HP 48/9 vs TI-92?
jef...@my-deja.com
real-life experience? Which is better/faster for which functions?
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
sasq...@hotmail.com
jef...@my-deja.com wrote:
> Anybody have both? Or a web site which compares features? What about
> real-life experience? Which is better/faster for which functions?
>
I owned 2 HP48(i have lost the 2) and i actually own a TI92+.
I will buy a HP49,so what do you want to know ?
Please ask specific questions.
henr...@usa.net
> Anybody have both? Or a web site which compares features? What about
> real-life experience? Which is better/faster for which functions?
>
I'm a HP48 owner/fan so that will probably affect my judgement. :-)
The TI92+ has a more "computer-feel" over it when you see it at first,
because it has roll-down menus and windows, while the opposite actually
is more true. The HP48 has a systemclock and a real databus and uses a
PC standard transfer protocol for transferring software, things the
TI's don't have.
The TI92+ is easy to start using because it uses the familiar but
rather cumbersome algebraic entry system. Cumbersome, because I've used
the HP48 stack for years and have become quite addicted to it. It
requires fewer keystrokes for most operations.
A beginner with a HP48 will first have to learn how to use the stack,
which is why many are frightened of using it. It doesn't take long to
learn and appreciate it, though.
The TI92+ has some really powerful symbolic capabilities. Almost
anything you throw at it, can be integrated or simplified.
The HP48 has much stronger numerical computation technology, because
the Saturn processor is built for it. The 68000 processor in the TI92+
isn't. Things that may take minutes on a TI92+ may take seconds on a
HP48.
On general build quality the HP's are always winners. The TI's are
infamous for bad keyboards, and the buttons break or fall off. It
definitely feels alot cheaper and makes much more noise than a HP
keyboard. A HP keyboard, if treated correctly can last I believe more
than 30 years. I've tried 20 year old keyboards which feel just like
new ones.
The HP's are generally taken more seriously among engineers, because
it's so darn flexible. They can do almost anything, where TI's are
limited to the programming which they are born with. TI doesn't support
the internal mathematical routines of the software other than in the
ultra slow built-in BASIC. HP's can use them in ultra fast assembly.
Take a look at hp48.ml.org and enjoy the wealth of software for the
HP's. You'll never find such a mass for the TI's. Most TI programs are
kids games anyway. :-)
The TI92+ GUI is fast, much faster than on the HP48
Now the new HP49 is almost everything the HP48 isn't. It has a really
fast GUI and more symbolic capabilities than the TI92+. I haven't tried
the HP49 but I've heard it feels like using a superfast HP48. The
keyboard is different, but should be just as good in quality.
I heard that a TI92 guy had tried the HP49 and then felt ashamed by his
TI92 afterwards... :-) It's that good.
Also it seems to be much easier to get acquainted with than the HP48
because the HP49 supports both RPN and algebraic entry.
Conclusion: I can use a TI, but I'd much rather use my HP48. I can
drive a Toyota, but I'd rather drive an Aston Martin. :-)
As for me, I'm going to buy the HP49 when it comes out.
Actually in my experience when people select their calculator it's
really a matter of taste, and you can't really affect people on what
calculator to use before they have tried it. Some will continue to use
TI's because 'it's cool' and others will use HP's because they have
learned to appreciate the features and quality of the HP's.
Go to a bookstore and see if you can get to try them all.
sasq...@hotmail.com
> I'm a HP48 owner/fan so that will probably affect my judgement. :-)
>
> The TI92+ has a more "computer-feel" over it when you see it at first,
> because it has roll-down menus and windows, while the opposite
actually
> is more true.
I completely agree.
When i think of what TI could have done of the TI92...
For this result TI could have used an i80186.
>The HP48 has a systemclock and a real databus and uses a
> PC standard transfer protocol for transferring software, things the
> TI's don't have.
Yes.
> The TI92+ is easy to start using because it uses the familiar but
> rather cumbersome algebraic entry system. Cumbersome, because I've
used
> the HP48 stack for years and have become quite addicted to it. It
> requires fewer keystrokes for most operations.
Moreover RPN is better to deal with complex symbolic computations.
It is really a pain to edit big expressions with algebraic command line
even if the history of the TI92 has a few RPN like options.
> A beginner with a HP48 will first have to learn how to use the stack,
> which is why many are frightened of using it. It doesn't take long to
> learn and appreciate it, though.
