Closed PowerWeb5 closed 5 years ago
Hello @PowerAccess , wow long post .. I believe you have the record :-)
Not sure I will be able to address all in one response, let me try a few essential things first:
Performance topic:
Strange
Display strategy
Bookmarks in Other bookmarks
Desync
Miscellaneous
I am sure I missed many of what you say. Can we do this first round though ? Thanks, aaFn
Wow, thanks for that detailed description PowerAccess and your reply aaFn!
I recently ran into similar performance issues and wanted to document it as well. My JSON FF bookmark export is ~11MB. I can hardly use the extension anymore because of ~20sec freeze of UI as described by PowerAccess. Extension used to work great and much faster before some (automatic) updates of FF and plugins. So don't know when it started: maybe ~4 weeks ago?
Even adding a new folder, or draging a link into the BSP2 UI-tree takes 10-30 seconds.
But anyway: THANKS! Since "Show parentfolder" died, this extension must have saved countless peoples nerves ;-).
System:
Bookmarks: 28710 Favicons to fetch: 0 (0.0%) Folders: 5199 Separators: 8 Oddities: 0
Hello @RiverRhine , thank you. As a first step, could I ask you to:
If you could repeat it 3 times to see if there are variations.
Hi aaFn, ok, here are the results:
1st hide+show Bookmarktree:
Load local store duration: 59 ms Background load local store duration: 5160 ms Background bypassed FF API for tree load Background tree load duration: 64 ms Background tree build duration: 62 ms Background save duration: 14968 ms Get tree duration: 1 ms PlatformOs: win Setting Windows variations structureVersion: -bnlist-spfldr-img16 disableFavicons_option: false pauseFavicons_option: true Display duration: 4833 ms Total duration: 4893 ms fontFamily = '"Segoe UI"' fontSize = '12px' Stats: Bookmarks: 28782 Favicons to fetch: 39 Folders: 5191 Separators: 8 Oddities: 0
BSP2 version: 2.0.62
2nd run:
Load local store duration: 53 ms Background load local store duration: 5160 ms Background bypassed FF API for tree load Background tree load duration: 64 ms Background tree build duration: 62 ms Background save duration: 14968 ms Get tree duration: 1 ms PlatformOs: win Setting Windows variations structureVersion: -bnlist-spfldr-img16 disableFavicons_option: false pauseFavicons_option: true Display duration: 4335 ms Total duration: 4389 ms fontFamily = '"Segoe UI"' fontSize = '12px' Stats: Bookmarks: 28782 Favicons to fetch: 39 Folders: 5192 Separators: 8 Oddities: 0
BSP2 version: 2.0.62
(I tested some more, but I saw no relevant variations in timing)
… now create and name new folder (takes ~23 sec to finish):
Load local store duration: 40 ms Background load local store duration: 5160 ms Background bypassed FF API for tree load Background tree load duration: 64 ms Background tree build duration: 62 ms Background save duration: 14968 ms Get tree duration: 1 ms PlatformOs: win Setting Windows variations structureVersion: -bnlist-spfldr-img16 disableFavicons_option: false pauseFavicons_option: true Display duration: 4132 ms Total duration: 4173 ms fontFamily = '"Segoe UI"' fontSize = '12px' Stats: Bookmarks: 28782 Favicons to fetch: 39 Folders: 5194 Separators: 8 Oddities: 0
BSP2 version: 2.0.62
In general UI is very slow. Highligthing a row in the tree under the mouse is delayed. The whole process of creating and naming a sub-folder takes ~25 seconds (!)
Thank you
Indeed, something is slowing you down, I see:
The first two are really not normal, especially considering that you have an SSD. The third one also looks long given you have an i7.
Could you open a new tab, and type in the address bar: about:performance
?
Then, if you have a way to copy the screen here, that would be great.
