TI89 Titanium -- opinions?
Gene Wright
Anyone have one of these yet?
In what areas does it appear better than the 49g+? (bundled apps?)
In what areas is the 49g+ better? (SD card, etc)
Gene
Veli-Pekka Nousiainen
news:199dd7ad.04062...@posting.google.com...
[VPN]
Aleph Mobius
apps it comes with are freely available to any TI89 owner, Titanium or
not. It now includes a USB link port so transfers are faster and it
has 2.5mb of FLASH instead of the TI-89's 780k or so. The case colors
and shape are also a little different. The processer and software are
unchanged.
Bhuvanesh
> "Gene Wright" <genewr...@hotmail.com> wrote:
> > Hi all.
> >
> > Anyone have one of these yet?
> >
> > In what areas does it appear better than the 49g+? (bundled apps?)
> >
> > In what areas is the 49g+ better? (SD card, etc)
> >
> The Titanic is f*ing slow
> [VPN]
For you and me, yes. It's generally fast enough for the typical
student. Remember that TI is not targeting professionals, much as we
would like them to.
--
Bhuvanesh
Steen Schmidt
Have you actually used a TI89 Titanium? Mind you the ordinary TI89 HW2 is
still faster than even the HP49G+ for a number of things. The Titanic is
probably not slower than the HW2 TI89, although I haven't seen AMS v3.00
personally.
Regards
Steen
Terry Bowman
flash.
I still wouldn't buy it though.
Terry
"Aleph Mobius" <ness...@aol.com> wrote in message
news:91017fd2.04062...@posting.google.com...
Timit? Hassan
The TI89 Titanic is not faster than the TI89 HW2 and is certainly
slower than it.
Among all newly released TI68k models only the TI92II and the HW2 were
faster than their predecessors.
I am not even sure that the HW2 are faster than the TI92II(TI92+E
module) as this one was up to 50 % faster than the TI92+ HW1.
And even if the TI89 HW2 is faster than the HP49G+ for some things it
is slower for many more things.
Plotting,most numeric computations,GUI and user language are the first
things which come to my mind.
So i agree with VPN,the Titanic and for instances all TI68k are slow.
"Steen Schmidt" <ssch...@nospam.dk> wrote in message news:<gWnDc.2576$%h7....@news.get2net.dk>...
Timit? Hassan
"Veli-Pekka Nousiainen" <DROP...@LETTER.welho.com> wrote in message news:<cbjrjs$9gm$1...@nyytiset.pp.htv.fi>...
Steen Schmidt
That's the problem with these calculators - you need to have used them
before you can speak about their performance. That's true about everything
in life, isn't it? I just advocate that people shouldn't jump to
conclusions - stick to facts and experience.
> The TI89 Titanic is not faster than the TI89 HW2 and is certainly
> slower than it.
Since it's based on the same hardware, I must assume you know from personal
experience that AMS 3.00 is slower than AMS 2.09? And that this performance
decrease will not propagate to the TI89 HW2 when AMS 3.00 is released for
that platform?
> I am not even sure that the HW2 are faster than the TI92II(TI92+E
> module) as this one was up to 50 % faster than the TI92+ HW1.
My personal experience tells me that it is.
> And even if the TI89 HW2 is faster than the HP49G+ for some things it
> is slower for many more things.
Have you conducted any form of test to show this? How do you weigh
importance of functionality?
> Plotting,most numeric computations,GUI and user language are the first
> things which come to my mind.
Lets' see...I'll just spend a moment digging through my last 300 hours of
pitting the TI89 HW2 against the HP49G+;
I have compared almost every aspect of these two machines, ranging over
object diversity, battery life, OS stability, algorithms, programming,
plotting, I/O, 3rd party apps, customization, so on and so forth. And of
course mathematical abilities, inspecting practicality/integration,
diversity, usefulness....and of course performance.
I gather you'd like to stick with the latter, as that seems to be the only
aspect of this you might put any weight to. I'll entertain with the very
short story only (all claims are isolated, i.e. measurement overhead is
deducted from the test results. All claims are general, in each case an
opposite scenario can easily be constructed):
- The HP is roughly twice as fast for real numerical computations.
- The HP is 2-6 times as fast for complex numerical computations.
- The HP has better numeric to exact conversion.
- The HP is the most versatile regarding indefinite integration, and often
the fastest too. Both have quirks in this area.
- The HP is *much* faster at series expansion, and it is more versatile too
(the TI really only sporting Taylor expansion).
- The HP sports a host of modulo arithmetic which the TI has not. The same
is true for a bunch of special functions and integer and polynomial
arithmetic. The latter is negated when taking Bhuv's MathTools for the TI
into account. A truly magnificent set of programs.
- The HP is more versatile regarding partial fraction expansion, mainly
stemming from the choice of complex or real mode.
- The HP is *much* more versatile when it comes to expression manipulation,
in part helped by the EQW. A couple of really high quality EQWs exist for
the TI as well now.
- The HP is *much* more versatile when dealing with trig functions, while
the TI is usually faster. The versatility of the HP must prevail here
though.
- The HP is 2-5 times as fast for symbolic matrix calculations, and 2-10
times as fast for numeric array calculations.
- The HP is much better when dealing with units. Beyond description better.
- The TI is 2-4 times as fast for integer arithmetic.
- The TI is faster at differentiation, except when auto simplification plays
a major role (typically involving trig functions).
- The TI is *much* faster at numeric integration.
- The TI is *much* faster at limits, except when we're dealing with simple
substitution.
- The TI is faster when doing symbolic and numeric sums. The Ti can also do
symbolic closed form products, while the HP cannot.
- The TI is much faster and more versatile at solving symbolic and numeric
equations, except the special case of univariate polynomials with strictly
numeric coefficients.
- The TI is generally better at factoring polynomials.
- The TI is usually the master of simplification, substitution and
evaluation, except again when its auto simplification routines trip over
each other. The HP is just generally half-slow, but rarely ice-age slow (The
only case that springs to mind is when dealing with RE, IM and ABS - forget
about it!).
- They both suck at integer factorization and primality tests (although the
TI seems to implement the most stable primality test by a hair).
- They are about equal dealing with vectors.
The TI has one fast parsing engine. It's truly amazing how quickly it can
apply its heuristics to expressions - I'm currently trying to mimick some of
these heuristics on the HP49G+, and the parsing is really a bottleneck (or
I'm not clever enough).
One aspect of the TI, I find very comforting, is that it often didn't fail
some complex computation, but instead kept going for hours and hours while
the HP would give up after a very short time if it couldn't find an answer.
Time limited algorithms are not good on a calculator in my opinion - they
are necessary on a mainframe when the algorithm has determined that a
solution within all probability will take millenia to compute, but on a
calculator I fancy the user to press CANCEL rather than the HP not factoring
something (not even a warning is issued, that the algorithm aborted because
of time constraints).
One thing I like about the HP is the lack of auto simplification - or, more
correctly, I'd very much like it to be there, but the user should be able to
switch it on and off. This should be so on the TI as well.
The keyboard layout on the TI89 is bad, but the keys on the HP49G+ sucks so
much I don't have words for it. The layout is perfect, but I miss keystrokes
constantly - if I enter one simple expression, say, with 25 keystrokes or
so, I see myself retyping maybe 4 or 5 entries (sometimes I have to press a
key 3 or 4 times before it registers). When the HP is turned ON, my first
keypress is often not registered. Delimiters is a problem area for me too -
I have a hard time getting my current HP to register delimiter entry like '
', " ", << >>, { }, [ ] and so on. Also the arrow keys, negate,
inverse.....well, I'll be d***** - It's *all the keys*!
And without getting into too much detail, TI-Basic isn't really slower,
generally speaking, than UserRPL. TI-Basic sucks at looping in most
instances, and UserRPL really shines when the stack is utilized (which not
that many people can do effectively really). TI-Basic is very limited, while
programming the HP has almost no limits (that is when judging on-calc
programming). The SDK & C development features are much better on the TI. By
a large margin.
> So i agree with VPN,the Titanic and for instances all TI68k are slow.
You agree with Veli-Pekka in answering something you really do not know
anything about - or that's how I see it.
It's not just HP as a company that's deteriorating - the quality of this
news group is going down the same hill (started around '99-'00 or so).
Regards
Steen
Steen Schmidt
> - The TI is 2-4 times as fast for integer arithmetic.
Except in the special discipline of factorial calculation.
