Chuck Guzis wrote: Well it is not standard with the windows I have. I found the PD stuff. Mind you if somebody came up with a better OS when the 386's came out, would we have windows today? ** Well I did say 8088, so that is more on topic, as I wanted to play around with small - C for the 8080 and cross compile for a FPGA home brew computer. To get on topic for all the people that have a Z80 and CP/M check this page out. http://www.schorn.ch/cpm/intro.html SPL is new language for the z80 and some replacement CP/M software too. ** and afford!

On Fri, 2 Dec 2005, Chuck Guzis wrote: MASM was never standard with PC-DOS or Windoze. But it WAS included by many OEMs with MS-DOS (for example Toshiba T300). It is included with Visual Studio. (Where does Microsoft get their product names???) was originally included with PC-DOS, but stopped being included (around 5.0?) also EXE2BIN.EXE -- Grumpy Ol' Fred cisin at xenosoft.com

On Fri, 2 Dec 2005, Scott Stevens wrote: DEBUG (the most useful program to ever come from Microsoft) is STILL included in Windoze, even XP. BTW, debug scripts can easily INCLUDE the save, and even the exit from debug.

Fred Cisin wrote: so that is how they upgrade! :D

On 12/2/2005 at 10:10 PM Jim Leonard wrote: Hmmm, I must have the DEC OMPILATION chipset, 'cause c800:0 has: DEC OMPILATION OR DI SASSEMBLY PROHIB ITED Who'da guessed that DEC (somewhere, somehow) is still making chips? Actually, it's an Intel 82815, but I got that from Settings->Control Panel->System->Hardware->Device Manager->Display Adapters. I could have also found out by right-clicking the desktop and selecting Properties->Settings->Advanced->Adapter (that is, if I had the right display driver installed). Or I could have checked the registry... Cheers, Chuck

On 12/2/2005 at 10:40 PM Jim Leonard wrote: My typo--it was C000:0. As I was typing the message, I wondered if 2000 still supported CGA. Probably not. Cheers, Chuck

Actually, this is my favourite: http://toastytech.com/guis/desk.html -- --------------------------------- personal: http://www.armory.com/~spectre/ --- Cameron Kaiser * Floodgap Systems * www.floodgap.com * ckaiser at floodgap.com -- MOVIE IDEA: The Never-Ending E-mail Signature ------------------------------

On 03/12/05, Chuck Guzis <cclist at sydex.com> wrote: I would guess that there must be standard place for windows to ask "What sort of device are you" for plug and play. The rest would up to the driver. I quite often find windows can at least identify the type of device wether or not it actually works. Dan

On Sat, 3 Dec 2005, Sridhar Ayengar wrote: and at c000:00BD on this one (Asus). It appears to vary with different manufacturers. -- Grumpy Ol' Fred cisin at xenosoft.com

On 12/3/05 12:55 PM, "Sridhar Ayengar" <ploopster at gmail.com> wrote: ^-- Took me forever to realize that was a joke ...

Jim Leonard wrote: ... OK, I'll bite: C:\Documents and Settings\Owner>debug -d c000:0000 C000:0000 55 AA 7A EB 4B 37 34 30-30 E9 4C 19 77 CC 56 49 U.z.K7400.L.w.VI C000:0010 44 45 4F 20 0D 00 00 00-84 02 9B 0E 00 00 49 42 DEO ..........IB C000:0020 4D 20 56 47 41 20 43 6F-6D 70 61 74 69 62 6C 65 M VGA Compatible C000:0030 01 00 00 00 B0 10 B2 EF-30 38 2F 30 37 2F 30 33 ........08/07/03 C000:0040 00 00 00 00 00 00 00 00-00 00 00 00 00 00 00 00 ................ C000:0050 E9 E7 E4 00 B0 10 21 04-03 D0 30 7E 3C 21 C0 80 ......!...0~<!.. C000:0060 00 00 00 00 10 00 00 80-22 00 A5 8B E9 D0 0D E9 ........"....... C000:0070 D7 0D 50 4D 49 44 6C 00-6F 00 00 00 00 A0 00 B0 ..PMIDl.o....... - - You are a genius if, from that, you can tell what card I have. (answer below if you scroll down far enough) Nvidia Gforce FX5200