Also, by chance, did you look at your task manager to see if there is not a CPU hog somewhere ?
performance data: BR
Ok, so something is definitely slowing BSP2 in your config, and it could be coming from inside it, although it's not supposed to keep cpu as all actions are short lived (I suppose Hoch means High ..).
Let's continue digging if you agree .. :-)
Hopefully, that will allow me to catch something.
Thank you in advance.
Love to help. Welcome!
Feedback:
although it's not supposed to keep cpu as all actions are short lived
Info: Yes, but I made the snapshot while/after opening BSP2. During the ~20sec UI locks and CPU high.
Red arrow = Close/open BSP2 tree:
Could you look at the Windows Task manager (click on show details then on the processes, and then expand Firefox to see its threads), and place here the captured image of the firefox + its threads line?
Seems to be different in W7Ultimate64. I used ProcExp:
and
FF was in idle above.
Could you open the browser console (Ctrl+Shit+J), click on the bin icon on top left to empty it, then do an operation on BSP2 which is slow like create a sub-folder, come back to the browser console and copy its contents here (text copy and paste) ?
Create new folder -> Browser console output:
as text:
TypeError: this.appMenuStatus is undefined browser-sync.js:364:24 updatePanelPopup chrome://browser/content/browser-sync.js:364 updateAllUI chrome://browser/content/browser-sync.js:197 maybeUpdateUIState chrome://browser/content/browser-sync.js:110 showFxaToolbarMenu chrome://browser/content/browser.js:486 onDOMContentLoaded chrome://browser/content/browser.js:1444 onDOMContentLoaded self-hosted:985 this._box is null browser-sidebar.js:328 get isOpen chrome://browser/content/browser-sidebar.js:328 get currentID chrome://browser/content/browser-sidebar.js:335 onWindowOpen chrome://browser/content/parent/ext-menus.js:994 handleEvent chrome://extensions/content/parent/ext-tabs-base.js:1592 editor is null UrlbarValueFormatter.jsm:303 _formatSearchAlias resource:///modules/UrlbarValueFormatter.jsm:303 update resource:///modules/UrlbarValueFormatter.jsm:59 formatValue chrome://browser/content/urlbarBindings.xml:584 _setValueInternal chrome://global/content/bindings/autocomplete.xml:598 set_value chrome://global/content/bindings/autocomplete.xml:282 URLBarSetURI chrome://browser/content/browser.js:2794 onLocationChange chrome://browser/content/browser.js:5021 callListeners chrome://browser/content/tabbrowser.js:746 _callProgressListeners chrome://browser/content/tabbrowser.js:759 _callProgressListeners chrome://browser/content/tabbrowser.js:4908 onLocationChange chrome://browser/content/tabbrowser.js:5244 _callProgressListeners resource://gre/modules/RemoteWebProgress.jsm:156 receiveMessage resource://gre/modules/RemoteWebProgress.jsm:295 startObserving moz-extension://dde5f1c4-8d49-4884-b363-30b001058176/privileged/FirefoxMonitor.jsm:247 _onOpenWindow moz-extension://dde5f1c4-8d49-4884-b363-30b001058176/privileged/subscripts/EveryWindow.jsm:48 _onOpenWindow self-hosted:989 _delayedStartup chrome://browser/content/browser.js:1721 _delayedStartup self-hosted:985 editor is null UrlbarValueFormatter.jsm:303 _formatSearchAlias resource:///modules/UrlbarValueFormatter.jsm:303 update resource:///modules/UrlbarValueFormatter.jsm:59 formatValue chrome://browser/content/urlbarBindings.xml:584 onSecurityChange chrome://browser/content/browser.js:5175 callListeners chrome://browser/content/tabbrowser.js:746 _callProgressListeners chrome://browser/content/tabbrowser.js:759 _callProgressListeners chrome://browser/content/tabbrowser.js:4908 onSecurityChange chrome://browser/content/tabbrowser.js:5271 _callProgressListeners resource://gre/modules/RemoteWebProgress.jsm:156 receiveMessage resource://gre/modules/RemoteWebProgress.jsm:312 