> - The TI is generally better at factoring polynomials.
That is when dealing with symbolic factors - numeric factorization belongs
to the HP.
The TI spends a long time displaying results in the history. That's very
annoying. About plotting; for simple functions, on a pixel-per-pixel basis,
the two are very equal in plotting speed. That sums out to an advantage
towards the HP, as it has fewer pixels to plot. For complex functions the HP
is 2-3 time faster than the TI.
Regards
Steen
Veli-Pekka Nousiainen
X
> Lets' see...I'll just spend a moment digging through my last 300 hours of
> pitting the TI89 HW2 against the HP49G+;
>> So i agree with VPN,the Titanic and for instances all TI68k are slow.
>
> You agree with Veli-Pekka in answering something you really do not know
> anything about - or that's how I see it.
>
> It's not just HP as a company that's deteriorating - the quality of this
> news group is going down the same hill (started around '99-'00 or so).
>
> Regards
> Steen
I must add here that I have actually owned a TI 89
After and before that I have used Virtual TI
and tried many times the real calculators
1) Numerical Matrix Inversion is terrible in the TI
2) Loops are slooooooooooooooooooooooow
3) The memory is not fully utilized, there are limits
A) Symbolics are faster in the TI
B) Display is sharper in the TI
My greatest dissapointment is that
TI did not use faster CPU for the Titanic.
They could have had 33,3 MHz and thus beat up the 49g+
Now there's nothing really new... )-:
I want the Qonos!!!
I can have a CAS that simulates TI, HP, Maple, MuPad, ...
It will also be my first handheld Linux system!
(I could install Linux to my Jornada 720)
(VPN)
Steen Schmidt
Yes, but have you used it to any extent to compare it to the HP49G+? Have
you used your HP49G+ to an extent so you can tell where there have been and,
more importantly, where there haven't been improvements over the HP49G? Are
you ignorant to the fact that the HP49G+ is slower than the HP48G for some
tasks? Does the HP49G+ suck?
Have you tried a TI89 Titanium, which were the model you stated facts about?
> 1) Numerical Matrix Inversion is terrible in the TI
There are programs to overcome this. One slow implementation shouldn't
result in a "f***** slow" statement. It's simply not true, unless you also
mean the HP49G+, HP49G & all the HP48s are "f***** slow"?
> 2) Loops are slooooooooooooooooooooooow
Don't use 'em or code in C instead. Not all types of loops are slow on the
TI either.
> 3) The memory is not fully utilized, there are limits
So are there indeed in the HP memory architecture. What about the 128 kB
maximum block size? What about no directories (officially) on the SD card?
> A) Symbolics are faster in the TI
Not all symbolics, but some.
> B) Display is sharper in the TI
Not any more - the HP49G+ has a sharper display than my TI89. Do you own a
HP49G+?
> My greatest dissapointment is that
> TI did not use faster CPU for the Titanic.
> They could have had 33,3 MHz and thus beat up the 49g+
> Now there's nothing really new... )-:
Exactly like the introduction of the HP49G. Where you as harsh towards that
as you are now towards the Titanic?
> I want the Qonos!!!
> I can have a CAS that simulates TI, HP, Maple, MuPad, ...
A CAS that *simulates* Maple? Come on - Maple costs a thousand bucks. I
wouldn't hold my breath for either TI nor HP simulation - TI guards their
ROMs ferosiously, and unless Hydrix have some sort of business agreement
with TI (could be, but I doubt it), they simply can't simulate the TI89.
Emulate maybe, but that'd be a huge task. Simulating or emulating the HP49G+
on the Qonos will probably make the simulation/emulation run slower than the
real thing. It's already running on an ARM core, and the XScale is not a
miracle speed demon for your information - it's used because of its price,
versatility and I/O (among other things, but not primarily for performance).
> It will also be my first handheld Linux system!
> (I could install Linux to my Jornada 720)
I like Windows Mobile 2003 better than Linux anyway :-)
Regards
Steen
Veli-Pekka Nousiainen
X
>> 2) Loops are slooooooooooooooooooooooow
>
> Don't use 'em or code in C instead. Not all types of loops are slow on the
> TI either.
That's new to me...I don't have a laptop to drag around
Although there are some tempting small models...
>
>> 3) The memory is not fully utilized, there are limits
>
> So are there indeed in the HP memory architecture. What about the 128 kB
> maximum block size? What about no directories (officially) on the SD card?
- User doen't see any block size when using the Flash (port 2)
- The Main RAM is ~240KB and a Code object can be creater than 64KB
- If you STOre something in a non-existent DOSDIR in an SD
it will be created
- where is the SD in any TI model???
>> B) Display is sharper in the TI
>
> Not any more - the HP49G+ has a sharper display than my TI89. Do you own a
> HP49G+?
Yes, I do: 131*80 pixels, now - since you own a TI-89
and you know it so well - how many pixels it has?
[VPN] - I Win - You loose, Man of Steen is rusty?
Aleph Mobius
> Apparently it does come with EEPro installed that is probably why the bigger
> flash.
The file is 327k on disk but I am not sure how much space it takes up
on the calc. This is also freely available to all TI89 owners. Unlike
the Voyage 200 for example, TI89 owners can get all the software
preloaded on the Titanium for free.
MM
"Steen Schmidt" <ssch...@nospam.dk> schrieb im Newsbeitrag
news:SRADc.1910$NH3...@news.get2net.dk...
<snip>
>Qonos will probably make the simulation/emulation run slower than the
>real thing. It's already running on an ARM core, and the XScale is not a
>miracle speed demon for your information - it's used because of its price,
>versatility and I/O (among other things, but not primarily for
performance).
I don't know, whether I can follow your objection here.
Are you saying, they try to emulate an HP49G+ on their Qonos device?
(i.e. implementing an ARM *emulator* on an ARM hardware core...)
That's kinda sick? Isn't it?
I think they emulate the Saturn on the XScale, as they did on the 49G+.
So since the XScale clock is @300-400MHz vs. 75MHz on the HP49G+, why should
the emulation on that XScale be slower then?
I also remember Eric Rechlin reporting, that the prototype was pretty fast
compared to the HP49G+ hardware.
--
MM
> Regards
> Steen
>
>
Steen Schmidt
You know you can't - there's really no need to be a wiseguy. I appreciate
very much that I can program the HPs in SysRPL directly on the calculator,
but for application development for the TI89 you could just as well use a
PC. I miss an opportunity to program the HP in C, whether it being on the
calc or on a PC.
To get back on track, here are some *facts* about branching performance on
the HP49G+ and TI89 HW2 (units are s^-1, higher numbers are better):
For-loop;
HP48GX: 366
HP49G+: 437-2377
TI89: 93-156
While-loop;
HP48GX: 247
HP49G+: 1395
TI89: 1357-1583
If-branching;
HP48GX: 236-296
HP49G+: 1040-1371
TI89: 1000-1267
I have for the comparisons sake included the HP48GX, since the TI89 was
already accused of glacial-slow looping performance at that time. Sure, the
For-loop implementation on the TI89 is poor, but other loop and branching
performance is very much on par with the HP49G+. Hence, avoid For-loops, and
you're pretty much in the clear.
> - User doen't see any block size when using the Flash (port 2)
Which is exactly the same as on the TI89 - the block size is transparant on
both calcs, except when you cross it (64 kB chunks on the TI, 128 kB chunks
on the HP).
> - The Main RAM is ~240KB and a Code object can be creater than 64KB
Yes, my TI89 has around 190 k RAM free, while my HP49G+ has around 215 k.
The asm limit on the TIs is actually 24 k. You don't consider object sizes
here, although that is at least as important as raw memory at disposal.
Examples:
Smaller on TI89:
Integers
Symbolic matrices
Expressions
Programs
Smaller on HP49G+:
Real numbers
Numeric arrays
For broad use the RAM on the TI89 and on the HP49G+ are equivalent in size
in my experience.
> - If you STOre something in a non-existent DOSDIR in an SD
> it will be created
But how do you delete the directory again?
> - where is the SD in any TI model???
But we were talking about performance here, not features.
> Yes, I do: 131*80 pixels, now - since you own a TI-89
> and you know it so well - how many pixels it has?
160*100.
> [VPN] - I Win - You loose, Man of Steen is rusty?
This is not a contest, Veli-Pekka. These news groups used to be about
enlightment and exciting new knowledge, not about mudthrowing, namecalling
and arguments for the sole sake of "winning".
