G
Gene E. Bloch
Flightless Bird
On Tue, 10 Aug 2010 18:00:52 -0500, Char Jackson wrote:
> On Tue, 10 Aug 2010 15:01:26 -0700, Ken Blake
> <kblake@this.is.invalid.com> wrote:
>
>>On Tue, 10 Aug 2010 11
66 -0700, "Gene E. Bloch"
>><not-me@other.invalid> wrote:
>>
>>
>>> As for the origin of "thread", if it's not just the obvious meaning and
>>> analogy, as in threading through a maze or threading a string of beads,
>>> then I am one who doesn't know the reason for calling them threads, and
>>> would like to learn it.
>>
>>
>>OK. It has to do with list processing. "List" is a technical term and
>>isn't just the ordinary use of the word. A list is a bunch of data
>>that has more than one sequence as follows:
>>
>>1. The physical sequence of the records.
>>
>>2. One or more logical sequences that are different from the physical
>>sequence.
>>
>>So the messages in a newsgroup are a list. Their physical sequence is
>>the order in which they were posted. Their logical sequence is the way
>>messages in a "thread" are connected--e. g. this message follows
>>yours, the one I'm replying to, even though there are usually several
>>messages between them in the physical sequence.
>>
>>So visualize each message as a shirt button, and think of these
>>buttons as parallel to each other in a row arranged in physical
>>sequence (date and time). Then think of a thread (this thread, for
>>example) as having all the buttons pertinent to it as connected by a
>>piece of thread that goes from buttonhole to buttonhole, skipping the
>>buttonholes in the messages that are not part of the thread.
>>
>>So a "thread" *is* a thread.
>>
>>Also note that (although it's not pertinent in a newsgroup) the
>>buttons have multiple buttonholes and can therefore participate in
>>multiple threads (in other words, have more than one logical
>>sequence).
>>
>
> That looks to be an overly complicated, and overly wordy, restatement
> of Gene's (and my own) understanding.
Well, I have to admit that the button/buttonhole analogy confused me enough
that i just went back to my understanding of threaded lists to try to guess
what Ken meant
I don't see the buttonhole thing being very different from my remarks about
threading through a maze or threading beads on a string. I also see the
name "threaded", as in threaded lists, being derived from the more homely
examples such as the ones I cited, and "thread" as used in news threads
being derived from those same ideas, rather than a thread (pun intended) of
etymology going from the everyday examples to the threaded lists to the
newsgroup threads.
BTW, I have seen beadwork where there are several threads running through
the beads, with one thread perhaps going through beads B1 B2 B3 and another
perhaps through A2 B2 C2, if you get my drift. There would be only one hole
in each bead though, large enough to allow a couple of threads to pass
through. Yikes! I just remembered the structure of core memory (magnetic
cores, of course). It's a lot like what I just described...
I believe I may have programmed data structures involving several sets of
links, such as records threaded simultaneously by name and by date of
birth. Note that computer threaded lists use pointers instead of cotton
lisle thread, providing generous opportunities for bugs
--
Gene E. Bloch (Stumbling Bloch)
> On Tue, 10 Aug 2010 15:01:26 -0700, Ken Blake
> <kblake@this.is.invalid.com> wrote:
>
>>On Tue, 10 Aug 2010 11
66 -0700, "Gene E. Bloch">><not-me@other.invalid> wrote:
>>
>>
>>> As for the origin of "thread", if it's not just the obvious meaning and
>>> analogy, as in threading through a maze or threading a string of beads,
>>> then I am one who doesn't know the reason for calling them threads, and
>>> would like to learn it.
>>
>>
>>OK. It has to do with list processing. "List" is a technical term and
>>isn't just the ordinary use of the word. A list is a bunch of data
>>that has more than one sequence as follows:
>>
>>1. The physical sequence of the records.
>>
>>2. One or more logical sequences that are different from the physical
>>sequence.
>>
>>So the messages in a newsgroup are a list. Their physical sequence is
>>the order in which they were posted. Their logical sequence is the way
>>messages in a "thread" are connected--e. g. this message follows
>>yours, the one I'm replying to, even though there are usually several
>>messages between them in the physical sequence.
>>
>>So visualize each message as a shirt button, and think of these
>>buttons as parallel to each other in a row arranged in physical
>>sequence (date and time). Then think of a thread (this thread, for
>>example) as having all the buttons pertinent to it as connected by a
>>piece of thread that goes from buttonhole to buttonhole, skipping the
>>buttonholes in the messages that are not part of the thread.
>>
>>So a "thread" *is* a thread.
>>
>>Also note that (although it's not pertinent in a newsgroup) the
>>buttons have multiple buttonholes and can therefore participate in
>>multiple threads (in other words, have more than one logical
>>sequence).
>>
>
> That looks to be an overly complicated, and overly wordy, restatement
> of Gene's (and my own) understanding.
Well, I have to admit that the button/buttonhole analogy confused me enough
that i just went back to my understanding of threaded lists to try to guess
what Ken meant
I don't see the buttonhole thing being very different from my remarks about
threading through a maze or threading beads on a string. I also see the
name "threaded", as in threaded lists, being derived from the more homely
examples such as the ones I cited, and "thread" as used in news threads
being derived from those same ideas, rather than a thread (pun intended) of
etymology going from the everyday examples to the threaded lists to the
newsgroup threads.
BTW, I have seen beadwork where there are several threads running through
the beads, with one thread perhaps going through beads B1 B2 B3 and another
perhaps through A2 B2 C2, if you get my drift. There would be only one hole
in each bead though, large enough to allow a couple of threads to pass
through. Yikes! I just remembered the structure of core memory (magnetic
cores, of course). It's a lot like what I just described...
I believe I may have programmed data structures involving several sets of
links, such as records threaded simultaneously by name and by date of
birth. Note that computer threaded lists use pointers instead of cotton
lisle thread, providing generous opportunities for bugs
--
Gene E. Bloch (Stumbling Bloch)