According to Peter Norton, the BIOS code in a PC scans memory starting at C000 and looking at every 2K block for the signature 0x55AA. (I am paraphrasing.) This marks a ROM extension, probably in a software cartridge or on a controller. What follows is supposedly a 1 byte length and a 3 byte jump instruction. The BIOS initialization will then branch to this code and execute it. The code can do anything, even take control of the system as is done with some software cartridges. This ROM block can be at any even 2K address. The IBM EGA adapter went at C000, so graphics cards tend to end up there. Comments in this block might be really informative or might simply try to look very much like some IBM module so as to facilitate interoperability. HTH, Gil

A couple of additions to what you've said : The extension ROM routine is called after much of the rest of the system ahs been initialised and after the software interrupt vectors have been set up. It is common for the extension ROM initialisation routine to point some of said vectors to itself, and maybe save the old contents somewhere so the original routine ROM routine could be called if the requested operation is not handled by the extension ROM. For example you could intercept the disk read/write INT vector, use routines in the extension ROM to handle, say, 8" drives on a special controller, and go back to the routines in the normal BIOS for drives 0 and 1. Other than this initialisation entry point (and the header/checksum), the BIOS places no requirements on the format of an extension ROM and never itself tries to call anything other than the initialisation routine. -tony

Jim Battle wrote: You didn't display the second bit; hit "d enter" again. It's much more descriptive one more bit down: -d C000:0080 0D 0A 52 33 30 30 20 49-6E 66 69 6E 65 6F 6E 20 ..R300 Infineon C000:0090 39 35 30 30 20 50 72 6F-20 42 49 4F 53 20 2D 20 9500 Pro BIOS - C000:00A0 50 2F 4E 20 31 31 33 2D-41 30 35 36 30 32 2D 31 P/N 113-A05602-1 C000:00B0 30 31 0D 0A 00 28 43 29-20 31 39 38 38 2D 32 30 01...(C) 1988-20 C000:00C0 30 32 2C 20 41 54 49 20-54 65 63 68 6E 6F 6C 6F 02, ATI Technolo C000:00D0 67 69 65 73 20 49 6E 63-2E 20 42 4B 2D 41 54 49 gies Inc. BK-ATI C000:00E0 20 56 45 52 30 30 38 2E-30 30 34 2E 30 30 38 2E VER008.004.008. C000:00F0 30 30 32 00 20 6E 65 62-66 31 33 30 38 2E 38 34 002. nebf1308.84 From this you can clearly tell I have a 9500 Pro. What does yours say from c000:0080 to :00f0? -- Jim Leonard (trixter at oldskool.org) http://www.oldskool.org/ Want to help an ambitious games project? http://www.mobygames.com/ Or check out some trippy MindCandy at http://www.mindcandydvd.com/

On 12/2/05, Jim Leonard <trixter at oldskool.org> wrote: Is XP guaranteed to map 0x000c0000 physical into c000:0 virtual in a VM86 session? Does it keep the whole ROM space mapped to its corresponding physical addresses? It certainly has to hide the video RAM... I would have expected it to replace the ROM with one of its own as well. Eric

I just found an HP Apollo PA-RISC 735/125 (1992) in very good condition. It's another example of HP's ridiculously clean under-the-hood design, and probably could almost be shown to industry outsiders as an object of art. </rant> I have to ask: Anyone have a copy of HP-UX that will run on this? I understand that NetBSD won't use the UltraSCSI and I like to keep the original configurations together anyway. John A.

... On Pentium systems, at least the current ones, a larger "page" is used to map most of the kernel memory into every logical address space with a single entry. IIRC the default mapping is 4 MB. Gil

Chuck Guzis wrote: -------------------snip snip------> I'd say one of the misfeatures of PC is legacy. At some point despite the pain some artifacts of XT or even AT hardware needed to disappear as real. In an age of small ram being 256MB a DMA chip that can't address more than 65k is a bottleneck. Especially when you consider the size of page files, swapfiles and program size in general. If you doing a 32bit system then a 32bit DMA is required. Then once you have blistering performance from the hardware the software must use it. Then you can emulate/simulate older legacy stuff in a way that is not hurting system performance inside the OS. Allison