The fact is that you, dare I say now, have again spoken about something you
don't know didly about. The TI89 Titanium - you've never touched it, and, I
suspect, never even seen one. You have *never* used AMS 3.00, so how can you
categorically state that it is slow? You don't know? You can't necessarily
compare it to an ordinary HW2 TI89 - which I've incidently also pointed out
can easily compete with the HP49G+ in many areas still. If you go back to my
original post on this subject, I merely asked you if you had seen one.
Apparantly not.
Regards
Steen
Steen Schmidt
> Are you saying, they try to emulate an HP49G+ on their Qonos device?
> (i.e. implementing an ARM *emulator* on an ARM hardware core...)
I don't know what they will/are doing.
> That's kinda sick? Isn't it?
It depends on what software you have access to, and how you are allowed to
use it.
> I think they emulate the Saturn on the XScale, as they did on the 49G+.
Maybe.
> So since the XScale clock is @300-400MHz vs. 75MHz on the HP49G+, why
should
> the emulation on that XScale be slower then?
It all depends on how the emulation is done and how the emulation layer is
handled in the OS. Sure, if the emulator has unhindered access to all
hw-resources, if all interrupts are stopped (i.e. the Qonos OS and
proprietary I/O is halted) and so on, the HP49G+ emulator might be faster
than the original. I'm just of that conviction that it'd be a smaller task
to emulate the ARM920T instead of the Saturn + writing new ARM wrappers.
Also, will HP let Hydrix use the original HP49G+ OS?
It's all speculation - we'll see when the Qonos is out.
I think we should concentrate on adjusting the clock in the HP49G+ instead.
The Samsung S3C2410X used is nominally spec'ed at 203 MHz (it could be a
special speed binning used in the HP though), and I've heard of people
making it run at 260-270 MHz without issues. It's a question if the thermal
envelope of the calc is compatible with such speeds.
It does seem that the CPU is currently throttled, hence not running at 75
MHz constantly.
> I also remember Eric Rechlin reporting, that the prototype was pretty fast
> compared to the HP49G+ hardware.
Yes, I read that. This could also be caused by external hardware such as
different keyboard handler etc. It'll be exciting to find out for sure
though.
Regards
Steen
Tom Lake
>
> You know you can't - there's really no need to be a wiseguy.
Yes, you can.
Scroll down quite far to find it. Are all these French people brilliant?
It seem like all the great software comes from them! 8-)
Tom Lake
Jean-Yves Avenard
Steen Schmidt wrote:
> Emulate maybe, but that'd be a huge task. Simulating or emulating the HP49G+
I can answer some of that.
We have a TI89 emulator already up and running, it's currently slower
than a real TI89 but I'm 100% convinced we will be faster than the TI89
in the final version.
> on the Qonos will probably make the simulation/emulation run slower than the
> real thing. It's already running on an ARM core, and the XScale is not a
> miracle speed demon for your information - it's used because of its price,
> versatility and I/O (among other things, but not primarily for performance).
Well, the main reason for choosing the xscale was its power consumption.
It's very low.
regarding HP49G speed , it's about 25 times faster than the real 49G
right now under linux ; should be slightly faster under ecos. it's
emulating the 49g (not the 49g+)
Jean-Yves
Samuel Nyall Stearley
need a hw3 patch if you want to get TSRs to run.
By the way what where you're tests for if-then branching? branching
is funny on the 89 as some kinds of loops use offsets to jump around,
but others don't. An example in tibasic sudeo code: It is slightly
faster with continues than without.
for ...
if cond_1 then
continue
else if cond_2 then
continue
else if cond_3 then
continue
endif
endfor
Bhuvanesh
> He doesn't need to used a TI89 Titanium to know that it is slow.
> The TI89 Titanic is not faster than the TI89 HW2 and is certainly
> slower than it.
I haven't noticed any significant slowdown. Can you be more specific?
--
Bhuvanesh
Steen Schmidt
> is funny on the 89 as some kinds of loops use offsets to jump around,
> but others don't. An example in tibasic sudeo code: It is slightly
> faster with continues than without.
I used best and worst case of:
for ...
if true then
endif
endfor
for ...
if false then
endif
endfor
for ...
if true then
else
endif
endfor
for ...
if false then
else
endif
endfor
while deducting the running time of
for ...
endfor
Regards
Steen
Steen Schmidt
> right now under linux ; should be slightly faster under ecos. it's
> emulating the 49g (not the 49g+)
That makes sense.
Regards
Steen
Bhuvanesh
> The asm limit on the TIs is actually 24 k.
The limit has been effectively removed on the TI-89T (the maximum
program size is now 64k), and is expected to be removed for the rest
of the 68k family when AMS 3.0x is released for those calculators.
--
Bhuvanesh
Bhuvanesh
> The file is 327k on disk but I am not sure how much space it takes up
> on the calc.
You can check this in the Var-Link menu, under the "FlashApp" tab.
> This is also freely available to all TI89 owners. Unlike
> the Voyage 200 for example, TI89 owners can get all the software
> preloaded on the Titanium for free.
Not all, only free Flash apps. V200 owners can also get these Flash apps.
--
Bhuvanesh
Bhuvanesh
> The file is 327k on disk but I am not sure how much space it takes up
> on the calc.
You can check this in the Var-Link menu, under the "FlashApp" tab.
> This is also freely available to all TI89 owners. Unlike
> the Voyage 200 for example, TI89 owners can get all the software
> preloaded on the Titanium for free.
Not all, only free Flash apps. V200 owners can also get these Flash apps.
--
Bhuvanesh
Bhuvanesh
Can you clarify what you mean by 'versatile'?
<<the TI really only sporting Taylor expansion>>
You know about my Laurent series program, right?
<<A truly magnificent set of programs.>>
Thanks; that's a great compliment :-)
<<The HP is 2-5 times as fast for symbolic matrix calculations, and
2-10
times as fast for numeric array calculations.>>
...and doesn't have that 32k limit on expression size, right?
<<The TI is faster when doing symbolic and numeric sums.>>
A couple of years back I would have mentioned Gosper and digamma type
summations as an HP advantage, but now we have those for the 68k too.
Regards,
Bhuvanesh.
Veli-Pekka Nousiainen
news:d1IDc.143398$j24....@twister.nyroc.rr.com...
So the TI is better after all! (at least for Man of Steen)
(-;
[VPN]
Veli-Pekka Nousiainen
news:662e00ed.04062...@posting.google.com...
TI is SW getting better - there is still development!
[VPN]
Veli-Pekka Nousiainen
> <<A truly magnificent set of programs.>>
> ...and doesn't have that 32k limit on expression size, right?
> A couple of years back I would have mentioned Gosper and digamma type
> summations as an HP advantage, but now we have those for the 68k too.
Bhuv, with you extension the TI can even beat the 49 in those
specific areas - BUT what expression size limit?
Does anyone need that long expression?
???
[VPN]
Veli-Pekka Nousiainen
>> [VPN] - I Win - You loose, Man of Steen is rusty?
(-;
^
There was a smiley - didn'y you C it?
> This is not a contest, Veli-Pekka. These news groups used to be about
> enlightment and exciting new knowledge, not about mudthrowing, namecalling
> and arguments for the sole sake of "winning".
>
> The fact is that you, dare I say now, have
> again
BULLS**T
> spoken about something you
> don't know didly about. The TI89 Titanium - you've never touched it, and,
> I
> suspect, never even seen one. You have *never* used AMS 3.00, so how can
> you
> categorically state that it is slow?
> You don't know?
I also read the TI newsgroup (or web-news forums)
I suggest yoiu read them too
You just too cynical Steen
You used to be more fun before HP killed Xpander etc...
> You can't necessarily
> compare it to an ordinary HW2 TI89 - which I've incidently also pointed
> out
> can easily compete with the HP49G+ in many areas still. If you go back to
> my
> original post on this subject, I merely asked you if you had seen one.
> Apparantly not.
>
> Regards
> Steen
No regards until Steen gets happier....
>VPN<
Aleph Mobius
> >
>
Have any of you tryed this program? I had it on my calc a while ago
but took it off. Any time it encountered an error and exited, a large
chunk of memory (50 or 60k IIRC) was not freed. After this happens 2
or 3 times there isnt enough RAM left do do anything. Have any of you
encountered this and/or sent a bug report to the author?