>
> The TI92+ has some really powerful symbolic capabilities. Almost
> anything you throw at it, can be integrated or simplified.
However it fails on some very easy examples and it is very limited in
some areas.
Plus its power is screwed by some memory flaws.
> The HP48 has much stronger numerical computation technology, because
> the Saturn processor is built for it. The 68000 processor in the TI92+
> isn't. Things that may take minutes on a TI92+ may take seconds on a
> HP48.
In fact i think that it is due to both M68K weakness for BCD
computations and to its algorithms.
I think that with better algorithms the TI92+ could be at least as fast
as the TI85.
For example it is significantly faster than the TI85 for numeric eigvl
and eigvc computations probably because it uses the same algorithm.
> On general build quality the HP's are always winners. The TI's are
> infamous for bad keyboards, and the buttons break or fall off. It
> definitely feels alot cheaper and makes much more noise than a HP
> keyboard. A HP keyboard, if treated correctly can last I believe more
> than 30 years. I've tried 20 year old keyboards which feel just like
> new ones.
>
> The HP's are generally taken more seriously among engineers, because
> it's so darn flexible. They can do almost anything, where TI's are
> limited to the programming which they are born with. TI doesn't
support
> the internal mathematical routines of the software other than in the
> ultra slow built-in BASIC. HP's can use them in ultra fast assembly.
Well,it is one of the biggest weakness on the TI92+ but TI has announced
a SDK including a C compiler for January 2000.
We will wait and see.
> Take a look at hp48.ml.org and enjoy the wealth of software for the
> HP's. You'll never find such a mass for the TI's. Most TI programs are
> kids games anyway. :-)
>
Yes,it is really a shame but without support from TI the potential
software programmers can't do almost anything.
We will see if the SDK will change that.
> The TI92+ GUI is fast, much faster than on the HP48
The TI92+ is fast until a certain point.
In some cases it is incredibly slow.
> Now the new HP49 is almost everything the HP48 isn't. It has a really
> fast GUI and more symbolic capabilities than the TI92+. I haven't
tried
> the HP49 but I've heard it feels like using a superfast HP48. The
> keyboard is different, but should be just as good in quality.
> I heard that a TI92 guy had tried the HP49 and then felt ashamed by
his
> TI92 afterwards... :-) It's that good.
I am the one who has tried the HP49 and felt ashamed but i am not a TI92
guy.
I just own a TI92+.
> Also it seems to be much easier to get acquainted with than the HP48
> because the HP49 supports both RPN and algebraic entry.
It is also because the HP49 interface is much faster than the HP48
interface.
And i think that the new equation writer is from far the best way to
work on symbolic expressions.
It is simply marvelous.
Dean Lewis
about the HP's, but I can probably give a few biased oppinions about the
TI.
Overall I found the TI-92 to be extremely easy to use. The quality seems
to be pretty good overall (I've dropped mine more than once with no
breakage). My only complaint was that dirt would get under the screen
over time which required disassembling the calculator inorder to clean
it. I've always found the hard plastic keys to have a good feel to them
and to be of good quality (I've heard HP users complain of the TI keys,
though, so perhaps its what you're used to). The only complaint I've
consistently heard on the TI news group was the screen not being very well
protected and occasionally cracking from a fall. I've never had this
happen to me, but enough people have mentioned it to make me beleive that
there is something to it.
One feature I particularly like is how the history pairs are set up on the
92. When used properly, it makes algebraic entry much easier and more
efficient to use than on the other TI calculators (TI-82, 83, 85, 86). If
I read part of the 49 manual correctly, it will have a similar setup when
in algebraic mode. I can't imagine why someone would prefer RPN over
algebraic entry, but I suppose since so many HP-48 users swear by it,
there must be something to it. I'm definately going to try to give it a
fair chance when I get the 49, but it will probably be hard for me to
change my ways now.
Another feature I had particularly liked about the 92 is the TI-Basic.
Although speed is definately not one of its plusses, I could very easily
and quickly make simple (and complex) programs to help me out.
Occasionally I found it to be too slow, but 99% of the time it was more
than sufficient. Overall, TI Basic is extremely easy to use and easy to
learn. If you're needing more power and speed, the new SDK kit for the 92
looks promising, but it's hard to say how great it will be until it
becomes available at the end of the year.
