Talk:Apollo Computer/Archives/2013

Merge Request
Perhaps this node should be called "Apollo computers, Inc" instead. -- mace

There are separate "Apollo Computers" and "Apollo Computer" articles. These ought to be merged. Dyl 12:55, Aug 17, 2004 (UTC)

I agree this needs to be fixed. The main article should be kept under whatever the actual name of the company was. If it was Apollo Computer, then Apollo Computers should have a redirect, or vice versa. It's nonsensical to click on a link to Apollo Computer, be redirected to Apollo Computers, and have the article be about Apollo Computer again. -- Kadin2048 20:32, 9 December 2005 (UTC)
 * It seems the plural now links here as it should. Generally I would say keep the "Inc." off unless there is another topic called Apollo Computer that is not incorporated. Now about the name, why was it called Apollo? What most of us assumed was it was a pun on the SUN workstation, since Apollo was the Sun god. Might have not been explicit, but might be mentioned. W Nowicki (talk) 19:57, 3 May 2011 (UTC)

When you say "most of us" to whom are you referring? It has never occurred to me that the name Apollo was somehow a reference to Sun Microsystems. SUN did not exist when Apollo was founded. I think it was just a name. Thomas144 (talk) —Preceding undated comment added 16:13, 22 September 2012 (UTC)

Unmatched sysadmin needs?
and low sysadmin-to-machine ratio that is still unmatched

I seem to recall that GM ran a HUGE distributed network of NeXT machines from a department consisting of two people. Much of this was due to NetInfo. I find it difficult to believe the Apollo was better. Maury 22:08, 16 August 2005 (UTC)
 * Ah, but Domain was seven years earlier! Need to state the facts without point of view. "Earlier" can be backed up with sources while "unmatched" is opinion and does not belong. We do need to wikilink and compare to the Single system image and cluster (computing) concepts. W Nowicki (talk) 19:57, 3 May 2011 (UTC)

>>> I worked at Apollo for 6 years and never knew if we had sys admins, if we did. I also never had a password the whole time.

Decline and Fall?
This whole article is kind of sophomoric, and a little dated. It's true that Sun basically destroyed Apollo but the question is why did customers like Computervision choose Sun equipment and not Apollo equipment? I think it had more to do with Sun's willingness to open up its system and sell boards and basically commodity software (Unix) when Apollo was still selling closed systems running proprietary software. The article makes it sounds like Sun products were more innovative, which is at best debatable (and in my opinion simply not true).

Also, this is the first I've ever heard of a dress code at Apollo (it's possible that Vanderslice introduced something like that in corporate offices, but I don't know), and I can't think of "many" engineers who left for any reason.


 * Alas, the Saxenian article is not unbiased. She clearly is saying why Silicon Valley is "better". The irony is that 18 years after her paper Sun is now defunct and all the jobs have moved to China and India. So needs to be carefully positioned as someone's opinion at a given time. W Nowicki (talk) 19:57, 3 May 2011 (UTC)

Dress code and Sysadmins
I worked in Engineering 1983-1988 and can verify that we never had a dress code.

NeXT may have been easier to administer (I don't know) but it came along later so is not relevant to the Sun vs Apollo comparison.

We didn't really have sysadmins, but we had a small staff that kept the network infrastructure going. I think there were three people supporting several hundred engineers.--Rees11 16:43, 21 September 2006 (UTC)

Not that anyone cares anymore, but...
I removed a short paragraph about the Apollo User Interface being difficult to use (something about hidden control characters). I always thought the Apollo DM was entirely obvious and intuitive (unlike say, a Macintosh which I find utterly baffling to this day). -Tom Gross (Apollo 1985-1990).


 * Right, anything is "easy" for those who know it and not easy for those who don't. Opinion. W Nowicki (talk) 19:57, 3 May 2011 (UTC)

Unreferenced
This article is largley unreferenced (one wallstreet article from the 1980's doesn't cut it); tag added appropriately. I also have a problem with some of the claims in this article, such as proprietary token-ring network. IBM was a huge proponent of token ring back in the 80's and early 90's -- it was by no means proprietary, unless the article means a different token implementation entirely (which should be explained, if so). Anyone familiar enough to help cleanup? /Blaxthos 07:09, 10 April 2007 (UTC)

Apollo token ring was different from IBM, although both were based on the earlier Cambridge token ring. It was indeed what we would now call proprietary, in that it was not well documented enough for anyone else to implement compatible hardware, and probably protected by patents. I don't have the time or inclination to track down references and re-write this article, but if anyone else does, I have a list of source material here: University of Michigan Apollo Archive: papers Rees11 16:14, 10 April 2007 (UTC)

If it's the Cambridge (England) token ring then that was actually a slotted token ring, which worked differently to the Apollo one. In fact I think that Apollo took the work that had been done at Prime on token ring and brought it over to Apollo. Texas Instruments, one of the accounts that I looked after in the UK, implemented the IBM Token Ring on their Apollo network which would indicate that it wasn't available at the outset of the Apollo design. 86.151.221.172 (talk) 14:49, 1 February 2011 (UTC)


 * Right, there were a bunch of incompatible LAN technologies. There were even a few non-standard Ethernet based ones too. Clearly needs to be clarified. Moreover, the whole protocol stack was not compatible with TCP/IP as noted. "Better" perhaps in some ways, but needs to be cited and stated as an opinion. Alas, the docs at Michigan do not seem there any more? W Nowicki (talk) 19:57, 3 May 2011 (UTC)

One individual?
The article says: "In 1988 the company incurred large losses in currency speculation, apparently due to the trading activities of one individual", and references John Markoff's 1988 NYT article.