Aleph Mobius
> ness...@aol.com (Aleph Mobius) wrote:
> > The file is 327k on disk but I am not sure how much space it takes up
> > on the calc.
>
> You can check this in the Var-Link menu, under the "FlashApp" tab.
Of course, but VTI won't load Flash Apps.
> > This is also freely available to all TI89 owners. Unlike
> > the Voyage 200 for example, TI89 owners can get all the software
> > preloaded on the Titanium for free.
>
> Not all, only free Flash apps. V200 owners can also get these Flash apps.
I see I was mistaken on that. The CellSheet app is preloaded on the
Titanium but is $15 for other TI89 users. The other 15 preloaded apps
on the Titanium are free to all TI89 owners.
Bhuvanesh
> X
> > ...and doesn't have that 32k limit on expression size, right?
> X
>
> Does anyone need that long expression?
It depends on whether you want to push the limits of what the device
can do :-). The size limit would affect the maximum matrix size, for
example. For approximate real numeric matrices, I think the maximum
size would be 126*126; for other element data types, the size would be
different.
Cheers,
Bhuvanesh.
Jesper Rasmussen
> of the 68k family when AMS 3.0x is released for those calculators.
Do you know when this will happen?
--
/Jesper
Mail sent to "From address" will not be read!
Use "news(at)jronline(dot)dk"
Veli-Pekka Nousiainen
Veli-Pekka Nousiainen
Thanks, Bhuv!
What about the speed of the new TI-Ti?
I think it's the greatest dissappointment among TiBORG?
read my lips: no new CPU (speed)...)-:
Cheers,
Vepanesh.
Steen Schmidt
> (-;
I like the HP better really.
Regards
Steen
Steen Schmidt
>
> Can you clarify what you mean by 'versatile'?
I only meant that the HP seems to be more consistent performance wise,
whereas the TI89 performance can deviate tremendeously from "small" changes
to input. It's mostly because of the high heuristic bias of the TI - even
though I tend to see this as an advantage.
Examples could be this set of integrands:
'1/(x^4-1)'
TI89: 0.278 s.
HP49G+: 1.100 s.
'1/(x^4+1)'
TI89: 3.010 s.
HP49G+: 2.348 s.
'1/(x^4-1)^4'
TI89: 26.127 s.
HP49G+: 4.154 s.
'1/(x^4+1)^4'
TI89: 114.205 s.
HP49G+: 3.414 s.
The Hp is the more consistent performance wise, while the TI has to grab the
bag o' tricks to integrate the latter two expressions. The difference isn't
huge in my opinion though, and the TI certainly have a huge advantage when
integrating polynomials for instance.
> You know about my Laurent series program, right?
Nope, not until now that is :-) I'll try it out.
> ...and doesn't have that 32k limit on expression size, right?
Right, although I rarely use expressins that large.
> A couple of years back I would have mentioned Gosper and digamma type
> summations as an HP advantage, but now we have those for the 68k too.
Yes, the Gosper implementation is actually pretty admirable. Your MathTools
has inspired me to start coding a bunch of special functions for the HP49G+.
I don't know if they'll ever be released - my HP calculator projects tend to
run out in the sand lately (lack of motivation I gather). I have a number of
half-finished projects lying around, including a very fast plotter (x10-20
over built-in), heuristic (>60k table already) symbolic integration &
differentiation (including special functions), arbitrary precision library
(the lib from Gjermund Skailand & Thomas Rast is much more complete),
numeric solvers aso.
I have actually started from scratch 3 or 4 times on my numeric solver -
each time I come to admire existing numeric solvers more and more :-)
Well, enough bable from me.
Regards
Steen
Ramiro Tasquer
> regarding HP49G speed , it's about 25 times faster than the real 49G
> right now under linux ; should be slightly faster under ecos. it's
> emulating the 49g (not the 49g+)
Are you going to make it available for download? (an x86 version)
or, the sources, so we can port it.
Ramiro "Tired of running emu48 over rdesktop" Tasquer
Jean-Yves Avenard
> Are you going to make it available for download? (an x86 version)
> or, the sources, so we can port it.
Unlikely.
Jean-Yves
Timit? Hassan
It is your superiority complex and the fact that you think that you
are smarter than everyone.
I thought that your failure with your advanced handheld project would
have made you more humble or less arrogant but obviously i was wrong.
If you think that both HP and the comp.sys.hp48 are losing their
quality,what are you doing here ?
And contrary to what you said i do know exactly what i am talking.
I used TI68k since the release of the beta version of the TI92(in
October 1995).
I have got or tested all TI68k version besides the Voyage 200.
I have talked with Ti R&D engineer and been deeply disapointed by
their serious faith that the TI92 was far superior to the HP48.
So unless i manage to always get TI68k with problems,i do know cleary
what i am talking.
Your statements almost prove my point that the HP49G+ is generally
faster than the TI68K.
Many of your statements are commong knowledge to me.
I find some other almost impossible to believe as in my experience it
is the opposite.
For example when you say that the TI is faster than the HP49G+ for
numeric sums.
Are you talking of the built-in command or of all ways to do a sum ?
Also how the hell have you manage to get your TI being 2-4 times
faster than your HP49G+ when my HP49G was faster than my TI92+ for
that ?
"Steen Schmidt" <ssch...@nospam.dk> wrote in message news:<PAyDc.135$KA7...@news.get2net.dk>...
> > He doesn't need to used a TI89 Titanium to know that it is slow.
>
> That's the problem with these calculators - you need to have used them
> before you can speak about their performance. That's true about everything
> in life, isn't it? I just advocate that people shouldn't jump to
> conclusions - stick to facts and experience.
>
> > The TI89 Titanic is not faster than the TI89 HW2 and is certainly
> > slower than it.
>
> Since it's based on the same hardware, I must assume you know from personal
> experience that AMS 3.00 is slower than AMS 2.09? And that this performance
> decrease will not propagate to the TI89 HW2 when AMS 3.00 is released for
> that platform?
>
> > I am not even sure that the HW2 are faster than the TI92II(TI92+E
> > module) as this one was up to 50 % faster than the TI92+ HW1.
>
> My personal experience tells me that it is.
>
> > And even if the TI89 HW2 is faster than the HP49G+ for some things it
> > is slower for many more things.
>
> Have you conducted any form of test to show this? How do you weigh
> importance of functionality?
>
> > Plotting,most numeric computations,GUI and user language are the first
> > things which come to my mind.
>
> Lets' see...I'll just spend a moment digging through my last 300 hours of
> pitting the TI89 HW2 against the HP49G+;
>
> I have compared almost every aspect of these two machines, ranging over
> object diversity, battery life, OS stability, algorithms, programming,
> plotting, I/O, 3rd party apps, customization, so on and so forth. And of
> course mathematical abilities, inspecting practicality/integration,
> diversity, usefulness....and of course performance.
>
> I gather you'd like to stick with the latter, as that seems to be the only
> aspect of this you might put any weight to. I'll entertain with the very
> short story only (all claims are isolated, i.e. measurement overhead is
> deducted from the test results. All claims are general, in each case an
> opposite scenario can easily be constructed):
>
> - The HP is roughly twice as fast for real numerical computations.
> - The HP is 2-6 times as fast for complex numerical computations.
> - The HP has better numeric to exact conversion.
> - The HP is the most versatile regarding indefinite integration, and often
> the fastest too. Both have quirks in this area.
> - The HP is *much* faster at series expansion, and it is more versatile too
> (the TI really only sporting Taylor expansion).
> - The HP sports a host of modulo arithmetic which the TI has not. The same
> is true for a bunch of special functions and integer and polynomial
> arithmetic. The latter is negated when taking Bhuv's MathTools for the TI
> into account. A truly magnificent set of programs.
> - The HP is more versatile regarding partial fraction expansion, mainly
> stemming from the choice of complex or real mode.
> - The HP is *much* more versatile when it comes to expression manipulation,
> in part helped by the EQW. A couple of really high quality EQWs exist for
> the TI as well now.
> - The HP is *much* more versatile when dealing with trig functions, while
> the TI is usually faster. The versatility of the HP must prevail here
> though.
> - The HP is 2-5 times as fast for symbolic matrix calculations, and 2-10
> times as fast for numeric array calculations.
> - The HP is much better when dealing with units. Beyond description better.
>
> - The TI is 2-4 times as fast for integer arithmetic.
