CharlotteTheHarlot

Understood. (EDIT: strike entire paragraph, Opera currently has no dependencies on 3rd party Java/JRE whatsoever for normal operation, see next my post). Where Javascript comes in is in the browser functionality required to process all the webpages coded by script kiddies (but don't get me started on this). Having cut my teeth on C, I am ok with Javascript. I am seeing references to houskeeping particularly garbage collection which I assume is a practice common to Java coded apps that employs Javascript. Perhaps it is comparable to C++ destructors. Since the Garbage Collection is hardwired into the Java Console (*) I'm betting that this maintenance is tied to one or both Java/Javascript (which are joined at the hip in this example). EDIT * that is the Tools > Advanced > Java Console. If you open it you can then press the letter 'h' and a menu appears: Commands: ---------------------------------------- c - clear console f - run finalization g - run garbage collection h - help l - list cached classloaders m - memory usage t - dump threads x - clear classloader cache (EDIT: that menu is dependent on 3rd party JRE, not seen when JRE is removed, see next post) This is where Opera and Java and Javascript all collide. I am finding many references to this Garbage Collection along side Javascript all over the place. It looks like a memory management technique/method. In fact it is even mentioned around 2006 in MSKB 919237 You may experience slow performance when you view a Web page that uses JScript in Internet Explorer 6. This accounts for at least one update to Jscript.dll. I'm really just casting a wide net here hoping that someone with expertise can shed some light on this. @triger49 ... I'm now convinced that Opera (at least for me) is not pegging the CPU at all, it appears to be 0% CPU utilization. If you ever get a chance to look at this again in ProcExp, please note the CPU useage for Opera.exe and let me know if it flatlines. My testing shows that Opera is suspended, which explains why all keyboard/mouse events are cached (by the OS no doubt), all download packets bounce, and any popped up menus are frozen on-screen during the 'freeze'. Most Process Monitors fail to segregate the individual process so that the Opera problem can be seen in isolation. See this thread.