startObserving moz-extension://dde5f1c4-8d49-4884-b363-30b001058176/privileged/FirefoxMonitor.jsm:247 _onOpenWindow moz-extension://dde5f1c4-8d49-4884-b363-30b001058176/privileged/subscripts/EveryWindow.jsm:48 _onOpenWindow self-hosted:989 _delayedStartup chrome://browser/content/browser.js:1721 _delayedStartup self-hosted:985 Unchecked lastError value: Error: ID already exists: open-link-in-new-temporary-container-tab ExtensionCommon.jsm:679 withLastError resource://gre/modules/ExtensionCommon.jsm:679 create chrome://browser/content/child/ext-menus.js:138 Tastenereignis ist in manchen Tastaturlayouts nicht verfügbar: Taste="i" Modifikatoren="accel,alt,shift" ID=“key_browserToolbox” browser.xul Cannot send function call result: other side closed connection (call data: ({path:"bookmarks.update", args:["kZiqxitKj5Ai", {title:"dbg", url:null}]}))
Info 1: No enties in browser console while show/hide/show BSP2 Info 2: Black magic, very strange!: First time after opening the Browser Console, the tree was fast and responsive for ONE create of a subfolder. Slowed down again afterwards (don't laugh pls ;-))
Tested : Disabled this plugins temporarily: and
... but doesn't make a difference. (have to go offline now till tonight ... I continue then.)
Info 2: Black magic, very strange!: First time after opening the Browser Console, the tree was fast and responsive for ONE create of a subfolder. Slowed down again afterwards (don't laugh pls ;-))
Ouch .. that makes me suspect that the save to disk is the problem as it is lasting 14s ... and there's one each time you make a modification to the tree (normal, we do not want to lose its state). This is usually very fast and goes unnoticed, but when being that long that must impact normal operations.
As a matter of fact, ever since FF 66 where Firefox migrated the add-on local storage to IndexedDB, I have seen random problems appear for users with big numbers of bookmarks. Unfortunately, there is no easy way to come back to the old mode once it has been migrated by FF automatically. See: https://www.mozilla.org/en-US/firefox/66.0/releasenotes/ https://blog.mozilla.org/addons/2019/02/15/extensions-in-firefox-66/ https://blog.mozilla.org/addons/2018/08/03/new-backend-for-storage-local-api/ https://www.reddit.com/r/firefox/comments/ar78wr/firefox_extensions_to_use_different_storage_type/egmdkul/ https://bugzilla.mozilla.org/show_bug.cgi?id=1406181
And there's nothing else in the trace that reveals anything meaningful :-(
One way to verify if my suspicion is right would be to:
If the last one is fast again, we're in the problem that stores are long (and also read).
14 seconds anyway is a lot !!
I am not sure how that can be optimized, I do not know IndexedDB's in Firefox, and I read that the sqlite form used for add-on local storage is not easily readable / controllable.
Maybe a simple naive question:
To find where your profile is, do the following:
I do not know where the add-on data is stored anymore, it was in the browser-extension-data directory in the past, but now it's somewhere else I do not know, however I presume it is still in the profile directory
Back again. Pulling things together to provide the details ...
Yesterday FF updtaed to latest (67.0.4 (64-Bit)). It seems to bee a little faster now ?!
Ouch .. ... I do not know where the add-on data is stored anymore, it was in the browser-extension-data directory in the past, but now it's somewhere else I do not know, however I presume it is still in the profile directory ... since FF 66 where Firefox migrated the add-on local storage to IndexedD ...
Ouch=ACK. A devs nightmare losing control about storage ;-). At least Moz had good intentions (perf+sec)) Anyway, going with Moz and FF through it's lows and highs for all the years ;-).
