n***@gmail.com
2010-11-07 12:04:40 UTC
Probably because it's too subtle and fuzzy, this hasn't been
discussed before: how did the amazing human-interface of
S3 and V4 originate ?
AFAIK 'oberon' was originated as a language, which was also
suitable for making an OS. So the OS & language doesn't
influence the human-interface that is evolved.
Since human interfaces are mostly independant of language
and OS. And usually best done by people with different skills.
I can't see that the OS & language naturally led to the HI.
Evaluating a HI has problems of subjective fuzziness.
So I'll just mention some serious problems of alternate
OS/HI that I've recently observed.
I wanted to test and hopefully use the 'said to exist'
public-domain electric-circuit-simulator: SPICE;
under linux. Has anybody else succeeded?
Such a project requires multiple files to be open
simultaneously, eg.:
* the application,
* the README/s,
* 2 or 3 instruction manuals,
* your own notes and log ......
Normal life with paper documents is a nightmare, where
even simple tasks can't be achieved without a computer.
Eg. "where's the exact wording [for legal reasons] of that
email about <the cat sat on the mat> ", needs the ability to
search a [preferably 'location and date-range' specified]list
of files.
Switching between displays/desktops to read and write the
various texts, while manipulating the application, is
intolerable. And the apparent convention of having
windows open at uncontrolled locations and of unknown
sizes, so that they obscure existing needed windows,
is a disaster, compared to ETH-O's control of location
and size of TextFrames.
BTW, when ETH-O's *DOES* become chaotic, and have
essential Frames covered/hidden; would it be difficult to
<move a hidden frame to the top> via:
= traverse the display-space and list the names of the
frames [in a popup if need be]
= select the required frame to be 'rolled' to the top;
my merely re-arranging the display-space's links ?
Why would you want to simulate the choas of real-life:
papers of different sizes arranging themselves chaotically
to cover you desk; when you can have the 'AI' of
computerisation, keep your documents uniformly stacked
like a skilled card-shark would achieve after years of
practice?
A common task that I often need to do, and which forces
me to use LinuxETHOberon [LEO] is extract URLs from
documents which might have 80 lines of:
"descriptive text with [7] more text [8].."
followed by 30 lines of:
1. <URL/hot-link corresponding to text above>
...
7. <URL/hot-link corresponding to "[7]" above
8. <URL/hot-link corresponding to "[8]"above.
For this task, you need 3 textFrames:
1. to read: descriptive text with [7] "
2. to, having read the text, and decided that you want
the corresponding URL/hot-link:
7. <URL/hot-link corresponding to "[7]" above
3. the destination where you accumulate the list of
extracted URLs/hot-links.
1 & 2 are different locations on the same textFrame,
and 3 could be at to a separate 'file'.
When I discuss this missing facility in linux
[I won't talk about Micro$loth], they say "use screen".
No!! I don't want to comit and pre-order a facility
to see multiple textFrames; I want to have the facility
immediately available IF/WHEN I need it.
I don't want to have another set of commands to
remember; when I can just look and read COPY.
I don't want a facility that splits the screen in half;
when I can split it to any proportion and I can do it
visually/manually/intuitively/flexibly ...
So how did ETH-O come to have such a superior
HI?
Thanks,
== Chris Glur.
PS. what is/was the aim of Bluebottle/A2 ?
IIRC it had something to do with faster multi-tasking.
IMO that's chasing the wrong goal.
Someone was talking about his 'advanced' x64, but
faster CPUs, which need their own air-conditioners
is usually the wrong goal; like the US auto industry
of the late 50s. OTOH an ARM device, which could
not-lose/restore its tasks after power-down would be
advanced.
discussed before: how did the amazing human-interface of
S3 and V4 originate ?
AFAIK 'oberon' was originated as a language, which was also
suitable for making an OS. So the OS & language doesn't
influence the human-interface that is evolved.
Since human interfaces are mostly independant of language
and OS. And usually best done by people with different skills.
I can't see that the OS & language naturally led to the HI.
Evaluating a HI has problems of subjective fuzziness.
So I'll just mention some serious problems of alternate
OS/HI that I've recently observed.
I wanted to test and hopefully use the 'said to exist'
public-domain electric-circuit-simulator: SPICE;
under linux. Has anybody else succeeded?
Such a project requires multiple files to be open
simultaneously, eg.:
* the application,
* the README/s,
* 2 or 3 instruction manuals,
* your own notes and log ......
Normal life with paper documents is a nightmare, where
even simple tasks can't be achieved without a computer.
Eg. "where's the exact wording [for legal reasons] of that
email about <the cat sat on the mat> ", needs the ability to
search a [preferably 'location and date-range' specified]list
of files.
Switching between displays/desktops to read and write the
various texts, while manipulating the application, is
intolerable. And the apparent convention of having
windows open at uncontrolled locations and of unknown
sizes, so that they obscure existing needed windows,
is a disaster, compared to ETH-O's control of location
and size of TextFrames.
BTW, when ETH-O's *DOES* become chaotic, and have
essential Frames covered/hidden; would it be difficult to
<move a hidden frame to the top> via:
= traverse the display-space and list the names of the
frames [in a popup if need be]
= select the required frame to be 'rolled' to the top;
my merely re-arranging the display-space's links ?
Why would you want to simulate the choas of real-life:
papers of different sizes arranging themselves chaotically
to cover you desk; when you can have the 'AI' of
computerisation, keep your documents uniformly stacked
like a skilled card-shark would achieve after years of
practice?
A common task that I often need to do, and which forces
me to use LinuxETHOberon [LEO] is extract URLs from
documents which might have 80 lines of:
"descriptive text with [7] more text [8].."
followed by 30 lines of:
1. <URL/hot-link corresponding to text above>
...
7. <URL/hot-link corresponding to "[7]" above
8. <URL/hot-link corresponding to "[8]"above.
For this task, you need 3 textFrames:
1. to read: descriptive text with [7] "
2. to, having read the text, and decided that you want
the corresponding URL/hot-link:
7. <URL/hot-link corresponding to "[7]" above
3. the destination where you accumulate the list of
extracted URLs/hot-links.
1 & 2 are different locations on the same textFrame,
and 3 could be at to a separate 'file'.
When I discuss this missing facility in linux
[I won't talk about Micro$loth], they say "use screen".
No!! I don't want to comit and pre-order a facility
to see multiple textFrames; I want to have the facility
immediately available IF/WHEN I need it.
I don't want to have another set of commands to
remember; when I can just look and read COPY.
I don't want a facility that splits the screen in half;
when I can split it to any proportion and I can do it
visually/manually/intuitively/flexibly ...
So how did ETH-O come to have such a superior
HI?
Thanks,
== Chris Glur.
PS. what is/was the aim of Bluebottle/A2 ?
IIRC it had something to do with faster multi-tasking.
IMO that's chasing the wrong goal.
Someone was talking about his 'advanced' x64, but
faster CPUs, which need their own air-conditioners
is usually the wrong goal; like the US auto industry
of the late 50s. OTOH an ARM device, which could
not-lose/restore its tasks after power-down would be
advanced.