My favorite feature on the TI-92 is the qwerty keyboard. I hear some
people say they love it, and some people say that it makes the calc too
bulky to carry around. It definately makes the calculator more bulky and
draws more attention than I would care for it to, but it makes programming
and just about anything else on the calculator much earsier to use. I
feel the keyboard is a big plus, but not everyone thinks that way.
One complaint I hear about the 92 is that the OS was written
inefficiently. In the fall, TI will be releasing a new version of the OS
for free that is supposed to be largely rewritten, so I'm sure there will
be improvements. Another plus is that it will raise the available ram to
over 600kb. I personally had never noticed any inefficiencies, but I
probably don't push my calc quite as hard as some people.
Overall I found the TI-92 to be very easy to use and very powerful. When
comparing the 92 to the 86 I used to have, I found that the ease of use of
the 92 made it much easier to use on a test. For me, that is a big plus.
I'm sure the 49 will end up being a remarkable machine. Hopefully both TI
and HP will keep trying to top eachother. There's nothing better than a
little friendly competition to keep better caclulators being produced by
both companies.
jef...@my-deja.com wrote:
> Anybody have both? Or a web site which compares features? What about
> real-life experience? Which is better/faster for which functions?
>
rse...@primenet.com
the keyboards, I prefer the 92, both because of its qwerty keyboard as
well as the feel of the keys. In terms of keys, I like the 92 keyboard,
followed by the 48 and then the 89. The weakness of the 92/89 is the
lack of RPN, lack of a reassignable keyboard, and the forced usage of
symbolic computation when it is probably undesirable as in the matrix
routines. I do not use the programming features on any of the
calculators extensively and as much as I prefer RPN for basic
calculation, I prefer algebraic languages for programming.
For my limited programming purposes, readability is more important than
efficiency / number of keystrokes. I have never had any problem with
dust under the keyboard and I have used the 92 in classrooms filled with
chalkdust, but I have a fairly late model 92 so TI may have improved the
seal over time.
A few notes - I think that HP should have continued to use a keyboard
layout similar to the current 48 and similar keys. I find elements of
frustration in both the TI and the HP models. For example, I do not
like the keystroke sequence for storing values in variables on the 48.
The 89 and the 92 are far better in my opinion. I much prefer the
storage sequence used on earlier HP calculators, which would require
adjustments for storage by variable name. On the 89/92, I can't
stand the fact that final closing parenthesis are not determined
implicitly as in all other TI calculators. Not being a student, the
standardized test issue is not important to me, but the portability
frequently makes the 48 or the 89 more appropriate than the 92.
I have presented my opinion on these matters on the TI-89/92Plus board
on the TI site, so you will see many of my comments there in older
[probably archived] postings.
There is an extensive body of literature comparing these calculators in
this newsgroup, in bit.listserv.calc.ti [I think that's the correct
name], and in the discussion groups on the TI site. And, of course,
all of the discussions are objective, well reasoned, and presented by
disinterested parties without strong opinions one way or the other. :)
In article <7lhhvm$2dd$1...@nnrp1.deja.com>,
Jeremy Fincher
>this newsgroup, in bit.listserv.calc.ti [I think that's the correct
>name],
actually it's bit.listserv.calc-ti :-)
Jeremy
----------------------------------
If i ever forget to capitalize a proper noun, forgive me. i'm a big fan of ee
cummings
My ICQ # is 28153190. My AIM/AOL name is either jemfinch02 or Cassius80.
Have a good day, and good luck in your endeavors!
rse...@primenet.com
Timité Hassan
> One complaint I hear about the 92 is that the OS was written
> inefficiently. In the fall, TI will be releasing a new version of the OS
> for free that is supposed to be largely rewritten, so I'm sure there will
> be improvements. Another plus is that it will raise the available ram to
> over 600kb. I personally had never noticed any inefficiencies, but I
> probably don't push my calc quite as hard as some people.
I have push the calc very far and the OS is very inefficient especially
regarding the power of the C.P.U.And i can give you some examples if you want.
The new OS will not bring significant improvements according to TI.
And it will not increased the RAM to 600 KB but the user Flash from 384 KB to
640 KB.
> Overall I found the TI-92 to be very easy to use and very powerful. When
> comparing the 92 to the 86 I used to have, I found that the ease of use of
> the 92 made it much easier to use on a test. For me, that is a big plus.
The TI92+ is very easy to use(even if i prefer RPN) but its power is very
questionnable.
