Category talk:Early computers

'Early' and/vs. 'Individual' computers

Categories of Computers

I added the Early Computers category to EDVAC, and someone removed it. I don’t mind that, because of the way the Early Computers category is currently defined. However, I think that we need to address some problems.

Currently “Individual computers” includes the early one-of-a-kind computers, plus some newer ones and “early computers” includes machines that were made in larger quantaties. I think that someone looking for the earliest electronic computers (ENIAC, EDVAC, EDSAC, IAS, etc) would expect them to be in the “Early Computers” category. They wouldn’t think of looking for “Individual Computers”. Yes, someone could get to the rest of the early computers listed under “Individual computers” - if they knew the name of one of them, but they might not know. If someone found “early computers” they would miss the pioneering computers.

I propose some possible solutions:

(1) Categories of “Early one-of-a-kind (or unique) computers”, “Early mass-produced computers”, and move the non-early individual computers to another category.
(2) Keep “individual computers” category defined as it is and rename it to “unique computers” or “one-of-a-kind computers”, rename “early computers” to “Early mass-produced computers”, and create the “Early one-of-a-kind (or unique) computers” category, and put the early unique computers in both “individual” and “early unique”.
(3) A broad “Early Computers” category, with two sub-categories - one-of-a-kind and mass-produced.
(4) “Early Computers” could include all early computers, one of a kind or mass produced.

Finally, as far as what constitutes “early”, I’d say before the IBM 360 (in 1964). --Bubba73 00:32, 2 Jun 2005 (UTC)

After thinking about the issue (me being the 'someone' above), I agree with you that the current naming scheme of these categories is highly problematic; in general, we certainly need to define and name WKP categories in the most commonsensical-logical way possible.

As for your suggestions I think the second one might be the best, but the term 'mass-produced' would be quite unprecise---since most early non-individual computers were manufactured in very low numbers, typically, say, some tens of machines. I therefore suggest we use the category name 'Early serially-produced computers'. BTW, the term 'unique' might be sliiiightly imprecise, due to the identical name a job control language of 1970s vintage, but I guess the mix-up possibility would be minuscule if at all existent. Perhaps we should go with 'individual' after all (or 'singular'?).

(Also, your definition of 'early' corresponds quite accurately with the one used so far; the S/360's introduction year seemed a reasonably good dividing line due to the general obsolescense of the concept of (even intravendor-designed, including, of course, IBMs own) incompatible systems that it inaugurated.) --Wernher 03:42, 3 Jun 2005 (UTC)

Thank you for your useful comments. "Individual" computer just doesn't capture what the category is about, since any particular computer is an individual computer. At this point I don't know anything about redoing categories, so I'll leave that up to someone else, and I hope to get more feedback too. Bubba73 02:38, 6 Jun 2005 (UTC)

After yet more contemplation regarding the category names, I now tend toward favouring "Early on-of-a-kind computers" and "Early serially-produced computers". Some 'research' using Google suggests that the "one-of-a-kind" term is the only one of our options that unambigously describes the machines in question. As for the other category, the variation without the hyphen is by far the more common term---but that may well be due to the ubiquity of U.S. English documents on the web as compared with the number of British English ones (?). Tell me what you think. --Wernher 10:20, 14 Jun 2005 (UTC)

I just notified the "Individual computers" category's creator about our discussion, in the hope of obtaining his agreement to a speedy renaming of that category (to, say, "One-of-a-kind computers"). If we aren't able to get his attention (he hasn't been contributing since 9 months ago), we have to list the category at the Categories for deletion page, specifying that we want the category to be moved/renamed rather that just deleted. As for the practical work of renaming as such, we can do it ourselves (I'm an admin).

Now, if we succeed with this, we'll have two top-level cats: "Early computers" and "One-of-a-kind computers". We then have several alternatives for cat'ing the machines in question. The simplest one might be to put the one-of-a-kind early computers into both those cats, and simply leave the serially-produced early computers as is in the "Early computers" cat (thus basing our cat scheme on the implicit fact that computers generally are serially/mass produced). Other alternatives would entail the creation of various numbers of subcats. --Wernher 07:33, 19 Jun 2005 (UTC)

I think that "one-of-a-kind" should be hyphenated, but I'm not certain. Another possible catagorization could be "first-generation electronic computers" (vacuum tumes) rather than just "early". And then they could be broken down into one-of-a-kind and serially produced. And then the discrete transistor computers could go under "second generation".

One slight problem with the 1-of-a-kind versus serially-produced, is the case of ORDVAC and ILLIAC I. They were twins, so does that count as serially-produced, even though they have different names? Some of the serially-produced ones had as few as three units, I think.

I think we should go forth with the following scheme for the time being: A) we'll rename 'Individual computers' to 'One-of-a-kind computers' and see to it that all the early computers now present in the 'Individual computers' cat are also put into the 'Early computers' cat (i.e., contrary to the present inclusion criteria for the two cats in question, but instead just like you argued in the beginning of this thread); B) as I indicated in my previous comment, we'll not create a separate cat 'Serially-produced computers', since all non-one-of-a-kind machines are, by definition, serially-produced ones.

The reasons I think we should perhaps avoid having "generational" categories of computers are that it would force us to decide on this quite controversial issue for each and every early computer; also, it entails using that scheme for more recent machines, for which the issue of a computer belonging to one generation or the next is usually even more hotly contested (as are indeed the very definitions of the generational spans during the last 30 years or so). --Wernher 2 July 2005 21:13 (UTC)

Your suggestions sound good to me. Bubba73 3 July 2005 00:00 (UTC)

Thanks; now let the world know: please vote YES to have the renaming done! --Wernher 3 July 2005 00:27 (UTC)