> - The TI is faster at differentiation, except when auto simplification plays
> a major role (typically involving trig functions).
> - The TI is *much* faster at numeric integration.
> - The TI is *much* faster at limits, except when we're dealing with simple
> substitution.
> - The TI is faster when doing symbolic and numeric sums. The Ti can also do
> symbolic closed form products, while the HP cannot.
> - The TI is much faster and more versatile at solving symbolic and numeric
> equations, except the special case of univariate polynomials with strictly
> numeric coefficients.
> - The TI is generally better at factoring polynomials.
> - The TI is usually the master of simplification, substitution and
> evaluation, except again when its auto simplification routines trip over
> each other. The HP is just generally half-slow, but rarely ice-age slow (The
> only case that springs to mind is when dealing with RE, IM and ABS - forget
> about it!).
>
> - They both suck at integer factorization and primality tests (although the
> TI seems to implement the most stable primality test by a hair).
> - They are about equal dealing with vectors.
>
> The TI has one fast parsing engine. It's truly amazing how quickly it can
> apply its heuristics to expressions - I'm currently trying to mimick some of
> these heuristics on the HP49G+, and the parsing is really a bottleneck (or
> I'm not clever enough).
>
> One aspect of the TI, I find very comforting, is that it often didn't fail
> some complex computation, but instead kept going for hours and hours while
> the HP would give up after a very short time if it couldn't find an answer.
> Time limited algorithms are not good on a calculator in my opinion - they
> are necessary on a mainframe when the algorithm has determined that a
> solution within all probability will take millenia to compute, but on a
> calculator I fancy the user to press CANCEL rather than the HP not factoring
> something (not even a warning is issued, that the algorithm aborted because
> of time constraints).
>
> One thing I like about the HP is the lack of auto simplification - or, more
> correctly, I'd very much like it to be there, but the user should be able to
> switch it on and off. This should be so on the TI as well.
>
> The keyboard layout on the TI89 is bad, but the keys on the HP49G+ sucks so
> much I don't have words for it. The layout is perfect, but I miss keystrokes
> constantly - if I enter one simple expression, say, with 25 keystrokes or
> so, I see myself retyping maybe 4 or 5 entries (sometimes I have to press a
> key 3 or 4 times before it registers). When the HP is turned ON, my first
> keypress is often not registered. Delimiters is a problem area for me too -
> I have a hard time getting my current HP to register delimiter entry like '
> ', " ", << >>, { }, [ ] and so on. Also the arrow keys, negate,
> inverse.....well, I'll be d***** - It's *all the keys*!
>
> And without getting into too much detail, TI-Basic isn't really slower,
> generally speaking, than UserRPL. TI-Basic sucks at looping in most
> instances, and UserRPL really shines when the stack is utilized (which not
> that many people can do effectively really). TI-Basic is very limited, while
> programming the HP has almost no limits (that is when judging on-calc
> programming). The SDK & C development features are much better on the TI. By
> a large margin.
>
> > So i agree with VPN,the Titanic and for instances all TI68k are slow.
>
> You agree with Veli-Pekka in answering something you really do not know
> anything about - or that's how I see it.
>
> It's not just HP as a company that's deteriorating - the quality of this
> news group is going down the same hill (started around '99-'00 or so).
>
> Regards
> Steen
Veli-Pekka Nousiainen
X
> But we were talking about performance here, not features.
>> [VPN] - I Win - You loose, Man of Steen is rusty?
>
> enlightment and exciting new knowledge, not about mudthrowing, namecalling
> and arguments for the sole sake of "winning".
Sorry STeen - I forgot the smiley
My greatest worry is that the there seems to be a lack of
humor, happiness in your latest comments.
> The fact is that you, dare I say now, have
> again
BULL...
but let me clarify: I was talking about TI vs TI-Ti
The new Titanic is so called on the TI newsgroup
because there is _NO_ _NEW_ CPU
The users waited for a speed-up, like a 33,3 MHz CPU
I was not originally comparing to a HP, but the older TI
> spoken about something you
> don't know didly about. The TI89 Titanium - you've never touched it, and,
> I
> suspect, never even seen one. You have *never* used AMS 3.00, so how can
> you
> categorically state that it is slow? You don't know? You can't necessarily
> compare it to an ordinary HW2 TI89 - which I've incidently also pointed
> out
> can easily compete with the HP49G+ in many areas still. If you go back to
> my
> original post on this subject, I merely asked you if you had seen one.
> Apparantly not.
>
> Regards
> Steen
Lighten up, Steen!
[VPN]
Steen Schmidt
> humor, happiness in your latest comments.
Well, happiness...you are probably right, I tend to look at the glass more
half empty than half full these days. Sort of OT here though, but you'd
probably know how it is...
> BULL...
Easy now :-) Sorry for that remark though - it was out of place.
> but let me clarify: I was talking about TI vs TI-Ti
The original question was: Opinions of the Titanic and how does it compare
to the HP49G+?
You delivered one sentence: "The Titanic is f*ing slow"
I merely pointed out that you haven't tried the TI-Ti yourself (hence can't
comment on its speed) and that the TI89 is faster than the HP49G+ for some
things. I then ask; does that mean you think the HP49G+ is f*ing slow too?
Remember, the HP49G and the HP48G share processors, but I dare to say there
is a bit of a difference performance wise? You can't label something you
haven't tried out for yourself. You could have said that you suspected the
TI-Ti to be nothing more than a dressed-up TI89, but AMS 3.00 seemed to
eliminate some of the restrictions found in earlier versions.
> I was not originally comparing to a HP, but the older TI
In what areas is the Titanic slower than a HW2 TI89 then?
Regards
Steen
Samuel Nyall Stearley
> because there is _NO_ _NEW_ CPU
It got the nickname titanic by the tigcc developers out of annoyance.
Steen Schmidt
> It is your superiority complex and the fact that you think that you
> are smarter than everyone.
You know...this post of yours did almost make me send one very angry email
to you, but I counted to 100, and will try to be very calm answering it.
You wrote this sentence:
"He doesn't need to used a TI89 Titanium to know that it is slow."
That's false. You cannot speak about a thing you haven't tried. You don't
have the knowledge we speak of here in your genes, and it can't be
telepatically transferred. What is so hard to understand about the need for
first hand experience to *know* something? It's not enough to *think* you
now. It's not enough for it to be plausible or even probable. How about
trying "being sure" on for size?
> I thought that your failure with your advanced handheld project would
> have made you more humble or less arrogant but obviously i was wrong.
See, this was what made me very angry. You, my friend, do not know what
you're talking about here. Our (no, I wasn't alone in that adventure)
project was not a "failure". It wasn't even close. See, I still have
invested a good deal of the *money* I made from the sale of the IP, and I am
still co-owner (and founder) of a company here in Denmark (Realconsult).
Most back then were sold of to Toshiba, whom we worked very close with
designing around the XScale (try taking a peek at their PDA line-up a couple
of years back, and see if anything seems familiar)
I won't go into much detail as to why we stopped, but it's centered around
one of my friends getting killed in a motorbike accident on Nürburgring and
me getting divorced and having to fight for my kid.
> If you think that both HP and the comp.sys.hp48 are losing their
> quality,what are you doing here ?
I don't really know. Nostalgia? You haven't noticed that 90% of the '95-'00
regulars aren't here anymore? You don't think HP is dropping on the quality
scale and have been doing for some years? You sincerely mean that the tone
of c.s.hp48 today is reminiscent of what it was 6-8 years back?
> And contrary to what you said i do know exactly what i am talking.
So you have used the Titanic - in what areas is it slower than a HW2 TI89?
See, I don't know, and I'm curious (mainly because I'm curious about AMS
3.00).
> So unless i manage to always get TI68k with problems,i do know cleary
> what i am talking.
I agree with you about the original TI92/HP48 business - a sad story. And
the TI92 was buggy as h***.
> Your statements almost prove my point that the HP49G+ is generally
> faster than the TI68K.
Yes, we agree on this. I don't agree with the statement that the TI89 HW2 is
always slower than the HP49G+ though - that was all I wrote.
> For example when you say that the TI is faster than the HP49G+ for
> numeric sums.
> Are you talking of the built-in command or of all ways to do a sum ?
Built-in summation function.
> Also how the hell have you manage to get your TI being 2-4 times
> faster than your HP49G+ when my HP49G was faster than my TI92+ for
> that ?