Dean Lewis
although in this case I wish you were correct. I was using RAM to refer to all
the available memory, which is not the correct way to describe the memory. The
total memory on the calculator currently is 572 KB (RAM + user data archive).
Once the memory is rewritten, the total available memory will be over 600 KB of
both RAM and user data archive. I'm assuming that the available RAM will stay at
188 KB, so the user data archive would increase from 384 KB to over 400 KB.
The two main inefficiecies that I have heard concerning the TI-92 are how it
handles large matrices and the for loop, which are poor tests to evaluate the
whole calculator with. I've used the TI-92 since about the time it was released,
so perhaps I'm too biased now to look at it fairly. Either way, I just don't see
where it 92 is so inefficient. If you do want to list the inefficiencies that you
know about, please do. Once I buy the HP49G, I wouldn't mind using your list to
compare the calculators. They are both plenty fast for me, but it would still be
fun to compare them.
Timité Hassan
Dean Lewis wrote:
> Actually, I think we both worded the the available memory wrong for the TI-92+,
> although in this case I wish you were correct. I was using RAM to refer to all
> the available memory, which is not the correct way to describe the memory. The
> total memory on the calculator currently is 572 KB (RAM + user data archive).
> Once the memory is rewritten, the total available memory will be over 600 KB of
> both RAM and user data archive. I'm assuming that the available RAM will stay at
> 188 KB, so the user data archive would increase from 384 KB to over 400 KB.
In fact the user data archive alone will be increased to 640 KB so the total
availlable memory will be inreased to 828 KB.However the user data archive will be
used for both storage and applications.
If TI completely removed the geometry and made it availlable it as a Free apps for the
TI92+ owner is supposed that it would be possible to increase the user data archive
until over 896 KB.
> The two main inefficiecies that I have heard concerning the TI-92 are how it
> handles large matrices and the for loop, which are poor tests to evaluate the
> whole calculator with. I've used the TI-92 since about the time it was released,
> so perhaps I'm too biased now to look at it fairly. Either way, I just don't see
> where it 92 is so inefficient. If you do want to list the inefficiencies that you
> know about, please do. Once I buy the HP49G, I wouldn't mind using your list to
> compare the calculators. They are both plenty fast for me, but it would still be
> fun to compare them.
Oh they are a lot of inefficiencies of the TI92 OS.The TI92 uses too slow algorithms
for numeric matrix computations.
It is even slower than the TI86 for most numeric matrix computations.
The funny thing is that the eigvc and eigvl functions of the TI92+ which are 100 %
numeric are much faster than those of the TI86 probably because TI used the same
algoritms on the 2 calc.
So why TI has not implemented the TI86 algorithms for all the other 100 % numeric
matrix computations ?
If TI is so interested by incredible accuracy on the TI92+ they could have implemented
arbitrary accuracy for numeric computations(i don't talk of computations with
integers) after all it is a Derive feature.
The For loop of the TI92+ has been designed in such a way that it is incredibly slow.
For k,1,1000:Endfor runs in 17 s on a TI92+ in approx mode and For(k,1,1000):End runs
in 5 s on a TI86.
According to the TI92 F.A.Q this slowness is due to the fact that the FOR loop is very
flexible.
I have written tons of programs for the TI92 series including some very complex and i
have never found the utility of this For
loop flexibility except for one program and even in this case you could use a while
loop without significant loss of speed.
So what is the utility of such flexibility when you don't need it in 99 % of the case
and when in the 1 % of case you need it
you can use While loop instead ?
Another example of the TI-Basic inefficiencies:
The calc awfully handle local variables.
For example you can't use local variables of a program in a subprogram of this
program.
And what do you think of the loss of the use of symbolic local variables from the TI92
to the TI92+.
Other examples of the OS inefficiencies:
Do you know that whatever your free RAM is the calc uses only 64 KB for its
computations ?
Do you know that the speed of the calc decreases when the number of variables
increases to be almost inexistent when you reach the limit of variables that the calc
can handle(once the 64 KB allocated to the V.A.T is full) whatever the Free RAM is ?
Have you noticed how slow become the calc once you have several big results in the
history ?
Do you know that the TI92+ is slower than the TI92II(until 50 %) in approx mode when
they should have at least the same speed like in Exact mode ?
Don't you find irritating the slowness of the text editor ?
What about the cursor speed in the graph screen ?
etc...
