--------------- FIDO MESSAGE AREA==> TOPIC: 203 C++ Ref: E4Y00006 Date: 04/28/97 From: HERMAN SCHONFELD Time: 05:00pm \/To: DANIEL MCGREGOR (Read 4 times) Subj: Newbie DM>Hello Herman Schonfeld! I just thought I'd drop in on your conversation DM>:) DM>26-Apr-97 15:57:32, Herman Schonfeld pointlessly wrote to Jay Moreau DM>They rambled on fou HOURS about: Newbie HS> Get Borland C++ v5.0, it's good to start off with and to keep using DM>it, HS> however, if your after speedy executables use either HS> 1) Watcom HS> 2) DJGPP HS> 3) MSVC++ HS> - Origin: Fox's Lair BBS Bris Aus +61-7-38033908 V34+ Node 2 DM>(3:640/238) DM> ^ DM>Well I know you're not in canada so I won't ask where you can get it for DM>less that $300 But, do you have any idea where I could get it for a DM>reduced price? no > This message was brought to you by DM> -=> Daniel McGregor (Who else? :) <=- DM>The Dragons Cave 2am - 5am Central standard time. ... For sale: Hourglass for timing Windows. --- Ezycom V1.48g0 01fd016b * Origin: Fox's Lair BBS Bris Aus +61-7-38033908 V34+ Node 2 (3:640/238) --------------- FIDO MESSAGE AREA==> TOPIC: 203 C++ Ref: E4Y00007 Date: 04/22/97 From: KEVIN CAMPBELL Time: 05:36pm \/To: PAUL WANKADIA (Read 4 times) Subj: self-destruct > On 11 Apr 97, Kevin Campbell wrote to Paul Wankadia -- >> I'm wondering how to, from an object's constructor, to self-destruct. >> E.g. if there is some kind of error, the object will NOT be created. > KC> Overload the new operator. > Oh crikey ... > KC> If the object is created, return a pointer to it, else return NULL. > ... but a good idea, IMO. Just tell me how to DO that ... Simple. It's just the same as a standard class function entry or operator overload. Define the function "void *operater new(size_t size)". Now make the code for it as void *xxxx::operator new(size_t size){ } Where xxxx is the function name. In the function, you simply allocate the memory, and return a pointer, alling a terminate function if you like if you can't get the memory. - Kev --- FMail/386 1.02 * Origin: Mail shipped from Deimos Spaceport (2:259/17) --------------- FIDO MESSAGE AREA==> TOPIC: 203 C++ Ref: E4Y00008 Date: 04/26/97 From: CRAIG A MCKAY Time: 10:56am \/To: ANTHONY TIBBS (Read 4 times) Subj: Random access files Y'ello Anthony! 23 Apr 97, Kurt Kuzba had to endure Anthony Tibbs's ramblings (shown below). KK>> char* FromName[36]; AT> "char *FromName [36];" ? This won't/shouldn't even compile AT> correctly. AT> However, the the method you specified is saying this: I want a pointer AT> to a pointer to a 36 byte char. No, it says "I want 36 pointers, of type pointer to char", It certainly would compile, and I have used sort of thing a million times: after all that is the structure of an argv. But, as you say, it is not right for what this chap wants to do. AT> I know it is possible to have a pointer to a AT> pointer, but not in this case.) Yeah, it doesn't make a whole lot of sense in this instance, all he needs is: char FromName[36] ; (Everyone seems very touchy in here so I thought I'd jump on my high horse too. :) Sorry...) Fido = 2:259/33 --Craigzilla++ Internet = McMoose@sol.co.uk WWW = http://www.taynet.co.uk/users/mcmoose/ --- GEcho 1.00 * Origin: The Kilted Bun, Letham Angus -- hame o' the McMoose (2:259/33) --------------- FIDO MESSAGE AREA==> TOPIC: 203 C++ Ref: E4Y00009 Date: 04/27/97 From: PAUL MADDOX Time: 11:11am \/To: ALL (Read 4 times) Subj: Wolf 3d gone wrong Hi. Sorry if this is a bit off-topic. I've got the source for Wolf 3d, and although I CAN get it to compile (using BC++3.0, the original host) it somehow compiles an EXE that's 419-ish K. Whereas the shareware version's