Is your HP49G faster than the TI92+ for integer arithmetic? I don't have a
TI92+ at hand to try it, but I can provide some numbers for my HÅ49G+ and my
HW2 TI89;
Example:
TI89:
:Prgm
:For x,0,2000
:EndFor
:EndPrgm
runs in 14.789474 seconds on my TI89.
:Prgm
:For x,0,2000
:2528578459656*103
:EndFor
:EndPrgm
runs in 18.736842 seconds on my TI89, which means 507 integer
multiplications pr. second.
HP49G+:
<< #1d #1000d START NEXT >>
runs in 0.4371 seconds on my HP49G+.
<< #1d #1000d START 2528578459656 103 * DROP NEXT >>
runs in 5.6277 seconds on my HP49G+, which means 193 integer multiplications
pr. second (DROP adds just a very slight overhead).
Similar tests for various (very much larger) integers:
(210-digit int)+(231-digit int): HP=257/s & TI=404/s
(210-digit int)-(231-digit int): HP=243/s & TI=328/s
(145-digit int)*(97-digit int): HP=10/s & TI=27/s
(145-digit int)/(97-digit int): HP=20/s & TI=26/s
(6-digit int)^(2-digit int): HP=7/s & TI=22/s
Sqrt(6-digit int): HP=10/s & TI=40/s
Result: the TI89 HW2 is 2-4 times faster than the HP49G+ for basic integer
arithmetic.
Regards
Steen
Reth
Based on your posts here I'd be more than happy not to see you posting
again.
You've got problems
Without any respect: Rethnews:RVkEc.3385$dQ5...@news.get2net.dk...
feanor512
> > > I didn't know you could code in C directly in a TI-89
> >
>
> It seem like all the great software comes from them! 8-)
>
Here's another one
http://mxm.ticalc.org/ontic.html
Dunno if it works as I haven't tried either of these compilers.
Steen Schmidt
> again.
Non-constructive bullshit. Explain your statement please.
Steen
Veli-Pekka Nousiainen
>> Based on your posts here I'd be more than happy not to see you posting
>> again.
>
> Non-constructive bullshit. Explain your statement please.
>
You really have changed more to the depressive/aggressive
or should I say chynical side
Just try to lighten up and ignore hostality, remain cool
Try to be polite even when others are not
(Don't use you HP for counting to 100, nor a TI :)
The humor and happiness will come back
when the bad times are long forgotten
Keep on writing opinions, Steen!
{VPN}
PS: Be aware - I'm going to e-mail flames to your personal mailbox
(-;
Tom Lake
The only problem with that one is that you need to send the source back to a
PC to be compiled with TIGCC and then send the executable back to the TI.
With the other C, there's no need to use a PC at all.
Tom Lake
Paul Floyd
>> If you think that both HP and the comp.sys.hp48 are losing their
>> quality,what are you doing here ?
>
> I don't really know. Nostalgia? You haven't noticed that 90% of the '95-'00
> regulars aren't here anymore? You don't think HP is dropping on the quality
> scale and have been doing for some years? You sincerely mean that the tone
> of c.s.hp48 today is reminiscent of what it was 6-8 years back?
In my experience, all of Usenet has been going downhill for more than 6-8
years. Though uncorrelated, HP's quality is also headed downwards.
A bientot
Paul
--
Paul Floyd http://paulf.free.fr (for what it's worth)
Surgery: ennobled Gerald.
sple...@yahoo.com
regulars are gone, where did they go??? ICQ? hpcalc.org is still updated
currently - does everybody just check around there? Just curious..
Paul Floyd <ro...@127.0.0.1> wrote:
> On Tue, 29 Jun 2004 23:11:24 +0200, Steen Schmidt <ssch...@nospam.dk>
> wrote:
> >> If you think that both HP and the comp.sys.hp48 are losing their
> >> quality,what are you doing here ?
> >
> > I don't really know. Nostalgia? You haven't noticed that 90% of the
> > '95-'00 regulars aren't here anymore? You don't think HP is dropping on
> > the quality scale and have been doing for some years? You sincerely
> > mean that the tone of c.s.hp48 today is reminiscent of what it was 6-8
> > years back?
>
> In my experience, all of Usenet has been going downhill for more than 6-8
> years. Though uncorrelated, HP's quality is also headed downwards.
>
> A bientot
> Paul
--
Shawn Legrand
Carpe Noctem! Carpe Diem! Hieros Gamos!
sple...@yahoo.com
Timit? Hassan
you have a superiority complex and that you are quite arrogant but i
sincerely apologize for this:
> > I thought that your failure with your advanced handheld project would
> > have made you more humble or less arrogant but obviously i was wrong.
>
> See, this was what made me very angry. You, my friend, do not know what
> you're talking about here. Our (no, I wasn't alone in that adventure)
> project was not a "failure". It wasn't even close. See, I still have
> invested a good deal of the *money* I made from the sale of the IP, and I am
> still co-owner (and founder) of a company here in Denmark (Realconsult).
> Most back then were sold of to Toshiba, whom we worked very close with
> designing around the XScale (try taking a peek at their PDA line-up a couple
> of years back, and see if anything seems familiar)
>
> I won't go into much detail as to why we stopped, but it's centered around
> one of my friends getting killed in a motorbike accident on Nürburgring and
> me getting divorced and having to fight for my kid.
>
I have gone too far here and again i apologize.
> You wrote this sentence:
>
> "He doesn't need to used a TI89 Titanium to know that it is slow."
>
> That's false. You cannot speak about a thing you haven't tried. You don't
> have the knowledge we speak of here in your genes, and it can't be
> telepatically transferred. What is so hard to understand about the need for
> first hand experience to *know* something? It's not enough to *think* you
> now. It's not enough for it to be plausible or even probable. How about
> trying "being sure" on for size?
Here then it is you who don't know what you are talking.
If you have spent as much time as me on TI calculators and following
their evolution as close as i have you would know that:
*Since the TI89,TI has never tried to improve the speed of its TI68k
calculators by software optimisation
*TI68k generally get faster by significant hardware modification such
as faster C.P.U or fixing of a hardware problem such as the second
memory RAM bank need to enable the CPU to use the data bus in 16 bits
mode.
So knowing this and knowing that the TI89 Titanium is to the best a
TI89 Hw2 with more Flash ROM or to the worse a TI89 version of the
Voyage 200,i can easily conclude that the TI89 Titanic is not faster
than the TI89 HW2.
> > If you think that both HP and the comp.sys.hp48 are losing their
> > quality,what are you doing here ?
>
> I don't really know. Nostalgia? You haven't noticed that 90% of the '95-'00
> regulars aren't here anymore? You don't think HP is dropping on the quality
> scale and have been doing for some years? You sincerely mean that the tone
> of c.s.hp48 today is reminiscent of what it was 6-8 years back?
The HP quality has decreased anyone will agree with that.
Your second statement was attack toward me so what did you expect me
to do ?
To agree with you perhaps ?
> > And contrary to what you said i do know exactly what i am talking.
>
> So you have used the Titanic - in what areas is it slower than a HW2 TI89?
> See, I don't know, and I'm curious (mainly because I'm curious about AMS
> 3.00).
I don't and will never have a Titanic or a Voyage 200 or any
supossedly new TI68k which brought nothing besides more user Flash
ROM.
And i don't need to get a TI89 Titanium to make accurate predictions
about it.
The A.M.S 3.00 is certainly there just to handle hardware
modifications such as the USB,to fix some bugs and to add some new
limitations or to remove some old ones,nothing more nothing.
> > So unless i manage to always get TI68k with problems,i do know cleary
> > what i am talking.
>
> I agree with you about the original TI92/HP48 business - a sad story. And
> the TI92 was buggy as h***.
>
> > Your statements almost prove my point that the HP49G+ is generally
> > faster than the TI68K.
>
> Yes, we agree on this. I don't agree with the statement that the TI89 HW2 is
> always slower than the HP49G+ though - that was all I wrote.
Where did i say that the HP49G+ is always faster than the TI89 HW2 ?
Please tell me where the hell i wrote that ?
> > For example when you say that the TI is faster than the HP49G+ for
> > numeric sums.
> > Are you talking of the built-in command or of all ways to do a sum ?
>
> Built-in summation function.
This is possible as the built-in function doesn't seem optimised.
However with RPN program this is another matter.
> > Also how the hell have you manage to get your TI being 2-4 times
> > faster than your HP49G+ when my HP49G was faster than my TI92+ for
> > that ?
