. SU 5 loads faster than SU 6, but in use both feel about as responsive, and both are significantly faster, especially with large models, than SU 4. My models are mainly quite small, though.
I run the Norton security applications, they seem to be quite resource-hungry, but apart from that they never have given me any trouble. At work they have some corporate system.
I haven't upgraded my video drivers for years, sticking to the tenet that "if it works, don't touch it". The manufacturers seem to regularly release buggy drivers.
The users of 3D software seem all to be in some kind of technological cul-de-sac. None that I know support multiprocessors except for rendering, and manufacturers have stopped developing faster singlecores.
Anssi
-
Craig...I don't know how to be clearer than I already have in the previous posts here, but here's a copy of my response to Brad's inquiry that might shed more light(?):
My problems had nothing to do with dwg files. My concern is that the interface and speed problems that made SU6 not my primary program got worse with the update.
I have been using SU6 lately only to export the jpegs I'm using for renders in PSP, as the pencil sketchy edge is a really nice addition. I've even been setting up my pages in SU5 because the "beachball selection effect", as someone aptly called it, caused too many jumps into hyperspace when using the scroll wheel and/or the orbit tool.
Here is one more example besides the "graphic card"(?) boggles I described in my post: With clean ram (first action after opening this file) going to xray mode takes 3-4 seconds in SU5, took 5-6 seconds in SU6 before the update, and takes a full 10 seconds with the latest state of the program. (This backward progression makes no sense to me BTW...and in short does indeed SUCK!...thanks for reminding me in the other thread.)
If the hardware requirements have changed, please give me more specific information than the "carefull not to mention brand names" clues on the Google SU page...I'm getting ready to upgrade my computer and really would like SU6 to function properly (or as near to SU5 as possible :`)
-
@unknownuser said:
Do I hear a 'I second that Coen'?
i second that, coen.
although i sympathize with all of you who are experiencing problems, i prefer to see the little fuckups for what they really are.
and this is because i prefer backing the SU developing team up in the hope that they will come up with an SU7 that is better than SU6 THAN to go looking for a new application to do the tasks i do with SU.
moving on to some new app means leaving all that surrounds SU behind. and this i am not prepared to do, because no other app has developed the following and unofficial support that SU enjoys.
-
Hi Tom, I think you have been clear. It certainly does sound bad, and it is very strange. We've made no change that would affect performance (any change that is optimized or affected by new hardware is a BIG change, and this M4 update is one of the smallest of the updates).
What I was hoping for, is to learn something new from any system house cleaning that you do. To see if there was something on your system that is causing the graphics card to cycle extra long, or cause any feedback latency and to see you fix that (and learn what the cause was) would be the best outcome here.
It really doesn't make any sense for you to have slow-downs, especially more slow-downs with M4, so we're perplexed.
Please keep us in the loop.
Thanks,
- CraigD
-
Well all of us running OS X 10.4.11 can upload to the 3d-warehouse again. That's a much appreciated fix. The new look of the 3d warehouse is nice too!
I love this program. It has become a big part of my life. I suspect most of the people on this forum feel the same way. Reading through the "Anti-SketchUp! Snobbery" thread would certainly seem to indicate this is the case.
Google SketchUp folks, I hope you just keep focused on making a truly great program even better!
Thanks,
Fred
-
Craig, I suppose if some cleanup idea Bruce suggests makes SU5 faster it may also make SU6 faster factored by the same amount it is now slower...stranger things have happened in this world of ones and zeros.
-
A great wealth of computer maintenance articles and other factoids are over at http://aumha.org/articles.htm. All his resources/forums/etc are still in the process of being updated for Vista. One place to start is Computer Health http://aumha.org/a/health.htm . On that same page is a link to Parasite Fight - yet another cause of mysterious problems. Some of the utilities for that fight will also help unclutter a system.
And I encounter some computer repair techs who dislike Norton products, saying they can needlessly junk up a system and just be another cause of problems.
-
@fbartels said:
Well all of us running OS X 10.4.11 can upload to the 3d-warehouse again. That's a much appreciated fix.
