Hi again,
I hesitated to report this problem as a bug, after I 1st discovered it, because it wasn't clear enough to me, that it might actually be a bug. But I've been continuing to use v 0.46 while watching my Task Manager, and want to report what more I've learned.
I'm still having the same issue of the CPU maxing out at 100% and freezing my system, by Inkscape (0.46)
alone. It tends to happen more often, and last longer, at higher zoom factors. I've noticed that gradients tend to magnify the effect, and the more gradients, the higher cpu use. The same with layers, although additional layers seem not to have as strong an influence as gradients. Also, additional....I'm not sure the right terminology here, but my best guess is "Z Layers" (as opposed to regular Layers). The more "Z Layers", the stronger the effect. I have not been able to determine whether regular Layers or Z Layers have the stronger influence, but gradients are stronger than either type of Layers.
I've also learned that this effect is stronger when I'm viewing an area on the canvas which contains images (or portions thereof). So for example, working an an image with a lot of gradients, and having trouble with this issue, I've found that if I scroll to an empty area of canvas, that I can zoom without limitations. But when I try to scroll back onto the image, scrolling is painfully slow. However, zooming with the image in view is MUCH slower than scrolling with the image in view. Also, if it's not clear by what I've said so far, this effect is less strong when images (or portions of) with fewer layers and/or gradients are in the viewing area. So the effect is not across the board, it depends on the complexity of what's in the viewing area. Also, this problem seems to get worse, the longer I work on an image. Often, I can Save and close the window, then open it up again, and have a faster response.
Further, I've learned that I can ease this issue somewhat, by closing visibility of layers that I don't need for the particular thing I'm doing on the canvas. As I mentioned already, reducing the zoom factor will also ease matters. So that if I can't add nodes by double-clicking, or am unable to do something, I can close unneeded layers, and/or zoom out, until the double-click (or whatever) will work. So far, I've been able to work around this, but a situation where I need a high zoom to properly position a node, or do something which the higher zoom prevents (by using too much memory), could easily present itself.
I've found I can also somewhat improve this issue by closing programs which tend to use more memory, which of course frees up more for Inkscape to use. I think I noted this in a previous reply, but I'm running 1.24 GIGs of RAM, which makes this issue nearly inconceivable to me. FYI, another InkscapeForum user has recently reported this same issue (
viewtopic.php?f=29&t=1151). However, they report their memory at 768 MB, and note their (presumably) highest CPU use by Inkscape at 72%.
Aside from being incredibly annoying, my concern is that I''m NOT a graphic artist, and thus far can only work on simple images. I'm still very much in the process of learning how to use Inkscape, and have not attempted anything too complex. I have to imagine seasoned Inkscape users are capable of creating much more complex images. So in addition to my 1.25 GIGs, I'm not in a position to really create such a problem by making very complex images.
So I'm still not sure if this behavior is within the realm of possibility for Inkscape, or if it's serious enough to submit a bug report. It's certainly bothersome enough, lol. But I'm not sure if it's expected, or possibly has a solution. What do you all think?
FYI, when this problem is happening, at worse, I wait between 1 and 2 minutes for a page up or down scroll. A single click on the up or down arrow, still takes nearly a minute. Zooming is nearly impossible, and again, I wait 1 to 2 minutes for a single click on the zoom button (bottom right corner) -- not holding down the button, just a click. If I hold down the zoom button, I can easily wait 5 minutes for the request to be completed. (Well actually, at worse, my system is frozen and I can't do anything, but it only lasts for a moment or 2 -- so I guess the aformentioned is how bad it gets just before my system freezes.)
For further comparison, the image I've been working on, as I've been observing this issue, has 15 regular Layers. On average, I would say each regular Layer has 2 Z Layers (some 1, some 3). It has 15 to 20 gradients, and admitedly, were I more experienced, I probably could have done what I want with fewer gradients. Also, because I'm still struggling to learn gradients, there are several I made, that I ended up not using. I'm not sure if they are counted in the memory, or would affect this issue. The other user I mentioned above, who is having this problem, noted the problem with a lot of clones. I have not used Clones, but have used Duplicate a lot. I haven't really set up a situation where I can compare the use of Duplicates and Clones as far as memory use. So anyway, I'm not sure if Duplicates should matter, but I mention it just in case. Um, I believe the canvas is the B8 size ~~ smaller than 500 x 400 px in any case. (And of course I can look up the exact size, if it matters.)
Also, I'm using Windows XP SP 2 (soon to be SP 3). I have something like 20 gigs free of hard drive memory and 1.25 gigs RAM. I couldn't say what
kind of RAM I have, but could probably look it up, if it matters. Could the page file memory be causing this, or have any potential for solving it? I guess I don't understand the page file memory very well. Is there any other aspect of memory which might be useful to note. Most of the data presented in the Task Manager I don't understand. But could it be helpful to look more closely at some of it?
So what do you all think? Is this expected behavior? Does it rise to the level of bug submission? Is there anything else I can do, besides what I've mentioned already (close other programs, close layers, zoom out, etc.) to either ease or solve the problem? Bug potential aside, this problem is quite annoying!
Thank you very much for your time and attention to this issue. I really appreciate your help. I'm glad to contribute whatever I can, as far as submitting a bug. But I wouldn't want to burden the process with a non-issue, either.