>
> Is your HP49G faster than the TI92+ for integer arithmetic?
Yes it was but unfortunately i can't do test anymore as my HP49G has
been stolen a several years ago.
I still have a TI92+ HW1 though.
>I don't have a
> TI92+ at hand to try it, but I can provide some numbers for my HÅ49G+ and my
> HW2 TI89;
>
> Example:
>
> TI89:
>
> :Prgm
> :For x,0,2000
> :EndFor
> :EndPrgm
>
> runs in 14.789474 seconds on my TI89.
>
> :Prgm
> :For x,0,2000
> :2528578459656*103
> :EndFor
> :EndPrgm
>
> runs in 18.736842 seconds on my TI89, which means 507 integer
> multiplications pr. second.
>
> HP49G+:
>
> << #1d #1000d START NEXT >>
>
> runs in 0.4371 seconds on my HP49G+.
>
> << #1d #1000d START 2528578459656 103 * DROP NEXT >>
>
> runs in 5.6277 seconds on my HP49G+, which means 193 integer multiplications
> pr. second (DROP adds just a very slight overhead).
Very curious but if what you say is true than the HP49G+ is more
disapointing than i thought.
> Similar tests for various (very much larger) integers:
>
> (210-digit int)+(231-digit int): HP=257/s & TI=404/s
> (210-digit int)-(231-digit int): HP=243/s & TI=328/s
> (145-digit int)*(97-digit int): HP=10/s & TI=27/s
> (145-digit int)/(97-digit int): HP=20/s & TI=26/s
> (6-digit int)^(2-digit int): HP=7/s & TI=22/s
> Sqrt(6-digit int): HP=10/s & TI=40/s
>
> Result: the TI89 HW2 is 2-4 times faster than the HP49G+ for basic integer
> arithmetic.
Your first comparaison give you a point but all your other
comparaisons give a clear advantage to the HP.
So what is exactly your point ?
I still don't see how you manage to conclude that the TI is faster
than the HP when almost all your tests show the opposite.
Btw try this on your TI89:
for i,1,3000:0->#("k"&string(k)):endfor
and tell me what happen.
But you should have a way to launch this other sequence:
for i,1,3000:delvar #("k"&string(k)):endfor
> Regards
> Steen
Steen Schmidt
I appreciate it. You unfortunately hit a sore spot there, so I reacted a
good deal more defensive than was necessary - I apologize as well.
> *Since the TI89,TI has never tried to improve the speed of its TI68k
> calculators by software optimisation
> *TI68k generally get faster by significant hardware modification such
> as faster C.P.U or fixing of a hardware problem such as the second
> memory RAM bank need to enable the CPU to use the data bus in 16 bits
> mode.
I agree, but haven't seen AMS 3.00 - I wouldn't be too surprised if your
predictions are correct though.
> So knowing this and knowing that the TI89 Titanium is to the best a
> TI89 Hw2 with more Flash ROM or to the worse a TI89 version of the
> Voyage 200,i can easily conclude that the TI89 Titanic is not faster
> than the TI89 HW2.
I never said I expected the Titanic to be faster than a HW2 TI89, but I
expect it to be very close performance wise, if not identical with it.
> Your second statement was attack toward me so what did you expect me
> to do ?
It wasn't meant as an attack on you personally, so I'm sorry if you got
stung.
> > Yes, we agree on this. I don't agree with the statement that the TI89
HW2 is
> > always slower than the HP49G+ though - that was all I wrote.
>
> Where did i say that the HP49G+ is always faster than the TI89 HW2 ?
It wasn't your statement I was referring to. You just seemed to back
Veli-Pekkas initial statement ("The Titanic is f*ing slow"), which I
extrapolated into VPN meaning the Titatinic was slower than the HP49G+
without exception. With the above sentence I tried to explain why I entered
this discussion in the first place - namely because I think it's too simple
an answer to give a guy who asks sincerely. But, what the heck - this is an
HP newsgroup, what should he expect when asking something like that here :-)
> This is possible as the built-in function doesn't seem optimised.
> However with RPN program this is another matter.
Yes, for numeric sums there is possibly some performance to gain.
> > Similar tests for various (very much larger) integers:
> >
> > (210-digit int)+(231-digit int): HP=257/s & TI=404/s
> > (210-digit int)-(231-digit int): HP=243/s & TI=328/s
> > (145-digit int)*(97-digit int): HP=10/s & TI=27/s
> > (145-digit int)/(97-digit int): HP=20/s & TI=26/s
> > (6-digit int)^(2-digit int): HP=7/s & TI=22/s
> > Sqrt(6-digit int): HP=10/s & TI=40/s
> >
> > Result: the TI89 HW2 is 2-4 times faster than the HP49G+ for basic
integer
> > arithmetic.
>
> Your first comparaison give you a point but all your other
> comparaisons give a clear advantage to the HP.
You misunderstand my sorry attempt at providing units - in the above higher
numbers are better, as the unit is s^-1 (10/s -> 10 per second -> 10 Hz if
you will). The TI is faster in *all* the above tests. For comparison, I have
just completed the same set of tests on my HP49G:
(210-digit int)+(231-digit int): HP49G=128/s
(210-digit int)-(231-digit int): HP49G=128/s
(145-digit int)*(97-digit int): HP49G=4.5/s
(145-digit int)/(97-digit int): HP49G=8/s
(6-digit int)^(2-digit int): HP49G=4/s
Sqrt(6-digit int): HP49G=3.5/s
The above means (for this particular test) that the HP49G+ is 2-3 times
faster than the HP49G for integer arithmetic, and consequently that the TI89
is 3-11 times faster than the HP49G.
> Btw try this on your TI89:
> for i,1,3000:0->#("k"&string(k)):endfor
You mean
for i,1,3000:0->#("k"&string(i)):endfor
right?
> and tell me what happen.
I got a memory error after a while (and one sluggish TI89 :-).
Regards
Steen
Bhuvanesh
> Here then it is you who don't know what you are talking.
> If you have spent as much time as me on TI calculators and following
> their evolution as close as i have you would know that:
> *Since the TI89,TI has never tried to improve the speed of its TI68k
> calculators by software optimisation
You can't know this for a fact unless you are a developer or tester at
TI or have tested each function extensively in each new AMS release.
> So knowing this and knowing that the TI89 Titanium is to the best a
> TI89 Hw2 with more Flash ROM or to the worse a TI89 version of the
> Voyage 200,i can easily conclude that the TI89 Titanic is not faster
> than the TI89 HW2.
No, you cannot conclude that without actually doing tests with the
TI-89 Titanium. It is quite likely that you are correct in your
assertion, but there is a chance that you are not.
> > I don't really know. Nostalgia? You haven't noticed that 90% of the '95-'00
> > regulars aren't here anymore? You don't think HP is dropping on the quality
> > scale and have been doing for some years? You sincerely mean that the tone
> > of c.s.hp48 today is reminiscent of what it was 6-8 years back?
>
> The HP quality has decreased anyone will agree with that.
> Your second statement was attack toward me so what did you expect me
> to do ?
I don't see why you think that statement was an attack on you. I think
Steen was simply commenting on the general quality of posts.
> Btw try this on your TI89:
> for i,1,3000:0->#("k"&string(k)):endfor
> and tell me what happen.
> But you should have a way to launch this other sequence:
> for i,1,3000:delvar #("k"&string(k)):endfor
I know, you brought this up earlier too. Oddly enough, I don't see it
in the bug list; I'll add it.
Cheers,
Bhuvanesh.
Bhuvanesh
> Here then it is you who don't know what you are talking.
> If you have spent as much time as me on TI calculators and following
> their evolution as close as i have you would know that:
> *Since the TI89,TI has never tried to improve the speed of its TI68k
> calculators by software optimisation
You can't know this for a fact unless you are a developer or tester at
TI or have tested each function extensively in each new AMS release.
> So knowing this and knowing that the TI89 Titanium is to the best a
> TI89 Hw2 with more Flash ROM or to the worse a TI89 version of the
> Voyage 200,i can easily conclude that the TI89 Titanic is not faster
> than the TI89 HW2.
No, you cannot conclude that without actually doing tests with the
TI-89 Titanium. It is quite likely that you are correct in your
assertion, but there is a chance that you are not.