Dean G. Lewis
> In fact the user data archive alone will be increased to 640 KB so the
total
> availlable memory will be inreased to 828 KB.However the user data archive
will be
> used for both storage and applications.
> If TI completely removed the geometry and made it availlable it as a Free
apps for the
> TI92+ owner is supposed that it would be possible to increase the user
data archive
> until over 896 KB.
When I read the TI web site, that is not how I interpreted the memory
distribution. Although I hate to be wrong about things, I hope your
interpretation is correct.
> Oh they are a lot of inefficiencies of the TI92 OS.The TI92 uses too slow
algorithms
> for numeric matrix computations.
> It is even slower than the TI86 for most numeric matrix computations.
> The funny thing is that the eigvc and eigvl functions of the TI92+ which
are 100 %
> numeric are much faster than those of the TI86 probably because TI used
the same
> algoritms on the 2 calc.
> So why TI has not implemented the TI86 algorithms for all the other 100 %
numeric
> matrix computations ?
I think TI made a balance between the speed of the sybolic matrix
computations and the numeric computations. Since the calculator is geared
towards students, they made the right decision. When professionals are
looking for speed, they will turn to C or perhaps Mathematica or Maple, so
having a calculator that can invert a 30 X 30 matrix in the blink of an eye
is not all that important. If the HP49G manages to perform lightning fast
numeric and sybolic matrix computations, then I might start to be a little
disappointed with the 92+.
> If TI is so interested by incredible accuracy on the TI92+ they could have
implemented
> arbitrary accuracy for numeric computations(i don't talk of computations
with
> integers) after all it is a Derive feature.
I would also like to see this feature, but I'm not sure I would categorize
not having this feature as an inefficiency.
> The For loop of the TI92+ has been designed in such a way that it is
incredibly slow.
> For k,1,1000:Endfor runs in 17 s on a TI92+ in approx mode and
For(k,1,1000):End runs
> in 5 s on a TI86.
> According to the TI92 F.A.Q this slowness is due to the fact that the FOR
loop is very
> flexible.
> I have written tons of programs for the TI92 series including some very
complex and i
> have never found the utility of this For
> loop flexibility except for one program and even in this case you could
use a while
> loop without significant loss of speed.
> So what is the utility of such flexibility when you don't need it in 99 %
of the case
> and when in the 1 % of case you need it
> you can use While loop instead ?
I really can't defend the speed of the for loop. At times I feel like I
could blink an eye faster than an iteration of the for loop. A person may be
able to questions TI's judgement in making the for loop flexible, but since
it was designed this way, you really can't categorize it as an inefficiency.
Now if the for loop ran just a slow on the TI86, I would have to categorize
that as an inefficiency since the slowness would not be caused by an
additional feature of the for loop.
> Another example of the TI-Basic inefficiencies:
> The calc awfully handle local variables.
> For example you can't use local variables of a program in a subprogram of
this
> program.
This I would have to categorize as the lack of a feature rather than an
inefficiency.
> And what do you think of the loss of the use of symbolic local variables
from the TI92
> to the TI92+.
I did find this very strange since nothing obvious was gained. It did kind
of give me the impression that perhaps something was accidently broken in
the code in the ROM that did not allow a program to utilize symbolic local
variables anymore.
> Other examples of the OS inefficiencies:
> Do you know that whatever your free RAM is the calc uses only 64 KB for
its
> computations ?
> Do you know that the speed of the calc decreases when the number of
variables
> increases to be almost inexistent when you reach the limit of variables
that the calc
> can handle(once the 64 KB allocated to the V.A.T is full) whatever the
Free RAM is ?
Now these are true inefficiencies. Fortunately they aren't the type of
thing that will affect 99.5% of the students using the calculators.
> Have you noticed how slow become the calc once you have several big
results in the
> history ?
This does affect me often and is quite annoying. I'm normally not a big
complainer, but this is something that really irritates me.
> Do you know that the TI92+ is slower than the TI92II(until 50 %) in approx
mode when
> they should have at least the same speed like in Exact mode ?
This had surprised me also when I first heard this. Fortunately this speed
reduction is in the user programs and functions rather than the built in
functions, although that is hardly an excuse.
> Don't you find irritating the slowness of the text editor ?
I don't use the text editor so I'll take your word for it. I guess I have
heard this complaint in the TI news group a few times.
