Nspire CX -- Maybe it's time for me to give up on calculators?
figureloop
I have been learning to use the new TI Nspire CX CAS. I have used TI
calcs from the TI-82, TI-83, TI-89, and now this. I skipped the
Voyage, and the early Nspire calcs because they didn't seem much more
useful than the TI-89. I also use Mathematica for more complicated
problems.
But I have little desire to invest in a laptop just to have MMA in the
rare cases that I want to do math on the go. For that, and also for
the frequent cases where I just want to pick up a calc, turn it on and
instantly compute something, a calculator is still handy. Also, since
I'm addicted to the algebraic entry, and I do a lot of complex math
(I'm into EE), an old-school calculator isn't for me.
Unfortunately, I'm finding the CX to be far more cumbersome to use
than I had hoped. The "document" concept seems like it could be
useful, but it's unable to reach it's potential with a non-QWERTY
keyboard. It seems like I have to do too much button pushing to get
things done, more so than on the TI-89. That said, the alpha kbd. on
the CX is of course better than the TI-89!
A few things also have me stumped or just a bit disappointed:
1. Mouse cursor fetching problem -- it seems like in some contexts if
I beat on the touchpad, the cursor will come out, but it's not
dependable. Perplexing.
2. I have to press a button to use scratchpad when I first turn it
on. Can it be set to just default to whatever you were doing when you
last powered down?
3. It seems like the whole universe of the graphing calc. is too
education oriented. Also, the desire to have a "locked-down" platform
with unnecessary limitations is incompatible with my usage hopes and
philosophy. I'm used to running my math software on a Linux PC
connected to embedded systems that I develop, and with my choice of
additional and free programming languages all over the place.
4. I had hoped the Lua scripting might be capable of tooling the
handheld up for use as a controller/readout for external
instrumentation. I'm not really sure if this is possible yet at this
point, but the overall attitude of "lock-down" surrounding this
machine inclines me toward not bothering to invest serious effort in
using it for these sorts of things.
5. The guide for the apps is all in the context of the PC software,
rather than the particularities of the handheld.
Anyway, I'm not trying to diss the CX, it's a great improvement over
the older generations of hardware and I seriously hope it helps
students and teachers accomplish their goals. But with the 900 pages
of documentation I'd have to read to fully "get" all the new apps,
concepts, and review the reference for the built-in functions, whereas
I am already fairly proficient with Mathematica and using a Linux PC:
I'm considering reaching the conclusion that I should just drop the
idea of a powerful hand-held engineering calculator anymore, and just
bite the bullet and buy a mid-range laptop and install Mathematica and
my favorite languages on it.
Has anyone else reached the same conclusion?
Thanks for comments.
P.S. Don't get me started on HP calcs.
Jim Fullerenex
--
To post to this group, send email to tins...@googlegroups.com
To unsubscribe send email to tinspire+u...@googlegroups.com
For more options, visit this group at
http://groups.google.com.au/group/tinspire?hl=en-GB?hl=en-GB
The tns documents shared by group members are archived at
https://sites.google.com/site/tinspiregroup/classroom-news/welcome-abouttime
Lionel Debroux
> The "document" concept seems like it could be useful, but it's unable
> to reach it's potential with a non-QWERTY keyboard.
I don't disagree with you, I'd just like to note that I think that part
of the motivation behind documents is to have separate sets of variables
contained into each document, which don't collide with each other.
> It seems like I have to do too much button pushing to get things
> done, more so than on the TI-89.
This matches my experience with the Nspire.
> That said, the alpha kbd. on the CX is of course better than the
> TI-89!
I disagree with the "of course" ;)
To me, larger keys, such as the ones found on TI-Z80 and TI-68k calcs,
are much easier to type with than the minuscule alpha keys of the CX.
And I have long, small fingers; people with fat fingers must be
considering the keyboard as quite unpleasant.
I know that the computer software provided alongside modern Nspires in
most countries alleviates the burden of typing large amounts of text on
such a small keyboard... but the computer software precisely cannot
fulfill at least one important (to me, anyway) use case: typing notes on
the go, up to ten kilobytes.
Mixing text and figures is easier on the TI-Z80 and TI-68k keyboards
than on the CX keyboard: text and figures are on the same keys, you
don't have to move your fingers and eyes as much.
I'm aware that my views are not universally shared: some people hate
having to use the alpha key.
> A few things also have me stumped or just a bit disappointed:
>
> 1. Mouse cursor fetching problem -- it seems like in some contexts
> if I beat on the touchpad, the cursor will come out, but it's not
> dependable. Perplexing.
I'd say that this is one of the aspects of the general mild/low quality
of the touchpad (and/or touchpad handling code), about which a number of
people are complaining.