They won't envy Apple for much longer. Based on the specs for Rosetta (the JIT PPC->x86 translator for the impending Intel Macs), it will only run PPC applications written for the G3 under Carbon. That means Classic apps -- including 68K apps -- will probably die in the future. -- --------------------------------- personal: http://www.armory.com/~spectre/ --- Cameron Kaiser * Floodgap Systems * www.floodgap.com * ckaiser at floodgap.com -- I'm in Pittsburgh. Why am I here? -- Harold Urey, Nobel laureate -----------

----- Original Message ----- From: "Jim Leonard" <trixter at oldskool.org> To: "General Discussion: On-Topic and Off-Topic Posts" <cctalk at classiccmp.org> Sent: Wednesday, December 07, 2005 6:22 PM Subject: Re: Legacy apps in Windows/OS X was Re: Old MS-DOS & Win Software apps the machine http://www.oldskool.org/ http://www.mobygames.com/ http://www.mindcandydvd.com/ I don't see the point of having so much legacy support in newer OS versions (it is more complicated and has more code to create bugs). If there is a need for such things then some company will develop a means of using your older software via an emulator. Besides what is so hard about keeping a legacy system in the house if you really need to run a 10 year old app on occasion? OS developers should concentrate on making their OS stable and reliable plus having good APIs, not in programming emulators and other add-ons.

----- Original Message ----- From: "Chuck Guzis" <cclist at sydex.com> To: <cctalk at classiccmp.org> Sent: Wednesday, December 07, 2005 8:26 PM Subject: Re: Legacy apps in Windows/OS X was Re: Old MS-DOS & Win Software I think Win2k server is the last MS OS that supports sharing directories with old Mac AppleTalk networks. For my smallish network that OS will last me a long time. So instead of complaining the latest and greatest doesn't support my setup I just keep using what works. Both Intel and MS didn't have to support their older designs, they did it because the product worked and sold reasonable well, and any major change in product that is incompatible with the older things you sold allows people to chose what product they want to buy all over again (if you need new apps you might as well consider a different OS/chip while changing). I think having a major commercial success is what stops innovation at those companies, if you have 5% market share you can pretty much try anything new since it will not matter as much.

It was thus said that the Great Patrick Finnegan once stated: No, he meant what he wrote. It's the CP/M compatibility layer built into MS-DOS. Address 5 of the PSP (Program Segment Prefix) contains a long jump to MS-DOS, and uses CL for the function code instead of AH. -spc (@#$@#!# thought I forgot my MS-DOS arcana ... )

On 12/7/2005 at 9:03 PM Patrick Finnegan wrote: With all due respect, I suggest that you review your X86 instructions and their effect on flags. Try running this little bit of code under your favorite debugger (DEBUG is fine) and have a gander at the flags register: xor ax,ax mov dl,-1 mov dl,dl int 3 Now, if your point of view held, the condition code should show a nonzero negative (signed) value. In fact, they show a zero positive value (i.e. the results of the XOR AX,AX). In fact, moves, be they register to memory, register to register or memory to register or memory to memory do not affect condition code bits. Else why have a nutty instruction like JCXZ? You coiuld just MOV CX,CX and jumpt on the zero result... Nope, I mean what I said--try it. This little tidbit goes all the way back to the time before DOS 1.0, when some level of compatibility with CP/M applications was implemented. Notice that I didn't say "CP/M-86", which uses int 224 to make system calls (which will not work under MS-DOS or Windows of any flavor). I suspect this is an artifact of old SC-DOS/QDOS. Cheers, Chuck

Chuck Guzis declared on Wednesday 07 December 2005 21:32: I'll admit it's been a while since I've bother to do much x86 assembly, but I don't doubt you're right. I think I might have been thinking more of "and dl, dl", which I'm pretty sure ought to update the flags. Pretty much any assembly language has useless constructs like that, though, just because it makes the instruction set easier to implement (and sometimes, compilers easier to write). For example, in RISC architectures, register 0 is generally fixed to read 0. When you would normally do stuff like: mov r1,r2 you can also do mov r0,r2 or even mov r0, r0 (using syntax there the matching intel syntax Sure there's a reason. JCXZ is one instruction, and (probably) uses up less code space than the other. You could "AND CX,CX" and do a JZ afterwards to do the same thing. I don't have a Windows system to try it on. I could probably find something running DOS (or OS/2) to try it on, assuming DEBUG is installed on the machine... I vaguely remember the PCB "jump to DOS" vector, but never used it when I wrote code (for the same reason I never used an FCB to touch files.. probably just too young :). Pat -- Purdue University ITAP/RCAC --- http://www.rcac.purdue.edu/ The Computer Refuge --- http://computer-refuge.org

