• Hello MLAers! We've re-enabled auto-approval for accounts. If you are still waiting on account approval, please check this thread for more information.

Beige G3, IDE and Apple's ridiculous implememtations . . .

Trash80toHP_Mini

NIGHT STALKER
68040
.  .  .  of anything that even smacks of an Industry STANDARD!

Did they really leave pin 20 implemented on the G3's internal "IDE" connectors on purpose?

I mean  .  .  .  really????? >:(<

edit: decided to leave the sub-standard SCSI connector alone for now  .  .  .  but really?

 
Last edited by a moderator:
To be fair, most PC motherboards until the beige G3's time period still had the key pin intact. Plenty of cables did not have pin 20 blocked off either, if it did, a firm press when installing would punch out the keyed hole. :p I don't think Apple intended the SCSI to be used as the primary storage interface on the beige G3, it was kept for backwards compatibility and general peripheral usage, which 5MB/sec was usually fine for. What was ridiculous was only having 3 PCI slots and 3 DIMM slots!

 
Nah, I meant when they butchered the SCSI Standard by adopting a sub-standard DB-25 connector. The Plus was the MacPro of its DTP era. They should have dropped the legacy FDD connector or mooshed it over to the side and gone with the Centronics 50 standard. To have carried that D-Shell waste of space all the way up through the SE/30 was laughable. The 512Ke was for the two floppy swappy masses and the dual FDD SE's after that.

Today Apple can't get rid of connector ports fast enough, go figure!

 
Last edited by a moderator:
The Mac Plus was the most powerful Mac (not counting the Mac XL) until the launch of the Mac II in March 1987. I would argue the modern equivalent of the Mac Plus is the iMac rather than the Mac Pro. The latter has its roots in the Mac II.

My guess is that some engineer thought that he/she was being very clever with the DB25. Most of the SCSI functionality in half the pins and still keep the floppy port. You do the floppy port a disservice. Given how expensive hard-drives were back then, a second floppy was for many the only affordable option. When I started out with Macs in 1989 in college, almost all of the student Macs were either Pluses or the 2-floppy SEs and the latter were prized because a. they don't high-density disks and b. you didn't have to spend an age swapping floppy disks between the disk with the system folder and the disk with the application and files on it. I can well understand why 2 floppy drives were worth having.

I always wonder why they didn't go with the Centronics 50-pin port. I have always assumed that it was based on cost. DB-25 is cheaper and almost as good. Unless it was an aesthethic choice.

 
That was my point, the Plus was THE Mac until the II and SE wee released. Top of the line = Mac Pro in any given generation. To call it less would be like judges at the Olympics "leaving room" for a higher score when the first gymnast deserves the ten. IMO you can't do the same to the Plus just because the next contestant was the Mac II. ;)

 
My guess is that some engineer thought that he/she was being very clever with the DB25. Most of the SCSI functionality in half the pins and still keep the floppy port. You do the floppy port a disservice.
My schematics of the SE/30 board indicate which pins from the 53C80 controller map to which pins on the 50-pin IDC connector and to the DB-25 connector. No functionality was lost by choosing a 25-pin connector for Apple, so it is kind of inaccurate to say "most of the SCSI functionality". Personally the DB-25 port isn't a huge obstacle, just convert it to some other more usable form.

That being said, the dual-floppy systems were for people who wern't willing to pay the costs of HDDs. I would think in about 93% of scenarios, FDD + HDD is better than 2x FDD. While the SE is decidedly low-end in comparison to the Mac II, it certainly brought many features that you could get excited about. It brought a real expansion slot to the compact form-factor, as well as an optional 40meg (or 20) HDD. To have an internal 40MB HDD, along with the floppy drive would have been pretty big back then. (If expensive). Apple did continue to sell the Plus alongside the new machines for those who wanted a cheaper single-floppy system.

 
Last edited by a moderator:
That was my point, the Plus was THE Mac until the II and SE wee released. Top of the line = Mac Pro in any given generation. To call it less would be like judges at the Olympics "leaving room" for a higher score when the first gymnast deserves the ten. IMO you can't do the same to the Plus just because the next contestant was the Mac II. ;)
Point taken. However, the Mac II was in development when the Mac Plus came out. I don't think that it was ever intended to be the best that Apple could produce, more of a reasonable successor to the 512, even if it would be teh most powerful mac for almost a year.