Digging into it:
After restart and waiting all things calmed down (like wait a few minutes doing nothing with FF, verify on Windows Task Mgr that cpu is idle .. etc ..) Do one of the things which can be slow Verify it is going fast this time Wait again all things calm down (and at least 14s or more ..) Do another one
Done. Result: 1st, second and so on are about the same. (can't reproduce the "once fast" ?!)
Starting here: FF idle, github page and BSP2 options page open:
My BSP2 options:
Operation profiled: "create subfolder (~5sec), type new folder name (~2 sec)+enter, folder created (~2sec)"
during operation:
... after operation:
Traces: (After folder operation)
BSP2 version: 2.0.62 Favicon display duration: 8058 ms
FF console (started fresh ):
Conclusion: "Somehow" faster after last update (strange coincidence?) but still slow.
To your suggestions/questions:
Where is your Firefox profile stored ? And is there free space room on that drive ? Maybe do a cleanup (with system files) and a defrag on that drive ... (optimize disk)
Prog+profiles all live on C=SSD. SSD-performance:
I do not know where the add-on data is stored anymore, it was in the browser-extension-data directory in the past, but now it's somewhere else I do not know, however I presume it is still in the profile directory
... yepp, anything else would be insane ;-).
BTW, I made an other test on my Macpro (SSD as well): Found an older FF version (ups, don't rember version) there!: BSP2 lighning fast like ever. After update to FF current: "Folder create" takes ~2 sec.
hmmm. Where to go next?! BR and thx for your work!
Thanks for the very detailed investigation.
Yes the Samsung 850 EVO is a good beast, I have two of them, speed like hell :-) And it looks like you have plenty of space, so that's not the problem indeed (was asking just in case ..).
Recurring schema, I see like last time two errors on moz-extension://dde5f1c4-8d49-4884-b363-30b001058176 inside the browser console trace.
To know which extension this is:
However, your last experiment does point at something related to FF version since this is the only thing varing between he two attempts you made, before and after the FF upgrade :-(
So that's defnitely it, somewhere, most probably with FF66 and the way they changed how add-ons data is stored internally, they degraded the function :-(
I am not sure what to do from there.
For now, I'd prefer to stay in an observation period if you agree. Being over-reactive while FF is stabilizing after their 66 version big change is probably not wise, and a loss of time if I find some compromise whiich only lives a few days until their next update .. I'll think more meanwhile if something low cost can be made as workaround.
Hi there! Thanks for the feedback and hints.
Click on Tools -> Add-ons, go to the little "parameter wheel" button, click on it to get a menu and then click on "Debug add-ons" This will bring you to a detailed list of modules with their internal UUID's You should find there the one with dde5f1c4-8d49-4884-b363-30b001058176 by searching on dde5f1c4
... didn't work. No such UUID (and extension) But
Another way: open a new Tab, type "about:memory" in its address field
Then click on the "Verbose" checkbox for "Show memory reports, and then on "Measure" After a few seconds, you'll get a very detailed report about the heap Search for the string dde5f1c4, and you should find a few lines saying to you which add-on this is
worked. What I found is not a visible extension, but this:
17 (100.0%) -- extensions
├───1 (05.88%) ── Extension(id={6d96bb5e-1175-4ebf-8ab5-5f56f1c79f65}, name="Deaktivierungs-Add-on von Google Analytics", baseURL=moz-extension://071d0795-02b6-4892-a956-89b4a0b7997c/)