I read through that article, and it's possible I'm missing something, but I see nothing of the sort mentioned there. The only similarity is a remark that "Last year, the company ran into financial problems because of mistakes related to its foreign currency trading." But that would make it a 1987 loss, right?, and there's no word that it was due to "one individual". —Preceding unsigned comment added by 64.81.170.62 (talk) 23:30, 7 May 2008 (UTC)


 * I added the reference to the NYT story because the statement had been unattributed. I believe you're right about the year and I'll change that. I was working at Apollo at the time and I do remember it being one individual but have no source for that right now. Rees11 (talk) 21:01, 8 May 2008 (UTC)
 * No source means must be removed. W Nowicki (talk) 19:57, 3 May 2011 (UTC)

3RCC - "me-toos"
I'd not have included Three Rivers: (a) it was much smaller than the three listed, (b) for much of its "life" it had no independent existence, being staffed by employees of Carnegie Mellon, (c) has been defunct much longer than those. Tedickey (talk) 00:42, 4 December 2008 (UTC)


 * Actually I would have included Three Rivers before Sun, because Sun came along later. But I'm probably biased. It's less significant than Sun or Symbolics. I remember there being a bunch of little graphic workstation companies at the time but can't name any others now. Rees11 (talk) 02:17, 4 December 2008 (UTC)


 * yes - there were a bunch (might be interesting to construct a category for those). But for the given context, I think that the best presentation would be to limit the list to comparable companies - and Sun and Symbolics fit that, while 3RCC doesn't.  (The topic for 3RCC says they were started 1974, but doesn't say at what point they actually setup offices that weren't inside someone's CMU office - my recollection is that it was around 1980, but of course that's not a reliable source ;-). Tedickey (talk) 11:20, 4 December 2008 (UTC)


 * Agreed. Suggest you wait a couple days then if no one objects remove 3rcc. Rees11 (talk) 15:54, 4 December 2008 (UTC)


 * sounds good Tedickey (talk) 16:46, 4 December 2008 (UTC)


 * Actually I think the PERQ should be mentioned in the history section, since it actually came out before. It was even more like the Alto, that is, it was a networked workstation, but not using a standard processor. W Nowicki (talk) 19:57, 3 May 2011 (UTC)

citations needed
For the citation needed for linking Apollo with DSEE with Clearcase, I was one of the developers at Pure/Rational and can unofficially confirm this sequence of events. However, being very new to editing wikipedia I did not want to edit the main page directly. Please forgive if this is also not the right forum or format for this.

A quick search brought up three references which might suffice for this citation (and perhaps the citations needed for IBM_Rational_ClearCase):


 * Software Configuration Management using Vesta pp. 165-167
 * A copy of the FAQ for the comp.software.config-mgmt group here: http://www.daveeaton.com/scm/CMFAQ.html
 * http://www.wordiq.com/definition/ClearCase

None are greatly satisfying, but perhaps good enough for now, or at least a direction to look?

Space E One (talk) 22:14, 15 December 2010 (UTC)

dual CPU design
The paragraph mentioning dual CPUs in the history paragraph is misplaced, since there is no other discussion of it, aside from what one might infer from the table which shows two CPUs for some models. The table itself lacks any mention of when the 2-CPU workstations were introduced. Lacking dates, the history section is vague. TEDickey (talk) 21:26, 28 December 2010 (UTC)


 * I would say these are details that belong in the Apollo/Domain article since that is about the hardware, while this is about the company (e.g. management, history, demise, etc.). So should move it there. W Nowicki (talk) 19:57, 3 May 2011 (UTC)

POSIX Compliance
The paragraph discussing the first Apollo computer states that it came out in 1981 and implies that it ran a POSIX compliant operating system (or frontend... not clear what that means). POSIX released their first standard in 1988 so at best this is just confusing. Maybe Aegis later became POSIX compliant? Needs clarification. Jrwjrw (talk) 05:06, 23 April 2011 (UTC)


 * Something like that. Apollo continued to improve their product until acquired by HP.  TEDickey (talk) 10:26, 23 April 2011 (UTC)


 * Generally the lede needs to be just a summary of the body. Indeed, saying it was POSIX in 1981 is just wrong, and a detail anyway. There is way too much overlinking too. Just wikilink the first time (or first time in a long section). W Nowicki (talk) 19:57, 3 May 2011 (UTC)