Let me see, as far as script related settings: I have all of these disabled ... 0 Allow resizing of windows 0 Allow moving of windows 0 Allow raising of windows 0 Allow lowering of windows 0 Allow changing of status field 0 Allow scripts to detect context menu events 0 Allow script to hide address bar 0 Open Console on error I have no user.js scripting enabled. I do see this re-routed in HOSTS ... 127.0.0.1 ssl.google-analytics.com I just re-enabled it (commented it out) ... # 127.0.0.1 ssl.google-analytics.com Well that wasn't it. Here is the log after the HOSTS change (I was just clicking on a bunch of open tabs, here and at Opera so the entries reflect the focused tab): --- 03 Jan 2010 04:53:27 --- [url="http://www.msfn.org/board/win9x-tools-realtime-process-monitors-t140830.html"]http://www.msfn.org/board/win9x-tools-real...rs-t140830.html[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace ;;;;;;;;; HOSTS re-enabled # 127.0.0.1 ssl.google-analytics.com ... 20010-01-03 04:58 --- 03 Jan 2010 05:09:30 --- [url="http://www.opera.com/docs/operafiles/"]http://www.opera.com/docs/operafiles/[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 05:09:30 --- [url="http://my.opera.com/community/forums/topic.dml?id=200095&abc=&page=3&skip=100&show=&perscreen=50"]http://my.opera.com/community/forums/topic...mp;perscreen=50[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 05:09:30 --- [url="http://my.opera.com/community/forums/topic.dml?id=260737"]http://my.opera.com/community/forums/topic.dml?id=260737[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 05:09:30 --- [url="http://my.opera.com/community/forums/topic.dml?id=295728"]http://my.opera.com/community/forums/topic.dml?id=295728[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 05:09:30 --- [url="http://my.opera.com/community/forums/topic.dml?id=310331"]http://my.opera.com/community/forums/topic.dml?id=310331[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 05:09:30 --- [url="http://my.opera.com/community/forums/topic.dml?id=256618"]http://my.opera.com/community/forums/topic.dml?id=256618[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 05:09:30 --- [url="http://www.google-analytics.com/ga.js"]http://www.google-analytics.com/ga.js[/url] Linked script not loaded --- 03 Jan 2010 05:09:30 --- [url="http://www.google-analytics.com/ga.js"]http://www.google-analytics.com/ga.js[/url] Linked script not loaded --- 03 Jan 2010 05:09:30 --- [url="http://www.google-analytics.com/ga.js"]http://www.google-analytics.com/ga.js[/url] Linked script not loaded --- 03 Jan 2010 05:09:30 --- [url="http://www.google-analytics.com/ga.js"]http://www.google-analytics.com/ga.js[/url] Linked script not loaded --- 03 Jan 2010 05:09:30 --- [url="http://www.google-analytics.com/ga.js"]http://www.google-analytics.com/ga.js[/url] Linked script not loaded --- 03 Jan 2010 05:09:30 --- [url="http://www.google-analytics.com/ga.js"]http://www.google-analytics.com/ga.js[/url] Linked script not loaded --- 03 Jan 2010 05:09:30 --- [url="http://www.google-analytics.com/ga.js"]http://www.google-analytics.com/ga.js[/url] Linked script not loaded --- 03 Jan 2010 05:09:31 --- [url="http://my.opera.com/community/forums/topic.dml?id=260737"]http://my.opera.com/community/forums/topic.dml?id=260737[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 05:09:31 --- [url="http://my.opera.com/community/forums/topic.dml?id=295728"]http://my.opera.com/community/forums/topic.dml?id=295728[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 05:09:31 --- [url="http://my.opera.com/community/forums/topic.dml?id=310331"]http://my.opera.com/community/forums/topic.dml?id=310331[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 05:09:31 --- [url="http://my.opera.com/community/forums/forum.dml?id=26"]http://my.opera.com/community/forums/forum.dml?id=26[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 05:09:31 --- [url="http://my.opera.com/community/forums/topic.dml?id=256618"]http://my.opera.com/community/forums/topic.dml?id=256618[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 05:09:31 --- [url="http://my.opera.com/community/forums/topic.dml?id=334421&t=1260761556&page=1"]http://my.opera.com/community/forums/topic...1556&page=1[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 05:09:33 --- [url="http://www.msfn.org/board/javascript-opera-complains-t140832.html"]http://www.msfn.org/board/javascript-opera...ns-t140832.html[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 05:09:33 --- [url="http://www.msfn.org/board/win9x-tools-realtime-process-monitors-t140830.html"]http://www.msfn.org/board/win9x-tools-real...rs-t140830.html[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 05:09:51 --- [url="http://ask.metafilter.com/87642/Opera-Browser-Woes"]http://ask.metafilter.com/87642/Opera-Browser-Woes[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 05:18:12 --- [url="http://www.msfn.org/board/add-reply-f23-t140832.html"]http://www.msfn.org/board/add-reply-f23-t140832.html[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace That last entry is from this very post I am editing. Could there be some other servers with obsfucated names relating to GA that also might be hiding in HOSTS? Thanks...CTH

While working on a problem in Opera unrelated to MSFN, I had reason to enable its error logging feature. Here is an example found in the log of something it flags as an error: --- 03 Jan 2010 02:13:50 --- [url="http://www.msfn.org/board/index.php"]http://www.msfn.org/board/index.php[/url]? Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace The thing that pops up over and over is that item Undefined variable: _gat. Means nothing to me, maybe it does to the administrators? I should note that I see entries from other websites mentioning variables as well (e.g., urchinTracker). Also it seems that the log reflects the currently focused webpage. Here are the last few entries right from the log while I have been here at MSFN: --- 03 Jan 2010 02:19:15 --- [url="http://www.msfn.org/board/win9x-tools-realtime-process-monitors-t140830.html"]http://www.msfn.org/board/win9x-tools-real...rs-t140830.html[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 02:19:25 --- [url="http://www.msfn.org/board/add-reply-f8-t140830.html"]http://www.msfn.org/board/add-reply-f8-t140830.html[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 02:19:52 --- [url="http://www.msfn.org/board/win9x-tools-realtime-process-monitors-t140830-pid-900998.html"]http://www.msfn.org/board/win9x-tools-real...pid-900998.html[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 02:20:17 --- [url="http://www.msfn.org/board/add-reply-f8-t137637.html"]http://www.msfn.org/board/add-reply-f8-t137637.html[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 02:32:36 --- [url="http://www.msfn.org/board/index.php"]http://www.msfn.org/board/index.php[/url]? Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 02:33:05 --- [url="http://www.msfn.org/board/forum.html"]http://www.msfn.org/board/forum.html[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 02:51:36 --- [url="http://www.msfn.org/board/index.php"]http://www.msfn.org/board/index.php[/url]? Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 02:54:15 --- [url="http://www.msfn.org/board/opera-10-t137637-pid-901001.html"]http://www.msfn.org/board/opera-10-t137637-pid-901001.html[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 02:54:42 --- [url="http://www.msfn.org/board/site-forum-issues-f23.html"]http://www.msfn.org/board/site-forum-issues-f23.html[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace --- 03 Jan 2010 02:55:07 --- [url="http://www.msfn.org/board/create-new-post-in-site-forum-issues-f23.html"]http://www.msfn.org/board/create-new-post-...issues-f23.html[/url] Inline script thread Error: name: ReferenceError message: Statement on line 2: Undefined variable: _gat stacktrace: n/a; see opera:config#UserPrefs|Exceptions Have Stacktrace I'm pretty sure that last line of each entry is stating that I have Javascript debugging disabled or something. Does anyone enable that? And would it reduce or increase these error entries? Just thought I would bring this to your attention.