EXE is only 100-ish K. When I exec the program the actual EXE appears to work fine, but I get bitmaps that are substituted for others, and crazy characters. Basically the engine works fine, it's just some problem with the bitmap files. Sooo. Is this because I'm using the shareware graphics files? TTFN - Paul. --- cPoint 2.17 * Origin: Nowhere, man. (paul.maddox@zetnet.co.uk) (2:250/102.23) --------------- FIDO MESSAGE AREA==> TOPIC: 203 C++ Ref: E4Y00010 Date: 04/28/97 From: DANIEL MCGREGOR Time: 08:40pm \/To: NICK SHREDERS (Read 4 times) Subj: Re: DJGPP OPTIMIZATIONS 2/3 Hello Nick Shreders! I just thought I'd drop in on your conversation :) 27-Apr-97 06:27:00, Nick Shreders pointlessly wrote to Daniel McGregor They rambled on fou HOURS about: Re: DJGPP OPTIMIZATIONS 2/3 CB>>> 1.8Mhz. As a result, I am very much aware that you have to AT>> Atari DM>> Power Mac? NS> Windows 95? Maybe, at least windows 95 only slows down your computer to 1.8mhz. It doesn't start that way :) > This message was brought to you by -=> Daniel McGregor (Who else? :) <=- The Dragons Cave 2am - 5am Central standard time. --- Terminate 4.00/Pro * Origin: The Dragons Cave -- 2am - 5am CST (1:140/237.2) --------------- FIDO MESSAGE AREA==> TOPIC: 203 C++ Ref: E4Z00000 Date: 04/28/97 From: WIM VELDHUIS Time: 09:06pm \/To: HERMAN SCHONFELD (Read 4 times) Subj: Video Herman Schonfeld wrote in a message to Wim Veldhuis: WV>Herman Schonfeld wrote in a message to Tim Esselens: HS> Protected mode is a flat memory model. There are so HS> segmentation limits so therefore you can allocate as much HS> memory as you want to an object. WV>This is not correct. In a flat memory model there are still segment WV>registers. The values in those registers however rarely change during WV>program execution. WV>In a 16-bit world this is also possible, but is usually referred to as WV>the tiny memory model. WV>Protected mode however is a complete different piece of cake. It has WV>nothing to do with 32 or 16-bit (It is possible in both) but with the WV>interpretation of the values in the segment registers. In real mode WV>these register are actually a part of the actual address. In protected WV>mode the values in the segment registers are used as index in a WV>descriptor table, which contains the actual segment adress information WV>AND the access rights for the segment (hence the name protected mode). WV>After all, under Windows 3.1 I awfully often got a General Protection WV>Failure, which is a good indication of protected mode . HS> I knew most of that but what do you mean flat mode? Flat memory model means the segment registers do not alter their value during execution of the program. HS> I thought protected mode and flat mode where the same? No, they are not. In 16-bit real mode, the tiny memory model is a flat memory model. Since segments here are only 64K, you reach the ends soon. HS> Is flat mode when you switch the cpu into protected and HS> change the segment limits then switch back to real mode? HS> (making it possible to allocate large segments) No. In protected mode you can have multitasking, here several processes each run in their own memory space. They are all running in a flat memory model. A device driver, for instance the hard disk driver, must be able to enter all those memory spaces, otherwise each proces needs to run its own OS. Device drivers are therefore programmed in segmented memory model. A device driver sets up the correct segment selectors in order to obtain access to the memory space of the processes it is servicing. Not every process however can enter another memory space. The protected mode scheme prohibits this. Only processes