Unfortunately the same doesn't appear to be the case for my OSX 10.5.1 iMac. I just tried uploading a simple model with a strange partial result.
a) the warehouse window was very slow to appear and took about a minute before I could type into it.
b) actually uploading produced a progress bar that just displayed the blue barber-pole pattern and sat there. I left it for several minutes but nothing changed. I was able to move the model around though.
c) on checking the warehouse via Safari I noticed that my model seems to be partially there. Instead of the downloading menu showing 'sketchup 6 .skp' it shows 'Collada (.zip)' which means nothing to me. What is Collada? apart from something you put a pina in?
d) attempting to download this 'collada' doesn't produce anything I can make use of.The model is http://sketchup.google.com/3dwarehouse/details?mid=355c41f107a8ff2cc274799cf1faeca7 in case that helps.
The version of this new SU6Pro shows as 6.4.120, my machine is a new iMac 2.4GHz core2duo with OS X 10.5.1 fully updated, 2Gb ram, etc. I've also noticed some odd hesitations in Pan; previous version did not hesitate if I pressed shift and mouse-middle-button to Pan, this one hesitates for about a second before moving the view. Oh, it also happens if I just use the Pan tool.
-
Hi Tim,
Collada is a 3D interchange format that can be opened in Google Earth (.dae file format): http://en.wikipedia.org/wiki/Collada
You should see your model as downloadable in .skp, .kmz, and .dae (Collada) formats. I don't think the upload went through completely. Please try again.
The SketchUp version is correct at 6.4.120.
I'm not sure why you're seeing any hesitations....do you have Fast Feedback and Hardware Acceleration on? Check your Open GL preferences in SketchUp: SketchUp > Preferences > Open GL
Let me know if this helps!
Thanks,
- CraigD
-
@craigd said:
Collada is a 3D interchange format that can be opened in Google Earth (.dae file format): http://en.wikipedia.org/wiki/Collada
OK...
@craigd said:
You should see your model as downloadable in .skp, .kmz, and .dae (Collada) formats. I don't think the upload went through completely. Please try again.
No improvement I'm afraid. I do have console log showing a lot of lines like this -
Dec 18 18:24:36 tim-rowledges-imac [0x0-0x63063].com.google.sketchuppro[2830]: SketchUp(2830,0xa0159f60) malloc: *** error for object 0xa039b6d8: Non-aligned pointer being freed
Dec 18 18:24:36 tim-rowledges-imac [0x0-0x63063].com.google.sketchuppro[2830]: *** set a breakpoint in malloc_error_break to debugand later ...
Dec 18 18:25:44 tim-rowledges-imac SketchUp[2830]: shareWithGoogleEarthCommunity
Dec 18 18:25:48 tim-rowledges-imac [0x0-0x63063].com.google.sketchuppro[2830]: adding: doc.kml (deflated 60%)
Dec 18 18:25:48 tim-rowledges-imac [0x0-0x63063].com.google.sketchuppro[2830]: adding: models/
Dec 18 18:25:48 tim-rowledges-imac [0x0-0x63063].com.google.sketchuppro[2830]: (stored 0%)
Dec 18 18:25:48 tim-rowledges-imac [0x0-0x63063].com.google.sketchuppro[2830]: adding: models/skpfile.dae
Dec 18 18:25:48 tim-rowledges-imac [0x0-0x63063].com.google.sketchuppro[2830]: (deflated 82%)
Dec 18 18:25:48 tim-rowledges-imac [0x0-0x63063].com.google.sketchuppro[2830]: adding: images/
Decand the last few that appear are -
Dec 18 18:27:36 tim-rowledges-imac [0x0-0x63063].com.google.sketchuppro[2830]: adding: textures.txt
Dec 18 18:27:36 tim-rowledges-imac [0x0-0x63063].com.google.sketchuppro[2830]: (deflated 39%)
Dec 18 18:27:36 tim-rowledges-imac SketchUp[2830]: KMZ21 export success
Dec 18 18:27:36 tim-rowledges-imac [0x0-0x63063].com.google.sketchuppro[2830]: adding: skpfile.kmz21
Dec 18 18:27:36 tim-rowledges-imac [0x0-0x63063].com.google.sketchuppro[2830]: (deflated 1%)- I guess the lack of something saying 'export success' after that might be an indication of some problem
And again, only the 'collada' version is on the warehouse list.
@craigd said:
The SketchUp version is correct at 6.4.120.
I'm not sure why you're seeing any hesitations....do you have Fast Feedback and Hardware Acceleration on? Check your Open GL preferences in SketchUp: SketchUp > Preferences > Open GL
Yup, both are on. One improvement of the iMac over my MacBook is that fast feedback actually works.
