Hi Tony and Mark, thanks for helping. I think it was reboot + cruft removal for the win. At least so far the responsiveness is back to normal. Cruft removal brings the folder size down to 35 Mb instead of 12Mb.
I was curious about the images to text/media ratio. I have a combined total of about 21 megs in image files, mp3 files, and config tiddlers. The other 14 M are all basically text tiddlers, and 14 Mb is a lot less than the 120 of yesterday but still pretty substantial. The largest two text tiddlers are 188kb and 51kb. The larger one is a long string of notes I took in Zotero, copypastaed into TW classic about 2009, and imported into TW5 when I upgraded from classic. It is all text with some formatting as far as I can see. The next one is all notes on a particular book. I do a lot of quotes with comments from books and articles so I don't have to dig out the original every time I write "real" stuff from my notes, so it is not like I type everything. It is probably a couple of volumes worth of linear text but not an encyclopedia. I write for a living and have used TW since 2006 for my notes and ideas, so there really is a lot of text. I'd be curious to know where the max memory settings are for Chrome, I think FF has a setting in the regular settings menu to only load pages on focus, which saves a lot of memory. On Chrome, I use an extension called "Tiny Suspender <https://chrome.google.com/webstore/detail/tiny-suspender/bbomjaikkcabgmfaomdichgcodnaeecf?hl=en-US>" which will suspend the tab after x minutes, replacing it in memory with a tiny URI file that brings back the page from the web on focus. It can be turned off per page, which is a good idea on TW5 because it is writable and gives a warning everytime Tiny Suspender goes into action, which is a pain. I seldom run out of memory and know immediately when I have. After a reboot, loading all my Chrome pages w/ tiny suspender active on everything but gmail and TW5, I only use up 4.5 gigs. Even if the whole 120 megs is loading, it should be fine once it is in, but as I said, I don't think everything has to load at once on Node.js, just the core and whatever tiddlers are open, which is one of the reasons I use it. On Sunday, October 21, 2018 at 3:38:43 PM UTC-10, TonyM wrote: > > Rich, > > Remember the browsers have their own internal memory setting that stop > them consuming too much of the host computers memory resources. Chrome and > FireFox have settings that allow this to be increased, especially > appropriate to TiddlyWiki users who's key application is the browser and > TiddlyWikis within that. These browser limits met I was just not utilising > the 16GB on my laptop, upping these has helped me, but is likely to help > you more if your computer has the RAM. > > I would be interested to know where you use the most memory in your > tiddlywikis, because the size you have could be a thousand encyclopedias, > do you include media and images? Few people could write one encyclopedia in > a life time, > > Regards > Tony > > On Monday, October 22, 2018 at 11:01:02 AM UTC+11, Rich wrote: >> >> On Sunday, October 21, 2018 at 3:49:30 AM UTC-10, Jed Carty wrote: >> >>> I am not surprised that the tiddler tool thing I made doesn't do well >>> with that many tiddlers, there are probably a lot of recursion errors. >>> Check if there are any edited tiddlers, they would start with >>> $:/plugin/inmysocks/ >>> Hi Thanks Jed, >> >> >> I deleted all the $_plugins/inmysocks/ tiddlers from the tiddlers folder, >> and there are not any $:/plugin... files at all (probably different >> characters allowed on nodejs/linux) >> >> On Sunday, October 21, 2018 at 3:49:30 AM UTC-10, Jed Carty wrote: >>> >>> I am not surprised that the tiddler tool thing I made doesn't do well >>> with that many tiddlers, there are probably a lot of recursion errors. >>> Check if there are any edited tiddlers, they would start with >>> $:/plugin/inmysocks/ >>> >> >> ~~~ >> Mark S wrote: >> >> I'm stunned with that 120meg number. That's 12x larger than I would call >>> top-normal. But maybe results on Linux are better (haven't tested there >>> yet). How much system memory do you have? >> >> >> Thanks Mark, >> >> I have been keeping notes and writing in TW since 2006. Some of the megs >> are from images, but there is a lot of text too. Eventually, three major >> publications should come out of it: two books and the Great American >> Hypertext Novel :) along with a bunch of subgroups on sound and hearing and >> technology and mediation. >> >> Memory: 16 gigs, but I think one advantage of nodejs is that the tiddlers >> are not all one file, so it does not load the whole thing into memory at >> once. >> >> After your crash, did you check your memory? Did you reboot your system? >>> Or close out your browser? For me FF steals memory growing larger over >>> time. I believe this happens both in Windows and Linux (though to a smaller >>> extent). It has to be periodically reset, and it does appear that TW >>> exercises it more. >> >> >> Did not check memory immediately after the crash, but when linux moves to >> swap, it slows everything down (even with SSD swap space), this is only >> slowing down TW5. Everything else is running fine. Don't use FF often for >> TW5, but checking in FF on a fresh start, it is slowed down there too, with >> the side effect of eating more memory than Chrome in this case. >> >> Didn't think to reboot and I may have not restarted Chrome either. will >> try and see if it fixes things. MS-DOS advice to the rescue! My bet is >> this fixes it :) I still always forget to try it. >> >> In your notes you say you added text slicer, but it sounds like you may >>> have also run it? Which would presumably increase the number of tiddlers. >> >> >> I have run the text slicer on one file and it increased the file count to >> 1185, so I am a bit concerned about that in the long run. the 985 count >> was pre-slice. >> >> I don't know about TiddlerTools, but I have encountered other plugins >>> that generate system tiddlers behind the scenes. Does your tiddler count >>> include system tiddlers? You can use a filter like [all[tiddlers]count[]] >>> in the advanced search to get a full count. >> >> >> Felix Hayashi's TiddlyMap is generating some files, but I have been using >> that without issue for a long time. Did just update it though so I will >> check with him about removing these files. The filter you gave counts 1129 >> tiddlers, 56 *less* than in the file folder count. I found 42 conflict >> resolution files from DropBox that are a byproduct of using the wiki on >> different computers. They are not counted by TW5 and I removed them to a >> backup folder, but they have also been a part of the workflow for years >> without issue. the others are probably retired media files or something. >> >> Good luck! >> >> >> Thanks! will update after reboot >> ~Rich >> > -- You received this message because you are subscribed to the Google Groups "TiddlyWiki" group. To unsubscribe from this group and stop receiving emails from it, send an email to tiddlywiki+unsubscr...@googlegroups.com. To post to this group, send email to tiddlywiki@googlegroups.com. Visit this group at https://groups.google.com/group/tiddlywiki. To view this discussion on the web visit https://groups.google.com/d/msgid/tiddlywiki/fa8027a0-8cb6-4f6a-9475-44fbb1d0d0ab%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.