> What about the cursor speed in the graph screen ?
> etc...
One way to speed this up is to press the second button and then the arrow
key. This will cause the cursor to skip several pixels (can't remember how
many) so it makes it much faster than just your normal scrolling in the
graph screen.
I suppose I'm probably down playing some of your inefficiencies that you
listed, but I guess that's probably because these are the type of things
that never affect me. Apparently they do affect you, though, so I can
certainly understand why you would be upset by them. If the HP49 is
everything that it is made out to be (which with HP's reputation it probably
is), then you will probably find what's lacking with the 92+ in the HP49G.
My only complaint about the 49G is that it does not have a qwerty keyboard,
although I am probably in the minority for wanting one.
Timité Hassan
Dean G. Lewis wrote:
> > In fact the user data archive alone will be increased to 640 KB so the
> total
> > availlable memory will be inreased to 828 KB.However the user data archive
> will be
> > used for both storage and applications.
> > If TI completely removed the geometry and made it availlable it as a Free
> apps for the
> > TI92+ owner is supposed that it would be possible to increase the user
> data archive
> > until over 896 KB.
>
> When I read the TI web site, that is not how I interpreted the memory
> distribution. Although I hate to be wrong about things, I hope your
> interpretation is correct.
I am sure that my interpretation is correct.There are over 160 KB of Flash not
used in the TI92+(neither by the OS nor by the User data archive).
If TI has fred up this up these 160 KB it would have increased the user Flash
from 384 KB to 544 KB of user Flash.
Now i suspect that TI has trim down the TI-OS to free up 96 KB of additionnal
Flash for a total of 640 KB.
In fact we can say that we are useing beta version of the TI92+ OS.
Shouldn't it be forbidden to sell not finished products ?
> I think TI made a balance between the speed of the sybolic matrix
> computations and the numeric computations. Since the calculator is geared
> towards students, they made the right decision. When professionals are
> looking for speed, they will turn to C or perhaps Mathematica or Maple, so
> having a calculator that can invert a 30 X 30 matrix in the blink of an eye
> is not all that important. If the HP49G manages to perform lightning fast
> numeric and sybolic matrix computations, then I might start to be a little
> disappointed with the 92+.
Well i don't want blink eye matrix computations on a calculator(is it possible
even with a m68k ?) but at least faster numeric computations than on a TI85.
TI has implemented 2 algorithms for the QR decomposition:one for exact matrix
and one for numeric matrix.
So i wonder why they didn't do the same for most useful matrix computations such
as matrix inversion,determinant,etc...?
I am sure that the HP49 will outperform the TI92+ for numeric matrix
computations as the HP48 already blow it away.
And it is highly probable that the HP49 will also outperform the TI for symbolic
matrix computations especially when i see the inefficiency of the determinant
algorithm of the TI92+ on symbolic matrix.
> > If TI is so interested by incredible accuracy on the TI92+ they could have
> implemented
> > arbitrary accuracy for numeric computations(i don't talk of computations
> with
> > integers) after all it is a Derive feature.
>
> I would also like to see this feature, but I'm not sure I would categorize
> not having this feature as an inefficiency.
I don't say that it is an inefficiency but it would have been great to have it.
The answer i have got from TI about this problem is quite stupid.For me it is a
design flaw because such flexible FOR loop are useful when there isn't While
loop.
But there is already a while loop on the TI92+.
However a simple solution would be to implement a second loop FOR let 's called
it FOR2 which will run like a normal FOR loop.
It should significantly speed up a lot of programs on the TI92+.
Let's give you an example:
On a TI92+ in approx mode:
0->s:For k,1,1000:s+1/k->s:Endfor runs in 31 s
For k,1,1000:Endfor runs in 17 s
This mean that the calculator spends more time to run the loop then to do the
calculations.
On a TI86:
0->s:For(k,1,1000):s+1/k->s:End runs in 35 s
For(k,1,1000):End runs in 5 s.
Do you see the difference ?
It is a crime to implement such a slow loop in a langage.
> > And what do you think of the loss of the use of symbolic local variables
> from the TI92
> > to the TI92+.
>
> I did find this very strange since nothing obvious was gained. It did kind
> of give me the impression that perhaps something was accidently broken in
> the code in the ROM that did not allow a program to utilize symbolic local
> variables anymore.
