--------------- FIDO MESSAGE AREA==> TOPIC: 251 OS2 BBS Ref: EDE00007 Date: 09/09/97 From: ROGER NELSON Time: 07:10am \/To: BILL KIRBY (Read 0 times) Subj: Adept and JAM Bill Kirby wrote in a message to Roger Nelson: BK> Thx Roger, I guess what I'm really trying to say. Is it's not BK> worth it! I didn't think it would be, although I toyed with the idea of using another mailer with Adept and then thought, why bother? BK> And most of all! YOU would not like it! I promise YOU ! BK> Am I going to waste my time Testing it...... NOPE ! I believe you. BK> Am I going to waste my time Testing Squish.. Yep ! Easy? Yep! BK> Does it work? Yep ! And works very well. BK> If all I need to make AllFix work is a NetMail *.MSG, I'm in BK> business ! I've seen Squish work (very limited) with Adept everyday BK> and I can promise you, when I use LINK, SQPACK, and ADD Allfix I BK> won't like it and the TIME is gonna become an issue! Having a *.MSG NETmail area just to use Allfix doesn't seem to be worth it, but I haven't used Allfix in years, so I don't remember using it for anything other than a file TIC processor. BK> Am I gonna waste my time testing Jam... Nope ! Me either. Regards, Roger --- timEd/2 1.10+ * Origin: NCS BBS - Mandeville, LA - (504) 626-7968 (1:3828/7) --------------- FIDO MESSAGE AREA==> TOPIC: 251 OS2 BBS Ref: EDE00008 Date: 09/09/97 From: ROGER NELSON Time: 10:56am \/To: RON MITCHELL (Read 0 times) Subj: Latest Adept stuff Ron Mitchell wrote in a message to Bill Kirby: RM> I've got about 75 megs of Adept stuff in the file area on my board, RM> and really need to regain some space. What can I safely delete and RM> still leave a full install for anyone that comes along and wishes RM> to try Adept? If you wish to keep a full install, grab the latest one. The full install begins with the letters "Adf." The upgrades begin with the letters "Adu." RM> I don't really understand all the numbering schemes and different RM> versions that are here. :) I think what I said (above) should help in the confusion. As far as the numbers in the various releases go, that is for the programmers alone to decide and while it may seem confusing to us, it makes perfect sense to them. It's sort of like a paper trail so that they can keep track of what they did in any given release. Regards, Roger --- timEd/2 1.10+ * Origin: NCS BBS - Mandeville, LA - (504) 626-7968 (1:3828/7) --------------- FIDO MESSAGE AREA==> TOPIC: 251 OS2 BBS Ref: EDE00009 Date: 09/09/97 From: EDFEL RIVERA Time: 07:17pm \/To: MURRAY WEISMER (Read 0 times) Subj: Re: ANSI EDITOR (OS/2) Hi. Well, I am trying to prepare my BBS for a comeback. I am still painting Menus, Hello screens and setting menu commands. Have not worked with QWK yet. BTW, I am using Ansi Artist to paint. It is DOS but it reads Wildcat @ codes. In the mean time, it could be excellent... Are you being able to do Local Downloads and Uploads with Adept? It is now my only concern. Regards! Edfel Rivera --- AdeptXBBS v1.11y (FREEWare/2) * Origin: SnoValley Shareware Exchange (1:343/108) --------------- FIDO MESSAGE AREA==> TOPIC: 251 OS2 BBS Ref: EDE00010 Date: 09/09/97 From: BILL KIRBY Time: 08:00pm \/To: RON MITCHELL (Read 0 times) Subj: Latest Adept stuff Hello Ron. 08 Sep 97 20:14, Ron Mitchell wrote to Bill Kirby: RM> Bill Kirby wrote in a message to All: BK>> DL the latest files: BK>> With the blessing of SafeHex. The following is now avail... BK>> 1. adu111xb7.zip = several fixes BK>> 2. gk23011.zip = Tic fixes RM> I've got about 75 megs of Adept stuff in the file area on my board, RM> and really need to regain some space. What can I safely delete and RM> still leave a full install for anyone that comes along and wishes to RM> try Adept? I have on my system: adf111h.zip adu111x.zip adu111yb8.zip I don't need any thing else..... However, the next file should be: adf111y.zip adu111y.zip And thats all I will need..... RM> I don't really understand all the numbering schemes RM> and different versions that are here. :) ....) Not many other people do either....) They are really only FIXs and or ADDs. Not to worry........ RM> L8r, Ron RM> ... On the other hand, you also have 5 fingers. Wrong.... 4 1/2. The Motorcycle has 1/2. Bill ... -= DownLoad Adept XBBS files @ http://www.prismnet.com/~bkirby =- --- Adept/2 XBBS adu111Yb8 (VAC)24 Node (FREEware)-GoldEd3.00a5+ UNREG * Origin: Twin Towers BBS, Austin Tx.