 
It's not a question of signal lines and their function. Alternate ground lines for elimination crosstalk on the cable would be the reason for using the full 50 pin Centronics connector. Depending on lines grounded on only the peripheral end for shielding may or may not suffice. I don't know, but the engineers that established the SCSI standards did. I've never cut a DB-25 to Centronics 50 cable apart to check, I've sort of been hoping that's how Apple may have done it.

AFAIK there's never been a fast SCSI2 implementation that used DB-25. There could be very good reasons Apple never adopted a faster controller or the raised the bar on external SCSI 1 performance. SCSI was a commercial grade electronics standard and the Mac had to meet the FCCs far more stringent standards for Class B certification. Among the reasons for limited SCSI performance may very well have been some terrible decisions made within the infinite loopiness in the implementation of SCSI on the Plus. Dunno, but it makes for an interesting discussion, hopefully it will become informative when EudiG (Gorgonops/Eudimorphodon) jumps in to slap me back down. That's a wonderful game we've been playing for over almost fifteen years now. ;D

It's absolutely true that most Macs (and micros in general) continued to be dual floppy systems. Saying Apple should have eliminated the external FDD connector on the top of the line Mac is a bit of hyperbole on my part for the sake of the SCSI discussion. ;)

 
Point taken. However, the Mac II was in development when the Mac Plus came out. I don't think that it was ever intended to be the best that Apple could produce, more of a reasonable successor to the 512, even if it would be teh most powerful mac for almost a year.
Point granted. (see hyperbole above) I've often heard it stated that the Plus was a stop gap Mac. Introducing the LEM for the next lineup as the HEM for that year makes for an interesting parallel to being IIvEXED in the next generational sea change. ;)

 
The Mac Plus was one of the most popular macs. It was available for what seemed like an eternity. To poor students like myself, it was almost affordable. It was within reach. It played all of the games that I had at my disposal in 1990 and it was enough for my brother to type up his master's thesis. All it really needed was more RAM.

 
The Mac Plus was one of the most popular macs.

<snip>

All it really needed was more RAM.
LOL! That's the story of every Compact Mac from day one*** when SJ's demo of "the computer for the rest of us" couldn't even handle his presentation. At the first of many of his overblown product announcements he demo'd the 512K to intro the first in the long line of RoadApples to come. (I refuse to accept LEM's overgenerous dilution of that traditional term to Compromised Macs.)

*** Nope! ;)   SE/30 doesn't count, it's not a compact Mac, it's a stunted IIcx with delusions of being a IIci. :p

 
LOL :lol: , that was kind of funny to read. I still find it hilarious that because the original Macintosh was compromised enough that they had to up the RAM for the demo. I don't know if you can even call it a tech demo, as a computer comparable to the one used for the demo didn't even ship until way later! I know I am basically rehashing what you said above, but I just find it so hilarious that I just had to post. ;)

 
By the time the Fast SCSI standard came out, the high density 50 pin connector was also adopted. If Apple had wanted to support external Fast SCSI they could have without putting the giant Centronics connector on the back.

A friend who worked higher tier support at Power Computing once told me that Apple never supplied external high speed SCSI because they didn't want to deal with the support issues. Folks forget, but the total supported cable length for faster SCSI standards is considerably shorter than for vanilla 5MB/s SCSI. Apple's philosophy was that fast devices should be in the box (after the Plus) and the external connector is really for slow peripherals like scanners.

 
Infinite loopiness to the max in that there philosophy**. SCSI-2 was published in August 1990**** only five months after the release of the IIfx. Apple didn't come up with a Mac with a fast internal bus until the Quadra 950(?) in mid-1992. I'll repeat that for clarity: WHAT FAST DEVICES? ;)

Radius put a Fast SCSI-2 external bus on the Rocket Daughtercard in 1992 at the very latest (date on my card) but I don't know when FWB, ATTO and the rest of the gang introduced SCSI-2 cards for NuBus putting Fast/Wide on the external bus playing field.