├───1 (05.88%) ── Extension(id={c607c8df-14a7-4f28-894f-29e8722976af}, name="Temporary Containers", baseURL=moz-extension://c9a3f5ff-bfd9-40ac-ac5e-7f792388c9c5/)
├───1 (05.88%) ── Extension(id={e824f0e4-72f7-4295-8879-d8ff4bf348d2}, name="Invidition", baseURL=moz-extension://35ed773c-4b12-4c1a-86e0-10fc3c2a094e/)
├───1 (05.88%) ── Extension(id=@testpilot-containers, name="Firefox Multi-Account Containers", baseURL=moz-extension://baa727c9-9294-41d0-87a7-735e6823769a/)
├───1 (05.88%) ── Extension(id=bookmarksearchplus2@aafn.org, name="Bookmark search plus 2", baseURL=moz-extension://c276880d-bc74-4ff2-9e27-6f1bba801d99/)
├───1 (05.88%) ── Extension(id=formautofill@mozilla.org, name="Form Autofill", baseURL=moz-extension://df6ea9b2-0815-4e76-9cb4-3e18ff94e1a9/)
**├───1 (05.88%) ── Extension(id=fxmonitor@mozilla.org, name="Firefox Monitor", baseURL=moz-extension://dde5f1c4-8d49-4884-b363-30b001058176/)**
├───1 (05.88%) ── Extension(id=hotfix-update-xpi-intermediate@mozilla.com, name="hotfix-update-xpi-intermediate", baseURL=moz-extension://13771935-216b-4440-a0c0-7d7dffe69f1e/)
├───1 (05.88%) ── Extension(id=https-everywhere@eff.org, name="HTTPS Everywhere", baseURL=moz-extension://dd58bf0c-209d-4678-bf36-4ffd90e62dfa/)
├───1 (05.88%) ── Extension(id=jid1-BoFifL9Vbdl2zQ@jetpack, name="Decentraleyes", baseURL=moz-extension://dfc646a2-51d1-4a19-9969-6502fa20242b/)
├───1 (05.88%) ── Extension(id=jid1-MnnxcxisBPnSXQ@jetpack, name="Privacy Badger", baseURL=moz-extension://cc108c4c-52ec-474b-a2df-61600874e87c/)
├───1 (05.88%) ── Extension(id=jid1-TMndP6cdKgxLcQ@jetpack, name="Mate Translate – Übersetzer, Wörterbuch", baseURL=moz-extension://d4afde02-d117-41b1-85df-934dd163ec8f/)
├───1 (05.88%) ── Extension(id=jsonview@brh.numbera.com, name="JSONView", baseURL=moz-extension://a5aac941-9cd3-4f66-bf17-1201f5bd6229/)
├───1 (05.88%) ── Extension(id=screenshots@mozilla.org, name="Firefox Screenshots", baseURL=moz-extension://6d0bd23f-c79b-4e25-abc3-3d61576bd57f/)
├───1 (05.88%) ── Extension(id=uBlock0@raymondhill.net, name="uBlock Origin", baseURL=moz-extension://b85953fb-6321-485b-9907-ced13d92c69b/)
├───1 (05.88%) ── Extension(id=wayback_machine@mozilla.org, name="Wayback Machine", baseURL=moz-extension://2fc2a9c5-5a4d-484e-bcf8-811988ba1456/)
└───1 (05.88%) ── Extension(id=webcompat@mozilla.org, name="Web Compat", baseURL=moz-extension://e71a1d4e-0911-48a1-9347-d999a015ef8c/)
with ├───1 (05.88%) ── Extension(id=fxmonitor@mozilla.org, name="Firefox Monitor", baseURL=moz-extension://dde5f1c4-8d49-4884-b363-30b001058176/)
... no idea what that is.
All occurences of that for reference. Just in case that makes sense to you:
So that's defnitely it, somewhere, most probably with FF66 and the way they changed how add-ons data is stored internally, they degraded the function :-(
ACK.
...find save strategies which are better with the new FF versions ,,, ... manage two formats at the same time :-(
Don't do that ;-). What a mess adding a second storage abstraction and migration layer!
For now, I'd prefer to stay in an observation period if you agree. Being over-reactive while FF is stabilizing after their 66 version big change is probably not wise, and a loss of time if I find some compromise whiich only lives a few days until their next update .
100 ACK, but on thing: Couldn't you maybe point the MozDevs to "that storage performance situation" so that they can test and improve the storage backend with that information? Then it's a little biased and not a random walk for their improvements ;-).