This Opera freeze problem is getting downright interesting. Unless I'm mistaken, the freeze seems to be almost on a clock of multiples of six minutes (sometimes 6, sometimes 12 etc). Here is what I have manually noted in the past couple of hours (rough times eyeballed from the tray clock): ;###### VERIFIED OPERA FREEZE: 00:02 ;###### VERIFIED OPERA FREEZE: 00:08 ;###### VERIFIED OPERA FREEZE: 00:14 ;###### VERIFIED OPERA FREEZE: 00:26 ;###### VERIFIED OPERA FREEZE: 00:33 ;###### VERIFIED OPERA FREEZE: 00:39 ;###### VERIFIED OPERA FREEZE: 00:46 ;###### VERIFIED OPERA FREEZE: 00:52 ;###### VERIFIED OPERA FREEZE: 00:58 ;###### VERIFIED OPERA FREEZE: 01:04 ;###### VERIFIED OPERA FREEZE: 01:10 ;###### VERIFIED OPERA FREEZE: 01:16 ;###### VERIFIED OPERA FREEZE: 01:28 ;###### VERIFIED OPERA FREEZE: 01:39 ;###### VERIFIED OPERA FREEZE: 01:52 ;###### VERIFIED OPERA FREEZE: 02:04 ;###### VERIFIED OPERA FREEZE: 02:11 ;###### VERIFIED OPERA FREEZE: 02:16 ;###### VERIFIED OPERA FREEZE: 02:28 ;###### VERIFIED OPERA FREEZE: 02:34 ;###### VERIFIED OPERA FREEZE: 02:40 ;###### VERIFIED OPERA FREEZE: 02:52 BTW: If anyone has some knowledge of Win9x compatible Process Monitors please see this thread. With a logging utility as described in that thread, I could produce a much better list. My current theory is that there is some scheduled housekeeping going on, this housekeeping for some strange reason places the browser into sleep. As I am poorly versed in Java/JRE programming I wonder if any members here who happen to know this topic might recognize sleep and/or scheduled housekeeping as common boilerplate practices (perhaps related to the so-called Javascript garbage collection?). I found an old discussion that mirrors my experience (see here). That thread is from 2008-04-01 and the version was Opera 9.26 (which I remember as being perfect). Nevertheless the description matches so IMHO we are probably blaming the 10.xx branch wrongly. EDIT: to everyone experiencing the freezes/pauses, you must read through this thread at opera. It is right on point and vividly describes the frustration. I only found that page through Google and it never turned up in all the local searches I have done over there at Opera during the past few weeks.

(reserved)