> 2. I have to press a button to use scratchpad when I first turn
> it on. Can it be set to just default to whatever you were doing
> when you last powered down?
Yeah, like DIAMOND + ON on the TI-68k/AMS platform.
> 3. It seems like the whole universe of the graphing calc. is too
> education oriented. Also, the desire to have a "locked-down"
> platform with unnecessary limitations is incompatible with my usage
> hopes and philosophy. I'm used to running my math software on a
> Linux PC connected to embedded systems that I develop, and with my
> choice of additional and free programming languages all over the
> place.
Wholeheartedly agreed, but this group is definitely not the place to
find a great many people sympathetic to your - and my - views.
The open development community is ;)
> 4. I had hoped the Lua scripting might be capable of tooling the
> handheld up for use as a controller/readout for external
> instrumentation. I'm not really sure if this is possible yet at
> this point, but the overall attitude of "lock-down" surrounding
> this machine inclines me toward not bothering to invest serious
> effort in using it for these sorts of things.
It's possible to output, from Lua, some data (without 0x00 bytes...) to
the hardly-accessible serial port that is part of the proprietary dock
connector with non-standard spacing at the bottom. People from the open
development community have been using it as a proof of concept for
hooking up a speaker or LEDs.
I'm not aware that Lua can read out data from the serial port, though.
The TI-Z80 and TI-68k platforms can produce and consume data on the
legacy I/O, and USB (84+(SE) & 89T) ports.
> 5. The guide for the apps is all in the context of the PC
> software, rather than the particularities of the handheld.
I can't comment on that, but I'll trust you :)
> Anyway, I'm not trying to diss the CX, it's a great improvement over
> the older generations of hardware and I seriously hope it helps
> students and teachers accomplish their goals.
Indeed.
> But with the 900 pages of documentation I'd have to read to fully
> "get" all the new apps, concepts, and review the reference for
> the built-in functions, whereas I am already fairly proficient with
> Mathematica and using a Linux PC: I'm considering reaching the
> conclusion that I should just drop the idea of a powerful hand-held
> engineering calculator anymore, and just bite the bullet and buy a
> mid-range laptop and install Mathematica and my favorite languages
> on it.
>
> Has anyone else reached the same conclusion?
Yes :)
Your other e-mail shows that you have very well realized that nowadays,
features in calculators are dictated by people with a mind of lockdown
and limitation: authorities regulating standardized tests, managers who
see creative usage of their calculators as a threat (instead of as a
win-win situation, which it had been for more than ten years before that
with the TI-Z80 and TI-68k series), etc.
Engineers and enthusiasts, like you and me, have different
requirements... but they're clearly not (no longer) the driving force in
the highly captive market you have correctly identified. The people who
decide about the calculators' features see them as pure teaching tools,
in a gated environment - and it's sad, to the opinion of a number of us.
In the open development community, there would be much fewer negative
feelings about the Nspire Clickpad / Touchpad / CX (CAS), if it had two
modes of operation *used at their full potential*:
* a locked-down mode for tests - which the Nspire series has, under the
form of the PTT mode;
* an open mode for usage by enthusiasts, engineers, etc., with official
documentation and SDK (like the TI-Z80 and TI-68k historically had for
some time, even if it was significantly incomplete) that helps people to
use the full potential of the hardware they have bought, and therefore
own, in the ways they see fit. That's what's missing on the Nspire series.
Reverse-engineering is a slow process, especially on such large amounts
of code, and a number of people interested in those aspects have stuck
to the TI-Z80 and TI-68k series (which haven't been used to their full
potential yet !) or moved on to other platforms...
Open development platforms definitely have an edge over closed platforms
when it comes to engineering: if a feature is missing, one can more
easily add it.
That's what occurred for many years with the TI-Z80 and TI-68k series
(well, there have been many more games than science programs, but that's
because games are easier than science programs), and which cannot really
occur with the Nspire series.
Regards,
Lionel.
Don Shepherd
You say the NSpire seems too "education-oriented." Of course, we've
been telling you for months now that the whole purpose of the NSpire
is education--students and teachers. The NSpire wasn't designed for
engineers, scientists, or professionals doing math research; those are
other markets. The NSpire was designed to teach kids about math in
innovative ways, and I think it succeeds there.
Is it perfect? Of course not. Your experience getting the cursor to
respond, I've had that problem too, and it's frustrating. What I'd
really like to see in the NSpire is a stylus and touch-sensitive
screen. While the touchpad is an improvement over the clickpad,
neither is as good as a stylus to give you exact control over the
interface.
I've resisted buying the new color NSpire because they don't have the
trade-in program anymore for teachers. When TI announces a stylus-
equipped NSpire CAS (OS4 maybe?), I'll get one.