Anyway: You created one of the "couldn't live without" - extensions here! Must have been "some hours ..."
As you said: Let's wait some "versions of FF" and come back to this, if things don't improve or get worse. If you update the FF devs, put me in that loop pls.
BR Andreas
BTW: In which timezone you are in?
Hello @RiverRhine,
On Firefox monitor, I found this: https://blog.mozilla.org/futurereleases/2018/06/25/testing-firefox-monitor-a-new-security-tool/ https://winaero.com/blog/mozilla-enables-firefox-monitor-extension-firefox-67/ https://monitor.firefox.com/
I am in the same timezone as you I believe :-) CET / CEST https://www.timeanddate.com/time/zones/cet
I thought of a quick workaround, not curing the problem, but moving it so that it is less painful hopefully:
Drawback would be a small chance of losing something in the BSP2 cache if the user exits from Firefox before a triggered 10s timer for save completely elapsed and the save could really occur. Hence the delay value should not be too big. And to limit such a drawback to a minority of users, only do that when I see the save time is long on their config.
Advantage should be that the save wouldn't perturb anymore immediate ongoing actions (only later ... with a chance the user doesn't see it because he stopped interacting).
How does that sound ?
Hi @aaFn, thx for the info about "monitor". An OK approach. But I hope Moz doesn't go the way to far to overcrowd/overfeature everything ...
Yes, CET here as well ;-)
Your suggestion sounds great: It would decouple the UI Interactions from the book keepings/sync in the background. The risk of loosing some links is def. worth a more fluent and nonblocking UI. The result would be a "real time UI tree" again, right?
Would love it!
The result would be a "real time UI tree" again, right?
That is the purpose yes .. let me try then.
Great. Small suggestion (and just a guess): There should be an event/hook for FF exit/termination. To avoid any data loss this could probably also call the "store/sync to disk" function additional to the call from the timer? Might be just some more lines? But: not a dev anymore, so forget pls if nonsense ;-)! Thx for your efforts again.
2.0.63 is out with the workaround to restore UI responsiveness. It appears to work on my tests, try on your side.
As said above, save to local store is delayed by 10s after last action on the UI, meaning that each action on the BSP2 UI restarts the 10s timer delaying the save.
I didn't find anything to hook for FF stop, the closest event would be browser.windows.onRemoved (https://developer.mozilla.org/en-US/docs/Mozilla/Add-ons/WebExtensions/API/windows/onRemoved), however it is triggered when any window is closed, including the property one for editing bookmarks, so I decided not to use it as an indication of possible FF close to force a save.
Impact, if a "large number of bookmarks" user with "slow save" (> 2s) closes before the 10s elapse, is that BSP2 will be de-stnchronized from Firefox on next restart, loading a state corresopnding to the last save which occurred.
Remedy for such rare cases will be to hit the Reload all bookmarks from FF now
button in the BSP2 options page, so nothing will be lost.
Looks like an acceptable drawback, which will hit anyway only a few user base, but for the best of coming back to UI responsiveness :-) (hopefully)
I also suspect that this long and strong cpu consumption by FF on saves with large numbers of bookmarks is at the origin of some of you guys seeing bookmark modifications occasionally not saved by FF in its internal DB, resulting in loss of work when you reload from the internal DB. At least, the phenomenon also started to appear fi FF66 and the new Indexed DB format on local storage, so that should have a high correlation factor. With delayed save by 10s, I am hoping this won't occur anymore.
@RiverRhine, @PowerAccess, let me know how that goes. Cheers, aaFn.
Man, that's it. Works like a charm here! "New folder test + drop link to it": There is still a ~1-2sec delay until the folder name dialog opens. Which is absolutely ok for me and no issue. After a while (?10s) the "heavy cpu" kicks in: Best to be seen in core 6 and 7 above. First spike: Dialog+ typing. Second spike: You know the magic..