Gang, I'm hoping someone can direct me to a utility to help me out. I'm knee-deep in trying to solve this nagging Opera 10.xx problem of constant freeze-ups (info found in this thread). I have been able to determine that it is not a problem of Opera nailing the CPU to 100% after all. In fact it appears to be a kind of time-out or sleep, in other words CPU utilization at 0% What I really need now is a process monitor that can target a specific process and/or a specific thread. What would be the absolute best tool for this job is a realtime monitor that can sound an alarm or write to a logfile when a designated process or thread goes to 0% (or any percentage we assign). Unless I am getting rusty, I have exhausted the capabilities of all of the following without finding that ability ... Microsoft SysMon (Win98se/WinME) ... Does not seem to allow individual process views. Instead there is a monolithic Kernel summing of all CPU activity. System Internals ProcExp (Process Explorer v11.11) ... Seems very accurate, and when Opera hits 0% it blanks the CPU field completely. Unfortunately there is no way to trigger an alarm that I can find. Its litttle graphic display also seems to be monolithic like SysMon which makes the live graph illustration in the System Tray little more than eye candy really. But someone please correct me if I am wrong. System Internals DLLview ... bare module info, no CPU. Enrico Del Fante ATM (Another Task Manager) ... As super detailed as this baby is, it also leaves out alarms and logging. I had high hopes for this one. EDIT: The author seems to have vanished and Wayback has pages only until 2003 (abandonware?). If anyone knows anything about this please share your info! In the ABOUT box the author mentions an online help file (none seems to be included in the distro). This missing help file could be useful to look at. NirSoft CurrProcess ... another nice one. This has much potential (e.g., dump process to file!) The only thing that he forgot was CPU activity. Microsoft pView95 ... only the barest of info. Microsoft pView95 (IgorNys modded) ... slightly more than the barest of info. Igor Nys PrcView ... better, more like ProcExp and ATM, same shortcomings MiTeC Process-Viewer PV ... much like pView Delattre PerfMonitor ... very impressive but monolithic (no specific processes). Delattre TMonitor ... not Win9x. Designed for multiple cores, this is gonna be a good one! Just released, free, found here. Merijn-Soeperman IbProcMan ... much like pView Merijn-PepiMK ProcAlyzer ... could not make this work (Win9x compatible?) EnTech Resource Manager ... sleek and pretty, but of minimal use (for the task at hand) IdeasAsylum TaskMonifier ... this is a .Net Framework 2.x application. Not gonna happen. Alastair L. Wilson MWTask ... really basic Win3x Taskman Folletto Malefico TaskManX ... slightly improved Win9x Taskman InsaneWare Task-Master ... nothing special, uses yellow on white text! un-useable EDIT: currently researching the following: * Igor M. Arsenin TaskInfo * Veign Usage Monitor * infiero Tail XP * WindowsProtector.com WindowsProtector I am open to all suggestions now. If I overlooked something in any of the above applications please correct me. Once again, I would love to find a way to automatically log to a file when a process CPU usage hits a threshold. This would be iron-clad info for the Opera devs to hopefully act upon. Currently I am logging these freezes by hand (BTW, I am seeing some kind of pattern where Opera complains about an MSFN script, see Site/Forum later). Later I'll re-organize this top post into an index of realtime monitor tools for Win9x. So if at all possible, please use ADD REPLY and avoid quoting the current top post (because it will later be editied and thus followups would likely contain orphan quoted material!) thanks!

Yup, and I thank you again Dave (see original post for an edit). Sorry 'bout that, should have replied directly. Instead I used a time machine to respond in advance to your dead-on accurate reply! @Everyone ... It seems they may have got one thing fixed in Opera 10 after all. There was this nasty tendency in Opera 9.xx for some random downloaded files to be defective even when the transfer tab/window pronounced the file as DONE. In the Opera 9.xx in Windows 98 thread see Posts #55 and #62 where I noted it. Having run a few hundred files through Opera 10.10 now, I have yet to find one that wasn't completed properly. Has anyone else seen the download problem at all in Opera 10.xx? All things considered though, I would take the once-in-a-while defective download over these random pauses/freezes!

I dunno. Always wondered about the tweaks to the tools between 98se->ME. They were not simply version string updates like some of the earlier versions (when you look in ExamDIFF there are many differences). I always had this gut feeling that some improvements had been integrated seeing how WinME does have better memory usage and management, but I certainly could be wrong. However, consensus seems to be that things like Defrag contain actual fixes. Anyway I was just wondering so that I could be on the same page so to speak, and try to see the exact same thing as everyone else here.