Chuck Guzis wrote: The "half-carry flag"? Huh? -- Jim Leonard (trixter at oldskool.org) http://www.oldskool.org/ Want to help an ambitious games project? http://www.mobygames.com/ Or check out some trippy MindCandy at http://www.mindcandydvd.com/

Tony Duell wrote: Is this "half-carry" flag addressable, or is it just something internal? A quick gander at 808x flags doesn't show a "half-carry" flag, unless I'm missing something obvious. -- Jim Leonard (trixter at oldskool.org) http://www.oldskool.org/ Want to help an ambitious games project? http://www.mobygames.com/ Or check out some trippy MindCandy at http://www.mindcandydvd.com/

I rescued an apple quadra 605 machine this weekend. Initially am going to use it to recover some old files off floppies, but eventually I want to run netbsd on it. And netbsd needs a full 68040, not the 68LC040 that's in the quadra 605. so does anyone have a 68040 25mhz they like to part with cheap? I'm located in Santa Barbara, California. thanks, Jeff

Jeff Davis wrote: I thought NetBSD works fine with the 68LC040, but wouldn't work with the 68*EC*040. Peace... Sridhar

On Thu, December 8, 2005 10:03 pm, Sridhar Ayengar wrote: I had to go look that up: 68040: cpu + mmu + fpu 68LC040: cpu + mmu (no fpu) 68EC040: cpu (no mmu or fpu) but it's not totally complete so some things crash. Btw, someone emailed me and I have a new cpu on the way. Sweet! Jeff

On 12/8/2005 at 5:18 PM woodelf wrote: Write a loop that executes, oh, 10,000,000 times. Get a stopwatch. Start the program and the stopwatch simultaneously, stop the stopwatch when the loop ends. Calculate the average time per instruction as (time to execute/(10,000,000*number of instructions in loop) :)

On 12/8/2005 at 6:33 PM woodelf wrote: MOV CX,9680H MOV DL, 98H TOP: ...all those stinkin instructions... LOOP TOP DEC DL JNZ TOP You'll execute the last two instructions 152 times. Cheers, Chuck

On 12/8/2005 at 7:03 PM woodelf wrote: Nope, but it did copy the number down wrong off my HP-16C. That should be 99H, not 98H loaded into DL. Alternatively, you could change the bottom JNZ to a JNS. Consider what happens. You execute one iteration of 38528 times through the inner loop. The remainder of 152 trips through the outer loop, you execute 65536 iterations each time for 9961472 total iterations. Add the 38528 for the first time and you get (surprise!) 10000000. Cheers, Chuck

woodelf wrote: They're still being made, but they're nowhere near as cheap as they were once upon a time. Peace... Sridhar

Jim Leonard wrote: I thought the pre-fetch queue was longer than that. Still most instructions I bet are 3 to 4 bytes long when you have to acess memory. I just tend to count memory cycles on instructions for ball park timing. What I do find odd that the 8086 came out first then a few years later the 8088. It still think the 8086/8086 could have had two instruction sets -- 8080 emulation and a better 16 bit design.

Chuck Guzis wrote: Hey, don't knock my favorite "nutty" 8088 opcodes! :-) It's because of opcodes like jcxz, xchg, repne cmpsw, add, and others that make 8088 asm coding so much fun 25 (!) years later. One of my favorite things to do when I have "free" time is to take some old 8088 code and speed it up with judicous use of stuff like that. Some of the earliest games for the 5150 were ports of 6502 or Z80 games and rewriting certain loops can speed the game up 500% or more. Some people like Sudoku puzzles, I like optimizing 8088 code for speed. Call me crazy 8-) -- Jim Leonard (trixter at oldskool.org) http://www.oldskool.org/ Want to help an ambitious games project? http://www.mobygames.com/ Or check out some trippy MindCandy at http://www.mindcandydvd.com/

Never mind, found it in the above thread... Jim Leonard wrote: -- Jim Leonard (trixter at oldskool.org) http://www.oldskool.org/ Want to help an ambitious games project? http://www.mobygames.com/ Or check out some trippy MindCandy at http://www.mindcandydvd.com/