Didn't test if UI is blocked then, I guess not.
Anyway: green light from my side. Many thanks and keep BSP2 up ;-).
Btw: I didn't see loss of data so far, but will have an eye on it in future.
Nearly 2 weeks, it seems we are good now with the workaround :-) Closing the issue, let me know if something comes back.
Still all green lights frome here. Again: Thx!
I have encountered several critical performance and reliability issues, which I've detailed below, together with suggested ways to fix those. I really appreciate your having developed this incredibly useful extension (with features that should be built into Firefox but it still sadly lacks) but also have some suggestions for a few quick enhancements that I believe will greatly improve its usefulness, such as with "Add Bookmark Here" and Quick Move/Rename/Bookmark Editing, which I've also detailed below together with details and suggested fixes for critical performance and frequent bookmark corruption/syncing issues (which occur despite having tried different options for workarounds, and like seen in screenshot and related issues referenced further below).
Critical Issues, Suggested Fixes & Enhancements
BSP locks up Firefox for 60 seconds (making it unresponsive) when have 40K bookmarks (30 seconds to display + 30 seconds after without any noticeable visual changes) every time I switch to/from the Sidebar regardless of settings (w/ or w/o Pause favicon loading and/or Old immediate loading) even with only 15 bookmarks/folders shown and had all favicons and bookmarks loaded and cached during same session.
Need to load favicons for visible bookmarks (for expanded top-level folders in BSP2) first, and load others JIT, by default
Bookmarks keep ending up in "Other Bookmarks" folder, and never able to be deleted, very frequently, and even deleting from BSP2 itself fails to auto-resync or auto-remove orphaned bookmarks from your own cached database (which should be occurring instead, as detailed below):
This has happened maybe 10 times over the past year, and always the same case: it ends up under Other Bookmarks folder.
This is not just an issue of them being "Out-of-Sync" (as the developer had referred to in reply to related issue #97) because of the following reasons, and I would suggest implementing auto-fixing/detection in the ways described below.
The bookmark always appears in the same "Other Bookmarks" folder (even though it was never in that folder previously).
When viewing/editing an orphaned/left over bookmark, you should be able to check and detect that at least when trying to edit it via BSP2's Properties dialog.
If you are monitoring changes to all bookmarks, then shouldn't you at least be able to detect and apply changes made to a bookmark even if you had "missed 1 change notification"?
All of the following methods for trying to move/update/delete this bookmark once I see it incorrectly shown under Other Bookmarks (when I would suggest that you handle all of the following cases, able to detect, update/resync that bookmark or at least remove the left-over bookmark from your cache when detecting no corresponding bookmark exists in Firefox Places DB):
There should be no reason that even attempting to Delete Bookmark for a bookmark in the BSP2 sidebar ends up failing to delete it from the BSP2 sidebar.
Add right click menu that should be shown when right clicking background of BSP2 sidebar (not a specific bookmark, eg. like available in Tree Style Tab).
Right click on Folder > Add Bookmark Here is critical to add
Quick Rename via Right-click on bookmark > Rename should be supported
Edit boxes should be shown at bottom of BSP2 sidebar to allow for fast editing as true Bookmark Manager replacement) for the following fields:
Enhancements/fixes for the Properties dialog (and critical performance issues with it)
Screenshot of Issues
You can see in the below screenshot how few top-level bookmarks and folders I have. You can also see, as the only bookmark under "Other Bookmarks" in BSP sidebar, yet another frequently encountered bookmark which remains out-of-sync, shown under "Other Bookmarks" (as always occurs for those, despite it never having been in that folder at any point) even after trying to right click > Delete Bookmark it in BSP Sidebar and even after the bookmark was deleted via Ctrl-D (as you can see with Star icon not light up, for the active tab opened for that URL):
Related Issues
97 BSP2 can not delete a bookmark
My Configuration