I just looked at the registries for two Win98se fresh installs. Both were identical with the following ... [HKEY_LOCAL_MACHINE\Software\Microsoft\Windows\CurrentVersion\explorer\FindExtensions\Static] [HKEY_LOCAL_MACHINE\Software\Microsoft\Windows\CurrentVersion\explorer\FindExtensions\Static\WebSearch] @="{07798131-AF23-11d1-9111-00A0C98BA67D}" [HKEY_LOCAL_MACHINE\Software\Microsoft\Windows\CurrentVersion\explorer\FindExtensions\Static\WebSearch\0] @="On the &Internet..." [HKEY_LOCAL_MACHINE\Software\Microsoft\Windows\CurrentVersion\explorer\FindExtensions\Static\WebSearch\0\DefaultIcon] @="C:\\WINDOWS\\SYSTEM\\shdocvw.dll,-111" [HKEY_LOCAL_MACHINE\Software\Microsoft\Windows\CurrentVersion\explorer\FindExtensions\Static\WebSearch\0\HelpText] @="Search the web" [HKEY_LOCAL_MACHINE\Software\Microsoft\Windows\CurrentVersion\explorer\FindExtensions\Static\WabFind] @="{32714800-2E5F-11d0-8B85-00AA0044F941}" [HKEY_LOCAL_MACHINE\Software\Microsoft\Windows\CurrentVersion\explorer\FindExtensions\Static\WabFind\0] @="&People..." [HKEY_LOCAL_MACHINE\Software\Microsoft\Windows\CurrentVersion\explorer\FindExtensions\Static\WabFind\0\DefaultIcon] @="C:\\PROGRAM FILES\\OUTLOOK EXPRESS\\WABFIND.DLL, 0" EDIT: I also took a look at the SYSTEM.1st file found in the root directory (it is binary like SYSTEM.DAT so you need to use something like RegDat), and I found that the FindExtensions 'Static' subkey does not exist in there yet. This tells me that during a fresh install, setup saves the SYSTEM.1st as a bare core registry snapshot (in fact the timestamp shows that this is done just before it writes SUHDLOG.DAT). Later during the install it decorates Windows with things like MSIE and OE and then, the Find Extensions show up. Please export and post here the following key from the machine with "viruses" ... [HKEY_LOCAL_MACHINE\Software\Microsoft\Windows\CurrentVersion\Explorer\FindExtensions\Static]

to Mijzelf, triger49, fortcollins ... What app were you using to see the 100% CPU usage? If it is SYSMON, is the Win98se or the WinME version? to Everyone ... Noticing on Opera 10.10.1893 that for every download I start, a new tab 'Downloads' opens. I could swear that this is new behaviour and that previously all downloads would get thrown into the already existing 'Downloads' tab. The preference setting, if any, eludes me. EDIT: It is found in: Downloads Tab > View > Never Show When Starting Download. Thank you Dave-H. I missed that setting and that thread with the answer. Using my better judgement I shall keep my opinion of this to myself.

You mean you used it just because you had it? I had to wait weeks before finding the right occasion: http://www.msfn.org/board/googel-t138944.html for this : http://www.msfn.org/board/googel-t138944-page-35.html Guilty. <scapegoat>Actually its Opera's fault I tell you! These freezes are driving me mad. I had to get it out there while this Win9x machine was still able to see the internet.</scapegoat> No really, I saw it on some Win7 forum around month ago (maybe Thanksgiving?) and tucked it away. I was laughing so hard. Still am. The last time some tech related idea made me crack up was those TV commercials like 10 years ago for some wholesale computer vendor ***. They would shoot animals out of a cannon. I just can't remember now . ***: Was it Outpost.com? Now this is gonna bug me for a while.

Why being sorry? http://tinyurl.com/ycehmrl jaclaz And with that I must bestow upon you the official and very honorable title of LinkMaster J . I am quite sure this award would pass with unanimous consent. BTW I swear I had never seen that lmgtfy link posted here before (or at Boot-land!). Since the search results say you only did it twice I feel somehow better as well. But I figured if anyone had used it, well, it had to be you. Now I'm off to look for a SomeoneBeatMeToItAgain.com for next time! @cure ... that link wasn't meant to imply you didn't search! I just needed a place to use it.

