OT: Google's latest dungheap

[Default User]

I understand your response, but ... the author of any kind of
software has to make assumptions about the speed and capability of
the target computer. As computers get faster and more capable, you
can't continue to support ancient hardware indefinitely. A 333Mhz
machine is about at the reasonable cutoff point.

Absolutely ludicrous. We're talking a fast game here, a damn web page.
A poorly written web page at that. There's nothing on there that
requires any complex processing or a fast computer.




Brian

 

[Guest]

This was on topic why?

Who claimed it was on topic? Certainly not the OP, considering the
subject line.

 

[Chris Dollin]

Who claimed it was on topic? Certainly not the OP, considering the
subject line.

It was posted here, and labelling something OT on the subject
line doesn't excuse an off-topic posting. I'm bemused by the
number of equally off-topic responses the post gathered. It's
not as though it's /almost/ on topic. We may hate what Google
does to newsgroups, and some bitching in asides isn't that
bad -- but starting a /new topic/ for the purpose?

How about I make posts on Eurogames or RDF or my favourite
prog-rock bands? I'll label them OT, so that will be OK, OK?

 

[Guest]

It was posted here, and labelling something OT on the subject
line doesn't excuse an off-topic posting.

My point is that the right question to ask is "This was appropriate
for this newsgroup why?" Pointing out that it's off-topic when the OP
clearly already knew that is pointless. The OP probably felt it was
appropriate for this newsgroup despite being off-topic.

 

[Chris Dollin]

My point is that the right question to ask is "This was appropriate
for this newsgroup why?" Pointing out that it's off-topic when the OP
clearly already knew that is pointless. The OP probably felt it was
appropriate for this newsgroup despite being off-topic.

Yes, that would have been a better question (and more polite than
what I /nearly/ wrote instead of twoty).

 

[Christopher Benson-Manica]

I've never really liked newsreaders that I've tried.

You might try tin if you haven't. I thoroughly despise every
graphical newsreader I've ever used, including Google. It certainly
obviates the need for a silly "preview" feature; any interface that
would make me wish for a "preview" feature is already fatally broken
IMO.

- All messages are available for a NG

Indeed, the only real point to using the Google interface for
anything, ever.

 

[Christopher Benson-Manica]

Actually, you can, but...

Why not?

....when the average hardware is substantially better than the ancient
hardware, the loss of the small population of ancient hardware users
is more than offset by the gain in average hardware users who want
more bells and whistles. Google doesn't care about the technically
inclined anyway - their raison d'etre is to swamp as many people with
sidebar ads as possible - so why should they trouble themselves to
placate those of us with the temerity to continue running a 333 MHz
machine?

It isn't acceptable to abandon older hardware if your customers require
otherwise.

We aren't the target customers, unfortunately, as Google has made
abundantly clear.

 

[Richard Heathfield]

Christopher Benson-Manica said:

We aren't the target customers, unfortunately, as Google has made
abundantly clear.

Oh, I wasn't talking about Google Groups. Surely nobody who has any choice
in the matter would *choose* to use Google Groups for anything other than
looking up old messages? No, I was addressing the general issue, not the
specific instance.

In C terms, if a C program runs acceptably fast on a slow machine, then
it'll be blitzingly fast on a modern box. I don't see a problem with that.
We used to be able to write programs that ran acceptably fast on slow
machines, so I can see no reason why we can't still do that.

 

[Charlton Wilbur]

RH> In C terms, if a C program runs acceptably fast on a slow
RH> machine, then it'll be blitzingly fast on a modern box. I
RH> don't see a problem with that. We used to be able to write
RH> programs that ran acceptably fast on slow machines, so I can
RH> see no reason why we can't still do that.

Users demand that computers do more, and computers have the capability
to do more so operating system designers and packagers run with it.

Fifteen years ago I and 50 to 100 other people were happy with a
computer that had less processing power, disk space, and memory than
the iPod I have in my pocket. I was very productive with that
computer, but I think if I went back to it today, I'd find it
cripplingly primitive.

Charlton

 

[Richard Heathfield]

Charlton Wilbur said:

RH> In C terms, if a C program runs acceptably fast on a slow
RH> machine, then it'll be blitzingly fast on a modern box. I
RH> don't see a problem with that. We used to be able to write
RH> programs that ran acceptably fast on slow machines, so I can
RH> see no reason why we can't still do that.

Users demand that computers do more, and computers have the capability
to do more so operating system designers and packagers run with it.

Perhaps we are at cross-purposes. I'm talking about code that is topical
within comp.lang.c - hence the "In C terms" and "a C program" in the above
quote. What do you think clc-conforming code can achieve on a modern
machine (other than performance) that it can't achieve on an older machine?
Modulo C99, nothing has actually changed in language terms. No sound or
graphics rendering, no network connectivity, nothing like that. So I stand
by my original point.