- www.prismnet.com/~bkirby (1:382/65) --------------- FIDO MESSAGE AREA==> TOPIC: 251 OS2 BBS Ref: EDE00011 Date: 09/09/97 From: BILL KIRBY Time: 08:16pm \/To: JIM RYSYK (Read 0 times) Subj: LATEST ADEPT STUFF Hello Jim. 09 Sep 97 09:34, Jim Rysyk wrote to Ron Mitchell: >> I've got about 75 megs of Adept stuff in the file area on my board, >> and really need to regain some space. What can I safely delete and >> still leave a full install for anyone that comes along and wishes >> to try Adept? Jim, you are going to make me toss and tumble ALL night, trying to figure out what the following is all about !!!! Roger, Please just in here and explain this...... JR> You could try to use a file processor that basically uses the REPLACE JR> field to erase the old version. That is, if the hatched .TIC file JR> has the name of the file it replaces in the .TIC. huh? JR> eg, the following sample is from my TICK ver 2.10 config ;) from the JR> DOS days. It seems my OS/2 Warp 4 system does not care, and still JR> executes the toss command from either a DOS or OS/2 Session: What am I missing ??? My Goad, I hope I never have to do this ! JR> IN C:\TBBS\NM\FILE\ JR> ZONE 1 C:\TBBS\NM\OUTBOUND JR> NET 2424 JR> NODE 242 JR> HOLD C:\TBBS\NM\HOLDDIR\ JR> QDIR C:\TBBS\NM\QDIR\ JR> REPLACE JR> ListFmt %3 %3 %1:-40.1 JR> AKA 20:416/100 JR> AKA 49:945/52 JR> AKA 60:4900/1 JR> AKA 81:416/247 JR> AKA 94:4160/2 JR> AKA 111:1416/1 JR> AKA 182:235/2305 JR> AKA 201:938/103 JR> RAID LOG >>> C:\F\LOG\logt-5.txt JR> AREA E:\F\LIST FLY_NET JR> Local ListName E:\F\LIST\LISTS.DIR JR> LOCAL REPLACE JR> 1:2424/15 password *& JR> AREA E:\F\Fido UCN_NA JR> Local ListName E:\F\Fido\BackBone.DIR JR> LOCAL REPLACE JR> 1:2424/11 password *& JR> The LOCAL REPLACE reads the REPLACE feilds to erase the old file that JR> got replaced. The BBS itself erases the files.bbs description if the JR> file does not exsist. JR> I thought all BBS operated that way for most of this decade? JR> -+- FLAME v1.1 JR> + Origin: user@able.gryn.org, Jobs Search 416-604-1221 JR> (1:2424/242) Bill ... -= DownLoad Adept XBBS files @ http://www.prismnet.com/~bkirby =- --- Adept/2 XBBS adu111Yb7 (VAC)24 Node (FREEware)-GoldEd3.00a5+ UNREG * Origin: Twin Towers BBS, Austin Tx.- www.prismnet.com/~bkirby (1:382/65) --------------- FIDO MESSAGE AREA==> TOPIC: 251 OS2 BBS Ref: EDE00012 Date: 09/09/97 From: DARIN MCBRIDE Time: 06:56pm \/To: JIM RYSYK (Read 0 times) Subj: SUBJECT:LOOKING FOR OS/2 > You have no clue of what I speak, do you. If something GOES WRONG. > i.e., A CRASH. JR> Perhaps. JR> If? JR> If something goes wrong? JR> Why would something go wrong? If there is a BUG in the software? It's called fault-tolerance. No software is bugfree. You need to handle faults. Even all-out failures - the unrecoverable faults - need to be recoverable. > > Hmmm... that multi-process under DOS can get expensive quickly. > > CPU usage on, say, my 486/66: > > 1. ~10% per active node (user online, downloading), > ~3% per inactive node (Binkley monitoring semaphore path, etc.) > 2. ~10% per active node (user online, downloading), > ~3% per inactive node (Binkley monitoring semaphore path, etc.) > 3. ~20% per active node (user online, downloading), > ~3% per inactive node (Binkley monitoring semaphore path, etc.) > 4. ~10% per active node (user online, downloading), > ~3% per inactive node (Binkley monitoring semaphore path, etc.) > > No real advantage yet, although #3 is looking worse and worse. JR> Not with your grunged figures. Where did you get the JR> impression to post that info for #4? Because I'm on a 486/66, not a Pentium. JR> Anyhow, like I said, I'm runing a private comercial JR> version of Binkley called TIMS. Your 10% per node, 3% JR> per inactive node statments are not factual. At least JR> for TIMS. I've not run freeware/shareware binkley for JR> many years, than only ran it on the remote N252 system JR> and never change its config for some years. Perhaps it uses less on your system. Perhaps you can get away with 1%+1% per inactive node (there is always overhead for each node, at least if you want a stable implementation). > Programming complexity: > > 1. Low > 2. Mostly low (because of threads) > 3. Low > 4. High (because of lack of threads - have to write your own thread > mechanism) > > Woops - #4 just jumped. Not that big of a deal for the end-user > since they don't have to write it. But remember - the higher