Dunno, Apple seems to have been waaay behind the curve when it came to mass storage I/O and didn't seem to care about it much it all. Pretty darn bad when their only serious market penetration was content production  .  .  .  I mean really? ::)

** This from the same gang who insisted "our customers don't use slots" while setting about creating an impossible mess of incompatible slot "standards."  That lack of compatibility, especially in the idiotic consumer level LC slot menagerie made it impossible for third parties to achieve economies of scale in production, thereby keeping cost levels high enough to ensure that statement was a self-fulfilling prophecy. Infinitely loopy! :p

**** Wikipedia

 
 I still find it hilarious that because the original Macintosh was compromised enough that they had to up the RAM for the demo.
The fact they had to do that is the source of my tongue-in-cheek insistence that I in fact own an example of the "Original" Macintosh because I have a 512k, not a 128. ;)

And, re the OP, I can also attest that many PC-based IDE implementations prior to the late 1990s didn't have the missing pin on the connector. One of the sort of amazing things about IDE was just how long it was around before it was "Standardized". Compaq introduced the first PCs using IDE in 1986 but it wasn't officially rubber-stamped by ANSI until *1994*. So far as I'm aware I don't think the missing-pin connector keying actually became an official compliance requirement until ATA-4, which was in 1998, and the Beige G3 *sort of* predates that. (With IDE the industry generally moved faster than the documented standards, so drives and controllers that supported various advanced features were usually shipping well before each revision of the ANSI standard became official. And, of course, it would be inevitable that some of these early adopter devices would end up being out of compliance with some aspect of the final standard when it eventually took effect.)

 
Last edited by a moderator:
The fact they had to do that is the source of my tongue-in-cheek insistence that I in fact own an example of the "Original" Macintosh because I have a 512k, not a 128. ;)
I would say that is fair. The 128K was a management "Oh my RAM is too expensive!" decision, making it in jt and I's minds, a RoadApple. If you *really* wanted to go for the original Macintosh feel, you would use an original 128K motherboard, along with non-marked (plain "Macintosh") case, and then solder in a 512K upgrade, but I do think the 512K (as Apple sold it) counts as the original. 512kE is a little murky, but a plain 512K can definitely be called the 'first'.

 
External SCSI is mostly used for scanners, tape drives, optical drives, burners, MO drives. People who need fast external SCSI tended to get a decent high end card and external RAID drives.

Could be worse and we could have to deal with that pre standards SCSI port on the Atari ST.

 
That mode was inflicted upon the Macintosh using world by Apple's substandard implementations, not Apple's response to user's wants or needs.

The problem with the fast inside/slow outside the box philosophy is: Apple never provided for fast internal drives! Such was true for even for the much vaunted Quadra 840AV flagship AFAIK. Can't find anything definitive on LEM or everymac about SCSI implementations, just the fact that there were two (unified!) buses in the 900 and 950. Talk about SCSI Voodoo, what a nightmare that must have been, keeping SCSI IDs unique across two buses! To use the fast(?) SCSI bus for mass storage in the 950, were you required to have the WGS card installed, negating its use for the Mac OS?

Why didn't Apple provide fast SCSI-2 for the internal drives of the x100 series? PPC was supposed to be revolutionary and they hobbled those machines with a horrid disk I/O bottleneck. At that point in time, such was even worse than what they did in lobotomizing the 512K prototypes the Mac team developed. Love the fact that they snuck the support for 512K into the 128K board right under the very noses of Scully & Cost Cutting Company.

Dunno, but this stuff is really interesting to me. All kinds of info on implementations and comparisons with other platforms are coming out of the woodwork!

 
Dunno, but this stuff is really interesting to me. All kinds of info on implementations and comparisons with other platforms are coming out of the woodwork!
I am going to try to document the speed differences between SCSI and IDE on my Quadra 630. I will also just for fun test SCSI on the SE/30, not expecting anything great there, but its worth a try. Anyway, I am hoping to get people to follow the steps I did for bench marking on their machines, so we can get some numbers to work with. Mostly I want to get someone to try a Q950, some other Quadra, and a Power Mac with SCSI, so we can compare some numbers. Certainly some of the numbers will be HDD dependent, but we should be able to get at least some notion of how the different storage mediums compare. A lot of people have the SCSI2SD V5, and while that is SD-card dependent, it gives a someone more consistent test drive. :)

Apple never provided for fast internal drives!
I was going to mention that earlier, about the only internal drive controller that could be called 'fast' was the WGS95 PDS card, but that only works with A/UX! :p I wonder how popular NuBus based storage options were. I know they were "popular", but I just am wondering what percentage of people might have had one.

Love the fact that they snuck the support for 512K into the 128K board right under the very noses of Scully & Cost Cutting Company.
I do too, its too bad that other 'slips' like that didn't happen with other RoadApple designs. Though, some similar stuff happened later. I really do believe that the Q630 was designed to run at 40MHz, but since it was designated to be a lower-end machine, it was forced (probably by management! ;) ) to run at 33MHz. Simply move two resistors to get the clock-generator to output 40MHz, and you are good to go! :D

 
Last edited by a moderator:
Back
Top