Note that we're talking about Win9x here of course. Any program that has a Search/Find application called from a DLL can put stuff in there accessible from the Start Menu. I am unaware of any trick to make an EXE work as a Find Extension. These Find Extensions are most easily added by creating subkeys under this key: [HKEY_LOCAL_MACHINE\Software\Microsoft\Windows\CurrentVersion\Explorer\FindExtensions\Static] Typically like this: [HKEY_LOCAL_MACHINE\Software\Microsoft\Windows\CurrentVersion\Explorer\FindExtensions\Static\[b]Program Named Subkey[/b]] @="{[b]Program's CLSID[/b]}" [HKEY_LOCAL_MACHINE\Software\Microsoft\Windows\CurrentVersion\Explorer\FindExtensions\Static\[b]Program Named Subkey[/b]\0] @="[b]Program Name (appears in Start Menu > Find)[/b]" [HKEY_LOCAL_MACHINE\Software\Microsoft\Windows\CurrentVersion\Explorer\FindExtensions\Static\[b]Program Named Subkey[/b]\0\DefaultIcon] @="[b]C:\\Somewhere\\SomeProgram.exe,0[/b]" The actual program is called by CLSID to a DLL though. That EXE there just supplies the icon. So you have to have some more, this is only an example ... [HKEY_LOCAL_MACHINE\Software\Classes\CLSID\{[b]Program's CLSID[/b]}] @="[b]Program Name[/b]" [HKEY_LOCAL_MACHINE\Software\Classes\CLSID\{[b]Program's CLSID[/b]}\InProcServer32] @="[b]C:\\Somewhere\\SomeProgram.dll[/b]" "ThreadingModel"="Apartment" That should be enough info for you to be able to prune out any items that you no longer want to appear under Start > Find. To delete the example I showed above you would delete this key: [HKEY_LOCAL_MACHINE\Software\Microsoft\Windows\CurrentVersion\Explorer\FindExtensions\Static\Program Named Subkey]

Interesting question. It didn't happen on mine (Win98se). I aimed the FIND dialog at C:\Windows and used trans[1].gif as the file to search for (I first made sure it was actually in there). The FIND returned all three copies that exist. Maybe the folder attributes need to be changed. None of mine are H (Hidden). File attributes for C:\Windows\Temporary Internet Files is SD File attributes for C:\Windows\Temporary Internet Files\Content.IE5 is SD File attributes for C:\Windows\Temporary Internet Files\Content.IE5\GRCX8XM1 is SD File attributes for C:\Windows\Temporary Internet Files\Content.IE5\WXEZ81MR is SD File attributes for C:\Windows\Temporary Internet Files\Content.IE5\YNW0T9DW is SD NOTE: S==System D==Directory Also note that I have the Start > Settings > Folder Options > View > Hidden Files set to Show all files.

I've had some time the last few days to dabble on Win9x and here I am in Opera 10 trying to figure out what they did to cause these interminable freezes. Most recent un-successful fix was killing the built-in spellcheck. So after you replied as above I had to go find out if I spelled it wrong! Its ok! And thanks for the award! But my real award-winning comments are reserved for emails to Opera developers. They would get auto-censored here and banned over there. Exactly right. And I can't wait til someone gets freezes/pauses on one of these (a picture from Opera website showing some cellphones and other toys they support): I just had to quote that statement. It is clear and concise and conveys exactly what I am thinking (minus the boiling anger that occurs every few minutes). Well said. You know one of the weird things about that is the fact that when Coke was about to change (remember that the consumers had plenty of fore-warning), people went out and hoarded the old stuff for personal consumption and because it may become a collector item, it pretty much sold out. Then, the new stuff came and most people had to try it, some again thinking it would be a collector item and probably sold out. Then the Classic comes out in replacement and sells more. It leaves you wondering if the whole thing wasn't a giant success anyway For a while there I thought Microsoft would employ the same trick with WinXP. They certainly succeeded in getting people to buy it before the cut-off deadline. Their mistake was not re-introducing the cash-cow WinXP (probably as MCE) and offering it for sale in perpetuity. That was a real marketing blunder! Instead, they appear to have bet it all on Windows Mojave, umm I mean 7. P.S. Reminder to all: no spell check on this post! Be kind to me.

These kind of markers can be placed in there by any program, Typically they are from shareware that requires registration of some kind and that value is treated as a flag to some validating function. I just checked my own gigantic collection of registry logs covering hundreds of computers back to Win95 and found no occurence of that exact string under that key. BTW: how come your HKEY_LOCAL_MACHINE/Software/Microsoft/Windows/CurrentVersion has forward slashes? It can only be HKEY_LOCAL_MACHINE\Microsoft\Windows\CurrentVersion. Anyway, I've been waiting for a chance to use this! Let Me Google That For You! Quick results show Foxit Software applications at the top of the list. P.S. to all ... sorry if someone else used that http://lmgtfy.com already! It's freakin hillarious to me.