> > I don't really know. Nostalgia? You haven't noticed that 90% of the '95-'00
> > regulars aren't here anymore? You don't think HP is dropping on the quality
> > scale and have been doing for some years? You sincerely mean that the tone
> > of c.s.hp48 today is reminiscent of what it was 6-8 years back?
>
> The HP quality has decreased anyone will agree with that.
> Your second statement was attack toward me so what did you expect me
> to do ?
I don't see why you think that statement was an attack on you. I think
Steen was simply commenting on the general quality of posts.
> Btw try this on your TI89:
> for i,1,3000:0->#("k"&string(k)):endfor
> and tell me what happen.
> But you should have a way to launch this other sequence:
> for i,1,3000:delvar #("k"&string(k)):endfor
I know, you brought this up earlier too. Oddly enough, I don't see it
Timit? Hassan
> timi...@yahoo.fr (Timit? Hassan) wrote:
> > Here then it is you who don't know what you are talking.
> > If you have spent as much time as me on TI calculators and following
> > their evolution as close as i have you would know that:
> > *Since the TI89,TI has never tried to improve the speed of its TI68k
> > calculators by software optimisation
>
> You can't know this for a fact unless you are a developer or tester at
> TI or have tested each function extensively in each new AMS release.
I am neither a tester or a developer but i have talked with a member
of TI R&D team and with people who work closely with TI for the
developpement of the TI89 and the TI92+.
And the conclusion of these numerous discussions is that TI doesn't
want to significantly improve their software otherwise they would have
fixed numerous problems known since the TI92 release in 1995/96.
TI has even refused to use algorithms freely given to them by a famous
french Math teacher who worked with them on the TI89/92+.
His name is among the names of the TI89/92+ developpers available on
A.M.S 1.x.
> > So knowing this and knowing that the TI89 Titanium is to the best a
> > TI89 Hw2 with more Flash ROM or to the worse a TI89 version of the
> > Voyage 200,i can easily conclude that the TI89 Titanic is not faster
> > than the TI89 HW2.
>
> No, you cannot conclude that without actually doing tests with the
> TI-89 Titanium. It is quite likely that you are correct in your
> assertion, but there is a chance that you are not.
You have the two calcs thus i assume that you have done the tests.
So enlighten us,is the Titanic faster than the TI89 HW2 for even one
kind of computation ?
> > > I don't really know. Nostalgia? You haven't noticed that 90% of the '95-'00
> > > regulars aren't here anymore? You don't think HP is dropping on the quality
> > > scale and have been doing for some years? You sincerely mean that the tone
> > > of c.s.hp48 today is reminiscent of what it was 6-8 years back?
> >
> > The HP quality has decreased anyone will agree with that.
> > Your second statement was attack toward me so what did you expect me
> > to do ?
>
> I don't see why you think that statement was an attack on you. I think
> Steen was simply commenting on the general quality of posts.
Ok,i has perhaps overreacted.
> > Btw try this on your TI89:
> > for i,1,3000:0->#("k"&string(k)):endfor
> > and tell me what happen.
> > But you should have a way to launch this other sequence:
> > for i,1,3000:delvar #("k"&string(k)):endfor
>
> I know, you brought this up earlier too. Oddly enough, I don't see it
> in the bug list; I'll add it.
This is not a bug.
This is due to a memory limitation.
If i have to guess,i would say that the V.A.T has a limited size of 64
KB.
However,TI should prevent the TI68k to reach this limit and to become
almost unusable.
Btw this problem existed on the TI92 but the limit was almost 800
variables.
> Cheers,
> Bhuvanesh.
Timit? Hassan
> I appreciate it. You unfortunately hit a sore spot there, so I reacted a
> good deal more defensive than was necessary - I apologize as well.
So we can move on ;-)
> > *Since the TI89,TI has never tried to improve the speed of its TI68k
> > calculators by software optimisation
> > *TI68k generally get faster by significant hardware modification such
> > as faster C.P.U or fixing of a hardware problem such as the second
> > memory RAM bank need to enable the CPU to use the data bus in 16 bits
> > mode.
>
> I agree, but haven't seen AMS 3.00 - I wouldn't be too surprised if your
> predictions are correct though.
Yes,unfortunately most of my predictions about TI68k are correct and
this explains why i am very bitter toward TI.
However i could be wrong on some very specific points.
I was for example surprised by the incredible improvements of the
solve functions from A.M.S 1.x to A.M.S 2.x.
It is outstanding to see what the new solve function can do.
It can even work as multiple equation solver.
> > So knowing this and knowing that the TI89 Titanium is to the best a
> > TI89 Hw2 with more Flash ROM or to the worse a TI89 version of the
> > Voyage 200,i can easily conclude that the TI89 Titanic is not faster
> > than the TI89 HW2.
>
> I never said I expected the Titanic to be faster than a HW2 TI89, but I
> expect it to be very close performance wise, if not identical with it.
Yes but i heard some people complain about the Voyage 200 being slower
than their TI92 thus i wouldn't be surprised if the Titanic is slower
than the TI89 HW2 for some operations.
Btw in the time i was very surprised to see that the TI92+ was slower
than the TI92II for several operations like numeric
summations,TI-Basic for loops,etc...
> > Your second statement was attack toward me so what did you expect me
> > to do ?
>
> It wasn't meant as an attack on you personally, so I'm sorry if you got
> stung.
Sorry,i over reacted.
> > > Yes, we agree on this. I don't agree with the statement that the TI89
> HW2 is
> > > always slower than the HP49G+ though - that was all I wrote.
> >
> > Where did i say that the HP49G+ is always faster than the TI89 HW2 ?
>
> It wasn't your statement I was referring to. You just seemed to back
> Veli-Pekkas initial statement ("The Titanic is f*ing slow"), which I
> extrapolated into VPN meaning the Titatinic was slower than the HP49G+
> without exception. With the above sentence I tried to explain why I entered
> this discussion in the first place - namely because I think it's too simple
> an answer to give a guy who asks sincerely. But, what the heck - this is an
> HP newsgroup, what should he expect when asking something like that here :-)
Yes ;-)
In fact i agree with VPN because the TI68k is slow for any serious
thing i try to do with it.
I used to write big and complex TI-Basic programs such as RPN
interface,symbolic interactive solver,etc...
Or even some simpler TI-basic like symbolic eigenvalues and
eigenvectors computations.
And whatever i did those programs were often too slow for my taste
even when for loops were not involved.
Btw do you know the funny reason why For loop is so slow on the TI68k
?
Off course i can write C programs but i would have preferred to have
this possibility on calc.
Because if i have to write C programs on PC,then i prefer to write
them for the PC itself.
> You misunderstand my sorry attempt at providing units - in the above higher
> numbers are better, as the unit is s^-1 (10/s -> 10 per second -> 10 Hz if
> you will). The TI is faster in *all* the above tests. For comparison, I have
> just completed the same set of tests on my HP49G:
>
> (210-digit int)+(231-digit int): HP49G=128/s
> (210-digit int)-(231-digit int): HP49G=128/s
> (145-digit int)*(97-digit int): HP49G=4.5/s
> (145-digit int)/(97-digit int): HP49G=8/s
> (6-digit int)^(2-digit int): HP49G=4/s
> Sqrt(6-digit int): HP49G=3.5/s
>
> The above means (for this particular test) that the HP49G+ is 2-3 times
> faster than the HP49G for integer arithmetic, and consequently that the TI89
> is 3-11 times faster than the HP49G.
Ok,i misunderstood as i though that you were giving computation time.
This is dramatic.
It looks like i was right not to buy the HP49G+ after all.
I have never used loops in my comparaisons.
I tried almost all functions both of the calculators have in common
and i test on rather big integers and the HP49G was almost always
faster than my TI92+ besides for matrix of integer computations.
The awful display time of the TI68k was a bit responsible of that but
i thought that the HP49G+,being several times faster than the
HP49G,would be faster the TI89 HW2 even without taking in
consideration the dispaly time.
> > Btw try this on your TI89:
> > for i,1,3000:0->#("k"&string(k)):endfor
>
> You mean
>
> for i,1,3000:0->#("k"&string(i)):endfor
>
> right?
Yes,in fact i used often k as the iterator on the TI92+ but those last
times i write only Java and C++ programs so now i use i more often as
an iterator.
> > and tell me what happen.
>
> I got a memory error after a while (and one sluggish TI89 :-).
Yes another problem which should have been fixed since the first A.M.S
as it was known since the first TI92.
> Regards
> Steen