At 1:40 PM -0800 12/7/05, Cameron Kaiser wrote: Personally I think there is a real opportunity for some software company here. I'm sure I'm not the only one with a dependance on old legacy 68k Mac OS apps. In fact the only app's that I use "Classic" support in 10.3.9 on my G5 2x2 to run are both 68k app's (ClarisDraw and the original "Master of Orion"). As long as someone can produce something that will run my legacy 68k apps well, and does it at an affordable price, I'll buy it. Basically I'd like something that allows easy integration of the App's with newer ones. It's probably going to be a year or two before I can think of getting an x86 Mac, so hopefully by then, someone will have come out with something. I'm also hoping that Virtual PC works more reliably, and faster. Worst case, I setup something like my 8500/180 running an old enough version of the OS so that I can also run "Warlords 2". BTW, I do own PC versions of both games, from before I got a Mac, however, they're limited to 640x480, while the Mac versions go up to something like 1152x870. Both games are also well on topic, but still two of my favorite games of all time. Hmmm, I just thought of another potential problem... I wonder how well my ADB-to-USB converter will work, I really don't want to give up the ADB Apple Extended II Keyboard I'm typing this on. They just don't make keyboards like this one anymore. Zane -- -- | Zane H. Healy | UNIX Systems Administrator | | healyzh at aracnet.com (primary) | OpenVMS Enthusiast | | | Classic Computer Collector | +----------------------------------+----------------------------+ | Empire of the Petal Throne and Traveller Role Playing, | | PDP-10 Emulation and Zane's Computer Museum. | | http://www.aracnet.com/~healyzh/ |

On Thu, 08 Dec 2005 00:27:31 +0000 Adrian Graham <witchy at binarydinosaurs.co.uk> wrote: It's not a matter of 'meeting the requirement' or else for the manufacturer. It's a matter of satisfying a customer and holding onto market share. I happen to prefer manual windows and non-central locking, and won't buy a car with electronic locks. That is my choice and car makers who want to sell a car to me will offer it. Much the same is true with Microsoft. They want to hold onto their legacy customers, in fact leveraging that legacy is a big part of how they've held onto and controlled the market. There come points at which it's probably to Microsoft's advantage to abandon legacy features, and that's when they do it. But they and a lot of the rest of us know that's really one of the only things they have going for themselves.

On 12/8/2005 at 12:27 AM Adrian Graham wrote: when Adrian, are you aware that here in the US we still use Edison (screw) base lamps? One of Edison's carbon-filament creations would fit very nicely in the reading lamp by my desk. I don't know if it's still there, but I recall an Edison lamp installed at a fire station in Livermore, CA that's been burning for many decades. Cheers, Chuck

There are actually a number of ancient bulbs around the country that refuse to die. Pretty neat, anyway. William Donzelli aw288 at osfn.org

On Thu, 8 Dec 2005 10:28:53 -0500 (EST) William Donzelli <aw288 at osfn.org> wrote: And I would bet that General Electric, Sylvania, and a few other concerns have people who study these bulbs to make sure they don't make any more like them.

I was under the impression that the way to make a bulb last for a good long time is to under-run it -- to run it off a much lower voltage than designed. I bet if you took a modern (110V) bulb and powered it off, say, 90V, it would have a long life. Of course if you do this you get a lot less light per watt of electical input power. Photofloods (do you get those in the States?) are the reverse. Bulbs that are deliberately overrun to give a high light intensity (and higher colour temperature) than normal bulbs of the same power. The trade-off is a very short life -- a couple of hours for some of them. -tony

At 12:27 AM +0000 12/8/05, Adrian Graham wrote: If you'll note, I'm not expecting Apple to provide this support, I'm hoping a 3rd party will provide it. My problem is that one of the key application I use that requires 68k and was dropped by Claris (aka Apple's App's division of the 90's), the last new version came out sometime around 1995. There isn't a better app, and it wasn't until this last year that there were any apps out there that can read the data files. Lack of space. My space for legacy HW is taken up with a couple of PDP-11's and a VAX. Though I must confess my Alpha is fast falling into the legacy category :^( If I could I'd have an Amiga, Apple ][, and Commodore 64 setup around here, but I don't have the space. It's a good thing you did :^) Zane -- -- | Zane H. Healy | UNIX Systems Administrator | | healyzh at aracnet.com (primary) | OpenVMS Enthusiast | | | Classic Computer Collector | +----------------------------------+----------------------------+ | Empire of the Petal Throne and Traveller Role Playing, | | PDP-10 Emulation and Zane's Computer Museum. | | http://www.aracnet.com/~healyzh/ |