UPDATE TO POST #1 The top post has been re-written to describe the current ZIP package suite: 2009-12-01. (downloaded on 2009-12-18). Win9x fans will notice no functional changes except for the fact that both FILEMON and REGMON are no longer distributed in the Suite. -CTH-

@Spitfire Is that the normal (MSI/Installshield) installer or the Classic one? @lightning slinger Having read through their comments I get the feeling the devs are just going to keep shoehorning new features endlessly without regard to existing problems, Win9x will probably be sacrificed to the gods of advancement. I only saw one comment from someone asking for bug fixes, everyone else acting sycophantic. This sounds like slow-motion software suicide. I am tempted to post the following: Devs: I got an idea, since we have like a 2% market share let's reduce it further by cutting the legs off everyone using Windows below Win2K. And we can revisit that next year and kill off Win2k and WinXP as well so everyone will buy Vista I mean 7. If we can do that maybe Microsoft will give Opera a better position on the browser selection screen! Then we'll really be in the bigtime.

I don't see it. Could you post the links you used for the download and for the OS support discussion. Thanks.

I'm glad you asked (!), because I pulled the INF from the Nusb33 installer just to be sure and I just noticed the exact same error in Nusb33 also. So, I was originally Windiffing against my corrected one. Anyway, here it is (the same line in both Nusb33 and Nusb33e) ... [Version] Signature="$CHICAGO$" Class=USB ClassGUID={36FC9E60-C465-11CF-8056-444553540000} provider=%MSFT% LayoutFile=LAYOUT.INF,LAYOUT1.INF,LAYOUT2.INF DriverVer=07/01/2001,5.1.2600.0 ;Modified - 06/01/2006, 01.01 ;Maximus Decim modified - 06/14/2006, 01.02 ;Maximus Decim modified - 08/31/2006, 01.03 add iPod, fix Sony Digital Camera ;Maximus Decim modified - 09/02/2006, 01.04 add LayoutFile ;Maximus Decim modified - 25/01/2007, 01.05 add wdmstub, remove USB FDD, add CnMemory ;Maximus Decim modified - 07/03/2007, 01.06 remove wdmstub, CnMemore replace Skymedi ;Maximus Decim modified - 28/10/2007, 01.07 add USB FDD (for DISKTSD.VXD 4.90.3001) ;Maximus Decim modified - 12/11/2007, 01.08 add EKE USB Floppy (VID_0424&PID_0fdc), [color="#FF0000"] add USBCompliance (for usbu2a.sys)[/color] [ControlFlags] ExcludeFromSelect = * [color="#FF0000"](everything else snipped)[/color] It should look like this (added semicolon and CRLF) ... [Version] Signature="$CHICAGO$" Class=USB ClassGUID={36FC9E60-C465-11CF-8056-444553540000} provider=%MSFT% LayoutFile=LAYOUT.INF,LAYOUT1.INF,LAYOUT2.INF DriverVer=07/01/2001,5.1.2600.0 ;Modified - 06/01/2006, 01.01 ;Maximus Decim modified - 06/14/2006, 01.02 ;Maximus Decim modified - 08/31/2006, 01.03 add iPod, fix Sony Digital Camera ;Maximus Decim modified - 09/02/2006, 01.04 add LayoutFile ;Maximus Decim modified - 25/01/2007, 01.05 add wdmstub, remove USB FDD, add CnMemory ;Maximus Decim modified - 07/03/2007, 01.06 remove wdmstub, CnMemore replace Skymedi ;Maximus Decim modified - 28/10/2007, 01.07 add USB FDD (for DISKTSD.VXD 4.90.3001) ;Maximus Decim modified - 12/11/2007, 01.08 add EKE USB Floppy (VID_0424&PID_0fdc), [color="#008800"]; add USBCompliance (for usbu2a.sys)[/color] [ControlFlags] ExcludeFromSelect = * [color="#008800"](everything else snipped)[/color] Like I said, there are also trailing spaces on a few lines that Windiff picks up. They were removed automatically from my fixed copy by UltraEdit so Windiff easily spotted them. I don't think they matter in INF files but might make a difference in something LFN friendly. Thanks! EDIT: just to clarify, after more research my Nusb33 and Nusb33e are identical after all. I guess the "e" just means English (doh!). So I in fact had the latest release all along. Hope I didn't confuse anyone. For the record, the latest release (newest file) appears to be dated 2007-11-16.

@maximus-decim ... I just Windiffed the latest Nusb (33e) to the previous I had installed (33) and found one minor difference in the USBSTOR.INF: see line 16 for an uncommented comment. Not sure if it breaks the INF or not. There are also a bunch of trailing spaces here and there which mean nothing (except to Windiff of course ;-) @Everyone ... Does anyone know if there is a Hoiby-like patch for the EXPLORER.EXE that comes in this NUSB distro (version is 4.72.3612.1700)?