the > complexity, the more likely they will have bugs. Bugs that will > CRASH the system. And higher complexity means higher cost, > generally. JR> Now I repeat, where did you get your information to post this for #4? I design software for a living. I know that it is a MUCH higher complexity to write a finite-state machine for nearly infinite states. A FSM that extends itself on the fly. Very difficult. It is generally easier to write the context switching and let the CPU handle your states. JR> Have you ever run the said TBBS muti line package using TIMS? Have you JR> ever visited a system runing same, and tryed to crash JR> it, calling back into it, while watching the BBS consul JR> to see what happens? Find a local system, and ask the JR> sysop if you can come over and try to do same. Then JR> report back on your sucess. You might even find a JR> sysop who allows your to monitor the consul remotly, JR> whail allowing you to try to crash it on a 2nd JR> alternative line. I know I offered to do this locally. JR> So I imaging other have also offered to do the same. Just because TBBS does it does NOT mean it is easy to do. We're not talking about a finished product - you complained about the DESIGN. The design says that one person implementing this BBS package needs to do it in a robust way. Option #4 does not give that to him. Perhaps a team of 5 or so programmers could do it - but there is complexity in numbers, and a definite cost, too! For a freeware package, one process per node/line is the best design. And since there is ZERO benefit to going to single-process, multi-line, I would do it in a commercial setting as well! JR> To program for a 96 line system it could be a short as JR> 10 lines of the sample JR> config. Feilds like SYSOP, NODE, FILES, PACKETS, MAIL, JR> and OUTBOUND are straight forware English words, easy "Configuration" IS NOT PROGRAMMING. I'm talking about writing the BBS package itself! The Pascal, C, C++, Modula III, Ada, Assembly, or whatever they used to write the BBS in. THAT is the complicated part. Any nimrod programmer can make their configuration files "easy to use." The hard part is writing the underlying logic to handle context-switching, serial access, inter-context communications, and everything else. > Fault-tolerance. i.e., if something crashes, how much of the system > will "bounce back". > > 1. Node dies. Batch file catches it, restarts. User is logged off. > No other node is affected. > 2. Node dies. Usually takes down entire BBS. All users logged off. > All nodes affected. If there is no batch file running it (no need, > right?), nothing is caught, and system stays down. Usually a > "launcher" process is used to simulate the batch file to restart the > system. > 3. Node dies. Batch file catches it, restarts. User is logged off. > No other node is affected. > 4. Node dies. Usually takes down entire BBS. All users logged off. > All nodes affected. If there is no batch file running it (no need, > right?), nothing is caught, and system stays down. I doubt there is > a "launcher" process because these get expensive under DOS. JR> Dies? Batch files? If Line 4 dies? JR> I've operated a 2nd TBBS system remotly for several JR> years. Its operated without any human onsite JR> assistance more or less for all of its life. Only JR> hardware failure ever brought it down. Congrats. That has ZERO to do with the design, though. JR> This system of mine here in TO also has to operate JR> without any human intervention for up to three months JR> at a time. Why is it you think its normal for a system JR> to lock up, or crash? It that a Maximum special JR> feature? Or is that special Sysop added feature you JR> added to your system? ;) My system operated without human intervention for months, too. Any system problem (mostly hardware that didn't want to co-operate - my Zoom modem talking to I think a GVC modem, when I had Binkley 2.6 standard edition and he had Platinum XPress for Win32, this would crash Binkley because Platinum XPress was told by FTSC to do something that was illegal when Binkley 2.6 was released) that happened was automagically taken care of by the EXACT SAME LOGIC in my batch files that was required for normal operation. It was a "feature" imposed on us by FTSC. Binkley 2.6 XE fixes the problem, as does the newer version of Platinum XPress. But when the problem WAS there, this "bad