Fifteen years ago I and 50 to 100 other people were happy with a
computer that had less processing power, disk space, and memory than
the iPod I have in my pocket. I was very productive with that
computer, but I think if I went back to it today, I'd find it
cripplingly primitive.

Nobody is asking you to go back to that. But just because I have a fast
computer (and the one on which I'm typing this is theoretically thousands
of times faster than the Atari ST that was the first "personal computer" I
ever owned), that doesn't mean I have to insist that everybody else should
buy a computer as fast as mine just so that they can use my programs.

 

[Guest]

Christopher Benson-Manica said:

Oh, I wasn't talking about Google Groups. Surely nobody who has any choice
in the matter would *choose* to use Google Groups for anything other than
looking up old messages?

Old Wolf has stated he has tried newsreaders, but hasn't found one he
liked. I am in the same position. There are people who do have choice
in the matter that choose to use Google Groups.

 

[Charlton Wilbur]

RH> Charlton Wilbur said:

RH> Perhaps we are at cross-purposes. I'm talking about code that
RH> is topical within comp.lang.c - hence the "In C terms" and "a
RH> C program" in the above quote. What do you think
RH> clc-conforming code can achieve on a modern machine (other
RH> than performance) that it can't achieve on an older machine?

Very, very little, if anything at all. But the subset of tasks that
can be usefully accomplished while limiting oneself to C within the
limits of clc-topicality is also vanishingly small, and largely of
interest to academicians and pedants. There is value in focusing only
on the portable and Standard-conforming aspects of C, or I wouldn't be
here, but there's more to C as she is spoke than that.

To do almost anything practically useful with C, you have to use
non-standard extensions. While specific extensions are off-topic
here, surely it is possible to acknowledge their existence!

RH> But just because I have a fast computer (and the one on which
RH> I'm typing this is theoretically thousands of times faster
RH> than the Atari ST that was the first "personal computer" I
RH> ever owned), that doesn't mean I have to insist that everybody
RH> else should buy a computer as fast as mine just so that they
RH> can use my programs.

No, but even if we limit ourselves to strict clc-topicality -- for
instance, a program that reads a file, performs some extensive
calculation on the data therein, and writes the result to another file
-- the nature of the computer on which the program is to be run has an
effect on the style of the program.

For instance, on that ancient computer, I wrote a ray-tracing program.
It read the specifications for the objects from a file and produced a
graphics file in a particular format based on the results. If I had
to write that program again, I'd write it very differently if I knew I
had to write it for a 15-year old Unix machine with 50 users than if I
was writing it for the computer I'm sitting at now.

At some point there's a tradeoff; I could, of course, take the program
I wrote yesterday and tweak it to allow it to run adequately on an
Atari ST. Practically speaking, if someone is paying me to do this,
the simple fact of comparing my hourly rate to the cost of a new
computer is enough to convince them otherwise, unless they're really
committed to that Atari ST.

Charlton

 

[Christopher Benson-Manica]

In C terms, if a C program runs acceptably fast on a slow machine, then
it'll be blitzingly fast on a modern box. I don't see a problem with that.
We used to be able to write programs that ran acceptably fast on slow
machines, so I can see no reason why we can't still do that.

It's always true, however, that there are some values of "slow" and
"acceptably fast" where a given program will no longer run
"acceptably fast" on a "slow" machine. Even a skillfully written
program may fail to perform acceptably when run on a 486 machine, and
there is nothing to be done about the situation. At some point the
possibilities for improving a program's functionality will outweigh
the benefits of supporting the small number of users running on legacy
hardware X. (This is not to forgive Google for not providing a
minimalist alternative interface for their "service", however.)

 

[Richard Heathfield]

Charlton Wilbur said:

RH> Charlton Wilbur said:


RH> Perhaps we are at cross-purposes. I'm talking about code that
RH> is topical within comp.lang.c - hence the "In C terms" and "a
RH> C program" in the above quote. What do you think
RH> clc-conforming code can achieve on a modern machine (other
RH> than performance) that it can't achieve on an older machine?

Very, very little, if anything at all.
Right.

But the subset of tasks that
can be usefully accomplished while limiting oneself to C within the
limits of clc-topicality is also vanishingly small, and largely of
interest to academicians and pedants.

I very, very carefully restricted myself to discussing C *code* rather than
a C *program*. I accept that people want all-singing all-dancing programs
with lots of gooey, and that we will perforce use extensions to provide
such bells and whistles - but of course it is not only possible but
commonplace to provide C libraries which are, if not strictly conforming,
at least clc-conforming.

