--------------- FIDO MESSAGE AREA==> TOPIC: 129 OP. SYS DEBATE Ref: E5E00005Date: 05/08/97 From: JACK STEIN Time: 09:49pm \/To: KEITH DOUGLAS (Read 2 times) Subj: Which is the best? Keith Douglas wrote in a message to Jack Stein: JS> 4DOS>e:\OPUS>start /? JS> Purpose: Start a program in another session. KD> Could it be that start is a 4DOS/4OS2 command? Yes, and it's worked here since the day I installed OS/2. Just one of many wise choices I've made in choosing which software blesses my computer. MS junk stays off as much as possible and GUI bloat is used only where eded. Of course 4OS2 and 4DOS can't be good, since few to no retailers install it on computers bought at Wal-Mart, and we all know 100 billion flies can't be wrong:-) Nothing beats a Packard Bell, WINDOSE and Command.com. Jack --- timEd/2-B11 * Origin: Jack's Free Lunch 4OS2 USR16.8 Pgh Pa (412)492-0822 (1:129/171) --------------- FIDO MESSAGE AREA==> TOPIC: 129 OP. SYS DEBATE Ref: E5E00006Date: 05/08/97 From: DENIS TONN Time: 03:46pm \/To: ERIC JONES (Read 2 times) Subj: Which is the best? Original from Eric Jones to Mack Barss on 05-08-1997 Original Subject: Which is the best? --------------------------------------- EJ> Anyway, as if it mattered, according to HELP's description of EJ> "seperate session" if you use the "add programs" function, OS/2 will EJ> automatically determine by itself if the windows program needs "seperate EJ> session" and act accordingly. So then, being in the shared WIN session EJ> is only the default if you add the program manually, or OS/2 decides the EJ> program doesn't need to have it's own seperate session. The way that OS/2 will "automaticly determine" the session type is from a file \OS2\INSTALL\DATABASE.DAT. There is a text version of this file \OS2\INSTALL\DATABASE.TXT where you can add/change any setting you wish and then "compile" to DATABASE.DAT with PARSEDB.EXE. A quick review of database.txt shows that the vast majority of windows sessions are defined to be run from a common win-os2 session. In fact, of all the apps listed, only Excel, PCS3270, and WordPerfect are defined to run in separate sessions. This might be because of the "critical" nature of these apps rather than any specific requirement of these apps to be run in separate sessions though. The user can still overide the default system setting and run them in a common session if they wish. "By default" most of the windows applications are run in a common session, the separate session setting is the one that is unusual. Also, the default session setting can be defined in the WIN-OS2 setup icon. "By default" the installation routine will set this up for common sessions. While Scott is correct that OS/2 offers the possibility to run Windows apps in a separate session (which Win95 does not), you are quite wrong in stating that this is the prefered or "default" way of running WinOS2 sessions. Denis All opinions are my very own, IBM has no claim upon them . . --- Maximus/2 3.01 * Origin: T-Board - (604) 277-4574 (1:153/908) --------------- FIDO MESSAGE AREA==> TOPIC: 129 OP. SYS DEBATE Ref: E5E00007Date: 05/08/97 From: ROGIER MEURS Time: 01:22am \/To: DARRELL SALTER (Read 2 times) Subj: Lemming-Ware Hi Darrell! Darrell Salter wrote to Rogier Meurs (Monday May 05 1997 at 13:14) RM>> This is no problem in Win95. In OS/2, it is, because of the RM>> synchronous single input queue. This is not about running several RM>> (socket) apps at the same time (which Win95 does without a problem). RM>> This is about starting them up. When an app grabs the queue in OS/2, RM>> your mouse-input is gone. In Win95, it isn't, because every 32-bit RM>> app has it's own queue. In OS/2, all PM apps share the same queue. RM>> When one app takes hold of it, you're out of control. DS> I think I understand what you're saying. It is my understanding that DS> this is the fault of the program and not the operating system. You're partially right. The