At 6:14 PM -0800 12/7/05, Cameron Kaiser wrote: I'll have to look at SheepShaver, I've not heard of it. Last I looked, Basilisk II isn't really what I'm after, though it might be fine for the games. I don't have to wonder why they stopped, I've seen the writing on the wall for a while now. How many people run Classic app's under Mac OS X? I'm under the impression that most users don't even bother with installing the classic emulation. A perfect example of Apple ditching legacy support is Classic Appletalk. It wasn't in 10.0, and I don't think it was in 10.1, it was in 10.2 and 10.3, BUT, they removed it from 10.4. Once they'd added it in, how much trouble was it to keep it in there???? Granted, for most people this didn't cause problems, but it's caused one person I know of to move to Linux, and it caused me to have to spend some serious time upgrading Samba on my VMS server, as I use Appletalk to allow my Mac to access my VMS box. BTW, I'm back on 10.3 and probably will be for a while, but that's because I think 10.4 is a waste of resources, and 10.3 does all I need. Until I need to run an app that requires 10.4, I don't think I'll be upgrading (and I bought 10.4 the day it came out)! As for why Apple is switching to Intel, it's my *personal* belief it's for their laptops. Zane -- -- | Zane H. Healy | UNIX Systems Administrator | | healyzh at aracnet.com (primary) | OpenVMS Enthusiast | | | Classic Computer Collector | +----------------------------------+----------------------------+ | Empire of the Petal Throne and Traveller Role Playing, | | PDP-10 Emulation and Zane's Computer Museum. | | http://www.aracnet.com/~healyzh/ |

On 12/2/2005 at 2:51 PM Fred Cisin wrote: I can't find it in VS 2005 aka VS 8. Maybe MS expects one to write all of one's assembly as C inline _asm statements? (P.S. Anyone noticed how SLOW VC 14.0 has gotten?) Cheers, Chuck

On 12/2/2005 at 7:36 PM Scott Stevens wrote: Yeah, well, you know that with the release of NT, Microsoft wanted a completely portable operating system, so they could release versions for MIPS and Alpha. Boy, that idea sure died in a hurry! ML's in the 2000/XP DDK, though I don't know if it's in the Vista DDK. I'd be surprised if it was. Cheers, Chuck

woodelf wrote: Ahem, there were better OSs at or shortly after the time 386's came out... but "better" does not mean that they were in wide use. Note the VHS vs. Betamax effect. There is plenty of stuff around there which does not primarily require 80x86s with DOS or Windows (or rather, just with emulators as a bootstrap for a native system). There were numerous scripting languages for CP/M and/or 8080/z80 based systems. Unfortunately, none really persuaded me to use them beyond the "hello world" level. They were typically at some sort of DOS batch level - if you want to really program something, it pretty fast results in obscure hacking as some constructs were not very well thought through. Compare this to the prior Bliss or DCL languages known from DEC, which I liked pretty much those days. Same goes for the PDP MAC assembler with its accompanying macro libraries. CP/M 3 had RMAC which would be comparable, but I have never really seen it in use with really sophisticated macros. Holger

On 03/12/05, Dan Williams <williams.dan at gmail.com> wrote: Pressed send too quickly, Here's a page with 386 software, all versions of windows and apps. If anyone is interested : http://www.angelfire.com/ca7/the386experience/downloads.html Dan

On Fri, 02 Dec 2005 12:00:56 -0700 woodelf <bfranchuk at jetnet.ab.ca> wrote: There are numerous mirrors of the old Simtel archive. And I personally know of a few sites that have mirrored some of the old 'shareware/shovelware' CDROM collections online. I won't broadcast the URLs here because they don't need the huge bandwidth bill, but you can find them. Often by just googling the filename of various old MS-DOS zipfiles that you remember and/or have. Private queries about the big sites I have found that mirror 'classic' DOS era CDROMs are welcomed. And maybe it's time for a 'CDROM exchange' type of thing. Not for warez purposes, just to back up the old shareware/freeware collections that many of us have a number of. It makes more sense to ship around CDs than it does for everyone to download everything individually.