A little thing has been gained but which is not really needed.Now on the TI92+
you can evaluate f(x+1) if f is a function defined with the variable x like for
example f(x)=sin(x)+x.
You can't on the TI92,you get an ERROR:Circular definition.
I have been said that these problems are linked to some TI92 core flaws.
It seems that the TI92 doesn't see a difference between a variable and a
symbolic name in an expression.
> > Other examples of the OS inefficiencies:
> > Do you know that whatever your free RAM is the calc uses only 64 KB for
> its
> > computations ?
> > Do you know that the speed of the calc decreases when the number of
> variables
> > increases to be almost inexistent when you reach the limit of variables
> that the calc
> > can handle(once the 64 KB allocated to the V.A.T is full) whatever the
> Free RAM is ?
>
> Now these are true inefficiencies. Fortunately they aren't the type of
> thing that will affect 99.5% of the students using the calculators.
Yes,these type of inefficiencies won't affect the current consummer but they
show the differences between TI92 and HP OS core power.
> > Have you noticed how slow become the calc once you have several big
> results in the
> > history ?
>
> This does affect me often and is quite annoying. I'm normally not a big
> complainer, but this is something that really irritates me.
I have even get an ERROR memory while i was scrolling a big result in the
history and i wasn't low in memory.I suspect that this problem is due to the
memory flaws i have listed above and to the lack of optimisation of the display
routines.
> > Do you know that the TI92+ is slower than the TI92II(until 50 %) in approx
> mode when
> > they should have at least the same speed like in Exact mode ?
>
> This had surprised me also when I first heard this. Fortunately this speed
> reduction is in the user programs and functions rather than the built in
> functions, although that is hardly an excuse.
Not really almost all the built-in functions are affected.It is the worst with
the sum(sigma) function.
And the TI92+ computes the matrix for 3D plottings slowly than the TI92II but it
has the advantage to display much faster 3D
plottings.
> > Don't you find irritating the slowness of the text editor ?
>
> I don't use the text editor so I'll take your word for it. I guess I have
> heard this complaint in the TI news group a few times.
>
> > What about the cursor speed in the graph screen ?
> > etc...
>
> One way to speed this up is to press the second button and then the arrow
> key. This will cause the cursor to skip several pixels (can't remember how
> many) so it makes it much faster than just your normal scrolling in the
> graph screen.
I know this way.It enables to skip 10 pixels at the type but i was talking of
the normal scrolling.
> I suppose I'm probably down playing some of your inefficiencies that you
> listed, but I guess that's probably because these are the type of things
> that never affect me. Apparently they do affect you, though, so I can
> certainly understand why you would be upset by them. If the HP49 is
> everything that it is made out to be (which with HP's reputation it probably
> is), then you will probably find what's lacking with the 92+ in the HP49G.
> My only complaint about the 49G is that it does not have a qwerty keyboard,
> although I am probably in the minority for wanting one.
Yeap.I also prefer QWERTY keyboard but i can use HP keyboard quite quickly,not
like the TI89 keyboard that i really hate !!!
Paul FLOYD
[TI92 memory discussion]
> Shouldn't it be forbidden to sell not finished products ?
So what do you call Windows 95 and Netscape that you used to post your
message, apparently?
(Within limits) I'd rather have something that works now than wait till
I retire to get something perfect.
A+
Paul
--
Paul Floyd Focal Ingenierie Sud
Mail: paul underscore floyd at focal dot fr
Sig tagline under contemplation.
jef...@webtv.net
Many many thanks for all the input.
I will definately be purchasing a
49, and will check out the 92+'s
new OS when it comes out. If
nothing else, just to check out
programming in C on a calc!
A couple other questions have come
up since my original post though.
One - will the 49 actually do
symbolic matrices nice and neatly
with built-in functions? (I failed
to see symbolic matrix functions on
the HP page.)
Two - I've had a problem w/ my 48
overheating in nominal heat.
Sometimes at even a mere 80
degrees, it just blanks out, and
only sometimes will it recover
memory. W/out a true computer to
link to, this makes backing up all
essentials on paper a real pain!
With the new case on the 49, it
looks sturdier, but will it be more
heat resistant?
jc
rse...@primenet.com
summer. [I don't recommend this, it has always been inadvertent.] I've
never had a problem with it. So maybe there is variation in quality.
[I'll have to test it in Flagstaff this Winter.]
In article <7m9bdi$mit$1...@nnrp1.deja.com>,