app _is_ at fault when it grabs the queue. It simply shouldn't. But: when the OS/shell _allows_ an app to block your input, the OS/shell is at fault too. One can never expect all software to be perfect, therefore OS's (usually) have inbuilt things to prevent an app to hang the system/block the input. In the case of Win95 vs OS/2, Win95 is clearly the better queuer (where BTW OS/2 is the better memory protector etc.). DS> I don't think the app is supposed to monopolize the queue. In any DS> event, I believe FP17 allows for a solution to this problem, by not DS> allowing an app to dominate the queue in this manner. But it's still a hack. OS/2 5.0 is expected to really solve it. The problem is, a rewriting of this part of PM breaks backward compatibility with OS/2 applications. This is no fun of course. And since the app availability on OS/2 isn't _that_ great, I don't think IBM is anxious to change it. Otherwise they would have really fixed it in v3.0. It's not like they can't fix it. Still, it's a problem that's to be taken into account and not to be ignored. Rogier http://lelystad.flnet.nl/~0meurs01/ --- FMail 1.20 Registered * Origin: File & Mailboard AD FUNDUM o32o-2821o4 Do-Ut-DeZ (2:2801/506) --------------- FIDO MESSAGE AREA==> TOPIC: 129 OP. SYS DEBATE Ref: E5E00008Date: 05/10/97 From: DARRELL SALTER Time: 12:00pm \/To: ROGIER MEURS (Read 2 times) Subj: Lemming-Ware Rogier Meurs wrote in a message to Darrell Salter: RM> Hi Darrell! Hey Rogier! RM> One can never expect all software to be perfect, therefore OS's RM> (usually) have inbuilt things to prevent an app to hang the RM> system/block the input. Agreed. DS> I don't think the app is supposed to monopolize the queue. In any DS> event, I believe FP17 allows for a solution to this problem, by not DS> allowing an app to dominate the queue in this manner. RM> But it's still a hack. OS/2 5.0 is expected to really solve it. RM> The problem is, a rewriting of this part of PM breaks backward RM> compatibility with OS/2 applications. This is no fun of course. And RM> since the app availability on OS/2 isn't _that_ great, I don't think RM> IBM is anxious to change it. Otherwise they would have really fixed RM> it in v3.0. It's not like they can't fix it. Still, it's a problem RM> that's to be taken into account and not to be ignored. This is the age-old problem of any operating system, trying to advance while maintaining backward compatability. A nasty tradeoff. Thank you for your comments. Cheers, Darrell darrells@idirect.com Yeah, I'm WARPed! ... Mafia DOS: "thisa you lasta chance [Y/N]?" --- * Origin: Sparky's - Ok! Not perfect, but... (1:229/1) --------------- FIDO MESSAGE AREA==> TOPIC: 129 OP. SYS DEBATE Ref: E5E00009Date: 05/09/97 From: BRIAN MURRAY Time: 08:17pm \/To: SCOTT LITTLE (Read 2 times) Subj: Which is the best? * In a message originally to Darrell Salter, Scott Little wrote: > DS> He is a die-hard WinWeenie > Oh, ye with an IQ of 5... must you resort to such childish > 'insults'? That rates right up there with 'Be nice, pinhead.' --- * Origin: Fiat Lux - Mostly Harmless Networking (1:209/345.100) --------------- FIDO MESSAGE AREA==> TOPIC: 129 OP. SYS DEBATE Ref: E5E00010Date: 05/10/97 From: DAVID BOWERMAN Time: 10:00am \/To: DARRELL SALTER (Read 2 times) Subj: Which is the best? Darrell Salter wrote in a message to Scott Little: SL> His name is Mack for a start, and second is IS an IFS because I put SL> the damn thing in the CONFIG.SYS myself. DS> Getting a little tense Scott? Is it too difficult for you to keep DS> track of all your tall tales? (My apologies to Mack). Darrell, the SUBSTP program to supply the equivalent to SUBST for use with OS/2 command line and and WPS sessions does install as an IFS. There are other ways to hook it but that (IMHO) is the simplest and easiest way with the minimum possibility of causing other problems. Regards, David --- timEd/2 1.10+ * Origin: Frog Hollow -- a scenic