There is value in focusing only
on the portable and Standard-conforming aspects of C, or I wouldn't be
here,

Right (see above).

but there's more to C as she is spoke than that.

To do almost anything practically useful with C, you have to use
non-standard extensions. While specific extensions are off-topic
here, surely it is possible to acknowledge their existence!

Sure. For example, I recently had cause to write a program for a Windows
user for whom a console is something you find on a flight deck. The bulk of
the processing could be done in straight C, and so I wrote a module just
for that, which will run quite happily on Linux or an ST or a mainframe or
whatever - and I did the gooey bits separately. It all works swimmingly
well, and most of it is still portable. More importantly, the portable bit
(the bulk of the code) is in a separate source file, which makes porting it
astoundingly easy. Can we say "ftp"?

<snip>

 

[Charlton Wilbur]

(quoting me)

RH> Right (see above).

RH> Not 'ere, there ain't. We have <cough> other newsgroups
RH> for discussing such matters.

But the point I was addressing is that while C *code* is no more or
less efficient than it was 10 years ago, C *programs* are expected to
do considerably more. Thus it is likely to be impossible, difficult,
or at the very least more expensive than it is worth to port a program
written for a computer with current-day specifications for an older
computer.

Sometimes the best solution really is to admit that your computer is
out of date (or incapable of delivering what you need, for any of a
number of reasons) and to buy a new one.

Charlton

 

[Lane Straatman]

Old Wolf said:

Isn't it still a beta?

<shrug> So don't use it. End of problem.

One certainly wouldn't use google in lieu of, say, comcast groups or the 4
buck a month service I had before I had all this wonderful broadband. I
_did_ use google while my communications were temporarily kaputt, using
terminals at the public libraries. The knock on google posts that I had
serially witnessed was that they didn't quote context, making the whole
exchange as ephemeral as chat (that wasn't all bad back in C_Dreamer days at
yahoo). It turns out, context is easy to quote; it's the posters' lack of
usenet savvy or ability to move something thematic foreword that was the
real problem. LS

 

[Old Wolf]

The OP probably felt it was appropriate for this newsgroup
despite being off-topic.

Yes; many people read and post this NG using google (as
evinced by the number of erroneous posts that have occurred
in the past due to bugs in the interface).

Anyway, there were already several messages on this
topic, but they were buried deep in some other thread
with an un-related subject line, so I deemed it
appropriate to start a new thread (perhaps I could have
just continued to post on the old thread, but changed
the topic line).

 

[Default User]

Isn't it still a beta?

One certainly wouldn't use google in lieu of, say, comcast groups or
the 4 buck a month service I had before I had all this wonderful
broadband. I did use google while my communications were temporarily
kaputt, using terminals at the public libraries. The knock on google
posts that I had serially witnessed was that they didn't quote
context, making the whole exchange as ephemeral as chat (that wasn't
all bad back in C_Dreamer days at yahoo). It turns out, context is
easy to quote; it's the posters' lack of usenet savvy or ability to
move something thematic foreword that was the real problem. LS

The early release of the new Google Groups did not quote when using the
default reply button on messages. This lead to a period of time where
GG posts were characterized by having no context. There was a way to
quote, but it was not obvious to most, although the information
generally became available to those interested. Eventually Google fixed
this, although it took far too long in my opinion.

On the whole, the designers of the GG interface have shown a level of
knowledge of usenet that borders on ignorance or incompetence. It
shouldn't have been that difficult a task, the prior art was copious.




Brian

 

[Lane Straatman]

The early release of the new Google Groups did not quote when using the
default reply button on messages. This lead to a period of time where
GG posts were characterized by having no context. There was a way to
quote, but it was not obvious to most, although the information
generally became available to those interested. Eventually Google fixed
this, although it took far too long in my opinion.

On the whole, the designers of the GG interface have shown a level of
knowledge of usenet that borders on ignorance or incompetence. It
shouldn't have been that difficult a task, the prior art was copious.

These sound like the same lamentations every time MS comes out with a new
OS. It might have to do with being the dominant search engine like windows
is the dominant OS. Maybe they do it ... because they can get away with it.
LS

 

[Default User]

These sound like the same lamentations every time MS comes out with a
new OS. It might have to do with being the dominant search engine
like windows is the dominant OS. Maybe they do it ... because they
can get away with it. LS

We're not discussing the search page, although elsewhere I've seem
people complain about changes to the image search. We're discussing the
Google Groups newsreader. The story on that since they first rolled out
the "Beta" has been one of incompetence and ignorance, well seasoned
with a dose of "don't give a damn".




Brian