I just want to confirm that the observations posted here previously by fortcollins appear to be dead-on accurate. The random pauses and freezes are becoming a serious issue. And, like most Opera bugs this one is not step-by-step reproducible. All throughout the 9.xx series I only noticed similar random freezes when saving MHT files, non-reproducible, perhaps one time out of every 20 saves. The pause/freeze paralyzes all Opera instances. This behavior still exists in the 10.xx branch of course. What has changed for the worse is that on the 10.xx branch the random pause/freeze occurs no matter what you happen to be doing. In fact it has happened (*) while I was typing this very post. All three versions in 10.xx have actually gotten worse! It smells like a memory management issue to me, but as has been stated here already the culprit is likely related to the disk cache. Regardless, as far as I'm concerned its virtual memory management is FUBAR. I will try out a few more 10.xx iterations before resurrecting 9.2x or 9.6x. IMHO, it is possible that the developers are making the classic mistake of testing on multi-core multi-GB RAM systems where the bugs get easily masked by super-fast hardware. And if this is the case, we are witnessing the early stages of the Opera post-Win9x era. Smart programmers test their compiled binaries on the slowest machines they can find. Stupid programmers never do. My suggestion for Opera users here at MSFN is to get involved by signing up and opening threads over there! I believe it is now or never. (*) freezes during this single post = 1 2 3 4 times in perhaps 10 minutes.

Doh! You're right! I never clicked that Fast Reply button before. Ok, I'm in the Fast Reply right now and see what you mean, no special items. Here I'll try something: Wrap in code tags ... Testing123Testing123Testing123Testing123Testing123Testing123Testing123Testing123Testing123T esting123Testing123Testing123Testing123Testing123Testing123Testing123Testing123Testing123Testing123T e sting123 This is Opera 10 with Identify as Internet Explorer selected. I'll edit this post to describe any width issues I see. P.S. I just noticed that clicking on More Options kicks the editor into the Full mode. UPDATE: right margin did become broken. This code was generated: Wrap in code tags ... <div class='codemain'> Wrap in quote tags ... <div class='quotemain'> Manually added codebox tags ... <div class='codemain' style='height:200px;white-space:pre;overflow:auto'>

You're right about that. I was just playing it safe is all (expect the worst hope for the best, etc). My first commandment is to protect the data. LOL ... no sarcasm really! I love FOR loops especially nested ones and that one looks to me like it will work. I just wanted to avoid the multiple dialogs about overwrite. In that second example I just knocked off one copy operation. There was definitely nothing elegant or efficient about what I posted, it was merely functional and lossless. When I have a backup present (and I always do myself, but clients never seem to) then I go for beauty and elegance! @horus, haven't used RichCopy but from that page you posted it sure looks like all the options you will need are in place. If I were you I might just back up that whole folder structure into a ZIP and then experiment away!

You may have noticed that sometimes an extra wide chunk of code forces the post that it appears in to stretch past the right side margin (which in turn causes horizontal scrollbars to appear in the main browser window). Other times it does not do this and instead the CodeBox it appears in has its own horizontal scrollbars. This keeps the entire thread and all its posts to stay within the margin as it should, with no horizontal scrollbars in the main browser window. We kicked this around for a while in this thread. Some posters saw the problem, but I could not. But it just appeared to me in this thread on Post#12 in the Batch code. So I saved the page (CSS is embeded within <STYLE> tags), looked around and experimented. The bad code (right margin pushing): this is how the HTML looks: <div class='codemain'>STUFF</div> The good code (no right margin problem): occurs if I use the Insert Special Item > CodeBox, and the following HTML/CSS is generated: <div class='codemain' style='height:200px;white-space:pre;overflow:auto'>STUFF</div> Now I am not sure if others are also selecting the same thing: Insert Special Item > CodeBox, but the fact remains that sometimes the inline CSS style= gets added and sometimes not. Anyway, to fix those extra wide posts, this is the missing ingredient within that <DIV that needs to present: <div class='codemain' style='height:200px;white-space:pre;overflow:auto'> I tested and verified this by adding the bolded section to the saved page in that Post#12 batch code. The result was that all posts were identical in width and properly within the margin. The browser (Opera 10) window scrollbars went away. The strange thing is, I do not see Code anywhere as an option, just CodeBox. So I fail to see how it gets entered anyway. I hope I explained it clearly!