running under the correct priviliges can do this trick. I hope this makes it a little bit more clear. WV>mvg/wr WV> HS> Thanks for replying! HS> ... !enif tsuj si gnihtyrevE HS> ___ Ezycom V1.48g0 01fd016b HS> - Origin: Fox's Lair BBS Bris Aus +61-7-38033908 V34+ Node HS> 2 (3:640/238) mvg/wr --- timEd/2 1.01.g3+ * Origin: LightHouse BBS ==> I am a H.U.G.O. Member ! (2:285/324.3) --------------- FIDO MESSAGE AREA==> TOPIC: 203 C++ Ref: E4Z00001 Date: 04/28/97 From: CAREY BLOODWORTH Time: 09:13pm \/To: JAY MOREAU (Read 4 times) Subj: NEWBIE JM>stories about which compiler to use... Currently I'm using Turbo C++ v3.00. JM>Is this an Ok compiler to start with? Or should I be using something lse? JM>Any help with this would be greatly appriciated... And Long Live C ;) Pros: It's okay to start with. Reasonably fast. Tolerable debugging ability. Cons: It doesn't generate the fastest code. Has few optimizations. Is only a 16 bit compiler. But, yes, it's okay to start with. It does have a problem when it deals with huge pointers. If you increment it, it may not always properly handle them. Generally though, you aren't going to run into it. --- QScan/PCB v1.19b / 01-0162 * Origin: Jackalope Junction 501-785-5381 Ft Smith AR (1:3822/1) --------------- FIDO MESSAGE AREA==> TOPIC: 203 C++ Ref: E4Z00002 Date: 04/29/97 From: HERMAN SCHONFELD Time: 04:41pm \/To: MARAT AFANASYEV (Read 4 times) Subj: gcc+iostream classes question MA>main.cpp:168 (main.o): Undefined symbol ostream::operator<< referenced MA>from MA>text MA> segment your settings are incorrect ... Error finding COLD BEER. Sysop not loaded. --- Ezycom V1.48g0 01fd016b * Origin: Fox's Lair BBS Bris Aus +61-7-38033908 V34+ Node 2 (3:640/238) --------------- FIDO MESSAGE AREA==> TOPIC: 203 C++ Ref: E4Z00003 Date: 04/09/97 From: PETER HAYWOOD Time: 01:45am \/To: PAUL MADDOX (Read 4 times) Subj: Re: DOS4GW variables Groovy hepcat Paul Maddox jived with All on 11 Mar 97 12:39:00! DOS4GW variables's a cool scene. Dig it! PM> Is there anyone out there that knows how to allocate memory in DOS4GW PM> for Watcom C++ 10.6? Just malloc or calloc it or use the "new" operator. PM> Basically, I want to move some variables into protected memory above PM> 640kb to free some conventional memory so I can shell another program PM> using system(). You don't need to move anithing around. If you're compiling and linking in protected mode using DOS4GW then all memory you allocate using standard C/C++ functions is allocated in extended memory. You can allocate memory below the 640k barrier if you need it (for some DOS or system related purpose, for example). Check the online help (Programmer's Guide) for more information. Wolvaen ... I am DOS of Borg! Prepare... oops, out of memory! --- Blue Wave/RA v2.20 * Origin: The Gate, Melbourne Australia, +61-3-9809-5097 28.8k (3:633/159) --------------- FIDO MESSAGE AREA==> TOPIC: 203 C++ Ref: E4Z00004 Date: 04/09/97 From: PETER HAYWOOD Time: 01:45am \/To: GARRY TIN (Read 4 times) Subj: Re: WATCOM debugger for dos Groovy hepcat Garry Tin jived with All on 16 Mar 97 05:53:00! WATCOM debugger for dos's a cool scene. Dig it! GT> Yo, is there anyone know where can i get a watcom debugger for dos? GT> 16 bit or 32 bit (wd.exe) Yeah, same place you get the compiler. :) Unfortunately, living here in the land of Aus, you're outta luck, my freind. The Watcom compiler costs well over 500 bucks here - if you can get it! (And if you're thinking of buying it, I know where you can get it in Melbourne.) Wolvaen ... I am Shaggy of Borg. You shall be groovy. --- Blue Wave/RA v2.20 * Origin: The Gate, Melbourne Australia, +61-3-9809-5097 28.8k (3:633/159)