design" you complain about kept the system going. JR> READ MY LIPS. For what its worth, I've operated my JR> other system (in net 252) remotly for years. Big deal. Nothing to do with the design. JR> If one line locks up, it does not affect the other line or user. It does Locking up? Minor detail. CRASHING is the problem in fault tolerance. JR> Software problem. I discovered it was a Windoz 95 user JR> who was one causing a line/user to lock up his session, Generally this is a hardware problem between your modem and his... OS has nothing to do with it. JR> BBS software to triger the INI string on this system. JR> I can't speak for other BBS packages, or even other JR> TBBS systems, or how anyone else configures their modem JR> strings. Purely a modem initialization/hardware problem, nothing to do with design. From a software engineering perspective, there is NO reason to go with a single-process multi-line system in a 32-bit multi-threaded OS, such as WinNT or OS/2. Absolutely zero! In DOS, for performance reasons, you may wish to do so - being able to run 32 nodes (+1 local) on a 386 requires this. However, that design consideration is only made by commercial system who have the manpower to tackle the complications inherent in such a system. And, since DOS is an unsupported platform (i.e., its author doesn't support it anymore), that's not much of an issue, either, is it. --- Maximus/2 3.01 * Origin: Tanktalus' Tower BBS (1:250/102) --------------- FIDO MESSAGE AREA==> TOPIC: 251 OS2 BBS Ref: EDE00013 Date: 09/06/97 From: JASON ANDREWARTHA Time: 01:54am \/To: ALL (Read 0 times) Subj: CRC ERRORS during download with SIO Some people may recall a problem i asked about a few months back regarding haveing crc errors when downloading under warp 4.0 with SIO running yet i've never had them under OS/2 2.1 or 3.0 well i've solved the problem and this might help some others who have had the same problem (i have tried Fix pack 3 and same problem) just take IBM1S506.ADD from warp 3.0 XR_3001 give it the command line /A:0 /U:0 /SMS /V all there you have it all crc errors are gone under warp 4.0 (fix pack 1 or 3). Regards, < Jas > --- AdeptXBBS v1.11y (FREEWare/2) * Origin: MILLENNIUM, Melbourne, Australia, +61-3-9312-3531 (3:635/737) --------------- FIDO MESSAGE AREA==> TOPIC: 251 OS2 BBS Ref: EDE00014 Date: 09/08/97 From: JASON ANDREWARTHA Time: 05:40pm \/To: ALL (Read 0 times) Subj: ADEPT bUGS Hi Guys i'd just like to let the guys handling adept know that the events files when running a spawn from an event the spawn can not be run due to the fact that adept translates cmd.exe /c e:\bat\maint.cmd to com.exe /c e:(a mess)bat(a mess)maint.cmd (its been line this since 1.11M) also i notice that when loging on localy the local bbs screens and the remote screens seem to get out of sincronisation. Last but not least the active line seting when you set line "2" inactive and go online with your terninal program while your online for 10-15 mins it diaplays make line inactive again so its anyonying when you realy go to activate the line again. Regards, < Jas > PS. im currently running 1.11Y3 . --- AdeptXBBS v1.11y (FREEWare/2) * Origin: MILLENNIUM, Melbourne, Australia, +61-3-9312-3531 (3:635/737) --------------- FIDO MESSAGE AREA==> TOPIC: 251 OS2 BBS Ref: EDE00015 Date: 09/08/97 From: JASON ANDREWARTHA Time: 10:51pm \/To: ALL (Read 0 times) Subj: Adept bugs (More) Hi could someone please pass this on the the guys deveolping adept. Vering im running 1.11y4 Report: (Follows) Haveing more than one (AdeptStartRexxthread) im a Adept rexx script causes Adept to shut down. im running them in my Chat.Cmd and have repeated this many times. one more anoying things is Meta F8 is being displayed as White and not greay and now meta FA is being displayed as White. i hope this is of some help to the next version. Regards, < Jas > --- AdeptXBBS v1.11y (FREEWare/2) * Origin: MILLENNIUM, Melbourne, Australia, +61-3-9312-3531 (3:635/737) --------------- FIDO MESSAGE AREA==> TOPIC: 251 OS2 BBS Ref: EDF00000 Date: 09/09/97 From: HERBERT ROSENAU Time: 07:52pm \/To: ALL (Read 0 times) Subj: Here it is Hi, The new of the OS/2 message editor SQED 1.14 is hatched in MXBBsNet, should be in the next day go through GFD and *is* uploaded to the maximus homepage: ftp.oeonline.com/incoming Tschau/Bye Herbert --- Sqed/32 1.14/development 109: * Origin: Geld macht reich! (2:2476/493)