backroad off the Infobahn (1:153/290) --------------- FIDO MESSAGE AREA==> TOPIC: 129 OP. SYS DEBATE Ref: E5F00000Date: 05/10/97 From: ERIC JONES Time: 08:51pm \/To: DENIS TONN (Read 2 times) Subj: Which is the best? *NOTE: IceEdit 2.30 [Eval] -=> Quoting Denis Tonn to Eric Jones <=- EJ> Anyway, as if it mattered, according to HELP's description of EJ> "seperate session" if you use the "add programs" function, OS/2 will EJ> automatically determine by itself if the windows program needs "seperate DT> The way that OS/2 will "automaticly determine" the session type is DT> from a file \OS2\INSTALL\DATABASE.DAT. There is a text version of this DT> file \OS2\INSTALL\DATABASE.TXT where you can add/change any setting DT> you wish and then "compile" to DATABASE.DAT with PARSEDB.EXE. DT> A quick review of database.txt shows that the vast majority of DT> windows sessions are defined to be run from a common win-os2 session. DT> In fact, of all the apps listed, only Excel, PCS3270, and WordPerfect DT> are defined to run in separate sessions. This might be because of the DT> "critical" nature of these apps rather than any specific requirement DT> of these apps to be run in separate sessions though. The user can DT> still overide the default system setting and run them in a common DT> session if they wish. DT> "By default" most of the windows applications are run in a common DT> session, the separate session setting is the one that is unusual. DT> Also, the default session setting can be defined in the WIN-OS2 setup DT> icon. "By default" the installation routine will set this up for DT> common sessions. That's what I've been saying! EJ> session" and act accordingly. So then, being in the shared WIN session EJ> is only the default if you add the program manually, or OS/2 decides the EJ> program doesn't need to have it's own seperate session. DT> While Scott is correct that OS/2 offers the possibility to run DT> Windows apps in a separate session (which Win95 does not), you are DT> quite wrong in stating that this is the prefered or "default" way of DT> running WinOS2 sessions. Excuse me? I never said putting win programs into seperate session was either prefered or default! Please have the coutesy of reading my message _before_ replying! ... OS/2 - better Windows than Windows, better DOS than DOS. ~~~ TGWave v1.20 Beta-07+ --- Telegard v3.02/mL * Origin: Courier Central \ Cashion, OK \ 405.433.2665 (1:147/92) --------------- FIDO MESSAGE AREA==> TOPIC: 129 OP. SYS DEBATE Ref: E5F00001Date: 05/10/97 From: KEITH DOUGLAS Time: 01:03pm \/To: ALL (Read 2 times) Subj: NT 'Security' From BugTraq: (NT's bug-issues mailing list) It is possible to remotely cause denial of service to any windows 95/NT user. It is done by sending OOB [Out Of Band] data to an established connection you have with a windows user. NetBIOS [139] seems to be the most effective since this is a part of windows. Apparently windows doesn't know how to handle OOB, so it panics and crazy things happen. I have heard reports of everything from windows dropping carrier to the entire screen turning white. Windows also sometimes has trouble handling anything on a network at all after an attack like this. A reboot fixes whatever damage this causes. This has been tested. It does what the above says. It can be blocked under NT 4.0 by denying local file sharing or something useful like that, but you know... --- FMail/386 1.22 * Origin: The Chrono Zone (514)363-6298 Lasalle, QC, Canada (1:167/310) --------------- FIDO MESSAGE AREA==> TOPIC: 129 OP. SYS DEBATE Ref: E5F00002Date: 05/10/97 From: KEITH DOUGLAS Time: 01:05pm \/To: MACK BARSS (Read 2 times) Subj: Relocating MB> for re-location to the US. I have accepted a position in MB> Minneapolis, MN MB> and the bbs will be relocated there. I hope to debate with Hope the new job goes well. See you soon. --- FMail/386 1.22 * Origin: The Chrono Zone (514)363-6298 Lasalle, QC, Canada (1:167/310)