- I guess the lack of something saying 'export success' after that might be an indication of some problem
-
I must admit I'm disappointed that the "shadow settings saved in scenes get reset between sessions" bug still hasn't been fixed. This drives me nuts every time I reopen an SU file having to go to every page, adjust the shadow settings up by 1 hr, then down by 1hr (to fix the bug), then update the page.... KNOWING that as soon as I close that session I'll have to do the same again.... and no it's not machine dependant. I'm surprised everyone hasn't been kicking up about this.
Apart from that, it's good to see 2007 dwgs supported now.
-
@unknownuser said:
@tim said:
No improvement I'm afraid. I do have console log showing a lot of lines like this -
Hi Tim,
I forwarded your log on to the Developers to see if it gives them some clues. Thanks for providing it.
Are you seeing the same problem on every model you try to upload to the warehouse? What if you take a simple box, paint it with a few textures and upload it?
Thanks in advance,
Tricia
-
So still hoping for an answer from the "@last" crew to my questions:
What specific hardware is now recommended as "sweet" for SU6?
Why the button response slowdown from SU5 to SU6 if it's not hardware related?
Brad, Craig, Tricia...you out there still?
-
Like you Tom, I only use SU5. I installed the new version of 6 and gave it a ride and still think 5 is as least as fast or faster.
Edit: Yesterday, I converted a v5 model to a v6 and today discovered the file size had doubled. I purged and purged and finally got it down to the proper weight. This is why the model seemed so sluggish. So now, I have to say that V6 is probably faster than V5.
-
Ron, didn't work...file size went down 50k (nothing to purge) and the wait time to go to xray went from 10sec to 14sec. I'm disgusted max now.
-
Hello everyone,
Jackson - regarding the "shadow settings saved in scenes get reset between sessions". What location do you have set on the Model Info > Location tab? I'm looking at a model right now that shows the issue trying to figure this one out. The model has a location of Scharendijke, Netherlands. In this model, if I look at the sunrise and sunset times, they are wrong. Do you see the same thing in your model? And I'm experiencing the same symptoms as it seems you are experiencing... I can reset the shadow fixing the sunrise/sunset times for the current session, but after saving the model and reopening it, the times/shadows are incorrect again. What is frustrating though is that I can't recreate the issue from scratch to figure out how those sunrise/sunset times got messed up...
Tom - does the slowdown you are seeing occur on all of your models?
Thanks,
Tricia
-
Tricia,
Thanks for replying- almost all my model locations are set to Edinburgh, Scotland (UK). I've just checked and three models which are displaying this bug have North Angles of 190°, 0° and 25. I'm pretty sure all my skp files display this bug, but not all of them have scenes saved with shadows. I'd be happy to email them to you.
Regards,
-
@jackson said:
Tricia,
Thanks for replying- almost all my model locations are set to Edinburgh, Scotland (UK). I've just checked and three models which are displaying this bug have North Angles of 190°, 0° and 25. I'm pretty sure all my skp files display this bug, but not all of them have scenes saved with shadows. I'd be happy to email them to you.
Regards,
Thanks Jackson .... I'll probably take you up on your offer.
Since I last posted, looks like one of the developers has an idea of what the issue may be. Having your model will help prove the proposed fix is a good one.
Best,
Tricia
-
Tricia...yes, all models that have any lag going from "materials" to "xray" take 20-40% longer in SU6. Same with other button punching changes and tool selections. SU6 sometimes seems faster orbiting...but studying what happens I notice the delay before going to a really fast wireframe and the delay after to get the faces back adds up to as long or longer than SU5.
Bruce...before and after task manager screen captures:
http://www.sketchucation.com/forums/scf/sas/SketchUp/ended-before.jpg
http://www.sketchucation.com/forums/scf/sas/SketchUp/ended.jpg
Set SU priority to high (couldn't find anything about affinity or cores...?): SU5 actually got slower (from 5 to 7sec), SU6 remained the same 13-14sec to change to "xray". I have 10gig available on a 30gig HD, recently ran ad-aware, but was connected to the internet...?
What's next? (Thanks a bunch for this BTW...from the bottom of my angst filled heart!)
-
Is an Inspiron 9300 one with 286 hardware? (1.6gig Pentium M, 128meg ATI Radeon Mobility 300, 2gig ram) Turning off 24 processes (55 down to 31) did drop CPU usage from 5% to 2%...? Still want the file, thanks?
Another general comment I'll add is that a 2meg+ file of geometry only seems to run faster than a 2meg+ file that has components. EDIT: :~p of course it does...nevermind!
Advertisement