Don Shepherd
Lionel Debroux
> You say the NSpire seems too "education-oriented." Of course, we've
> been telling you for months now that the whole purpose of the NSpire
> is education--students and teachers. The NSpire wasn't designed for
> engineers, scientists, or professionals doing math research; those
> are other markets. The NSpire was designed to teach kids about math
> in innovative ways, and I think it succeeds there.
I did understand, early on, that the purpose of the Nspire was teaching
math and science (well, at least, science minus computer science, since
until recently, the only way to program the Nspire on-calc was the BASIC
whose abilities, in some areas, are lower than those of the 20-year-old
TI-81...).
It was clear, from the feature set, that TI was not interested in a
subset of its (former) users anymore :)
But we aren't going to rehash past discussions here.
You all know what I think of a number of premises that led to the Nspire
as we know it, and I know that people can have different views and needs :)
It's likely that I wouldn't currently have any Nspire if I wasn't the
maintainer of TILP, and I have a CX CAS only because it was given to me.
Lionel.
figureloop
> To be honest, I was literally praying for a non-qwerty keyboard before I
> bought Nspire CAS (grayscale)----- collegeboard forbids any calculator with
> a qwerty. For students a handheld that cannot be used in APs and SAT is
> close to useless.
> Also, I believe the "document" concept of Nspire is the most revoluntionary
> part.
> There's not much meaning to compare computer software like Mathematica to
> Nspire handheld. As for mathematical capability, the former is way more
> powerful than the latter. I won't bother to recommend Hp calcs unless they
> get higher screen resolution.
>
> Best Wishes,
> Jim
Please help me understand the testing and calculator issue. I was
under the impression that calculators with CAS are banned. Is this
correct, or not?
If it is true that the CAS is banned from tests, then the CAS
calculator could have had its keys rearranged as QWERTY, to make it a
more productive power-math calculator.
It's not quite that simple, since the "EE" and other keys would have
to be rearranged as well. Whatever.
I think the ability to work in a document is a useful enhancement over
what I am used to with the TI-89. I can sure think of ways that the
whole environment could be better unified however.
For ex: Like Python, have the documents actually constitute a library
namespace. So if you have a document "myfunc" that defines a function
f, then that function could be accessible from any other document by:
import("myfunc")
myfunc.f(2)
<result>
Also, I wish functions were first class objects.
figureloop
[a lot edited]
>
Thanks for the interesting response, Lionel.
> Wholeheartedly agreed, but this group is definitely not the place to
> find a great many people sympathetic to your - and my - views.
> The open development community is ;)
though is fighting it. I mean, if TI and the educators want to have
their thing, so I can either buy it can accept it for what it is, or
buy something else that is "open" or has whatever feature I want. Why
would I buy a machine knowing full well that the manufacturer is
trying to keep it canned in some way, only to try to fight it, hack
it, then get upset when they break my hacks with the next OS release?
Is it price that drives this behavior? I mean, that to have a
computing platform at all which one can explore may be beyond the
means of some and so they only have a $150 calc. to work with?
Actually I'd have a hard time believing that the skills of the
remarkable hackers which develop tools, kernels, new languages, etc.
for these calcs. could possibly have been acquired unless they also
had access to regular PCs, the internet, etc. So again, why bother
fighting it? Just program something else which is open, like the PC.
> 4. I had hoped the Lua scripting might be capable of tooling the
> > handheld up for use as a controller/readout for external
> > instrumentation. I'm not really sure if this is possible yet at
> > this point, but the overall attitude of "lock-down" surrounding
> > this machine inclines me toward not bothering to invest serious
> > effort in using it for these sorts of things.
>
> It's possible to output, from Lua, some data (without 0x00 bytes...) to
> the hardly-accessible serial port that is part of the proprietary dock
> connector with non-standard spacing at the bottom. People from the open
> development community have been using it as a proof of concept for
> hooking up a speaker or LEDs.
> I'm not aware that Lua can read out data from the serial port, though.
sound of things, it's remarkable that this isn't some proprietary
encrypted link. Maybe I shouldn't give them ideas ;-)
Wait a minute, how does the Vernier stuff work? It's the fact that
the calc. was designed to connect to sensor instruments that often
perked my interest in the possibility of building my own instruments
and driving them with the calc. There must be a data input pathway
for the Vernier stuff to work, huh?
> > Mathematica and using a Linux PC: I'm considering reaching the
> > conclusion that I should just drop the idea of a powerful hand-held
> > engineering calculator anymore, and just bite the bullet and buy a
> > mid-range laptop and install Mathematica and my favorite languages
> > on it.
>
> > Has anyone else reached the same conclusion?
>
> Yes :)
> Your other e-mail shows that you have very well realized that nowadays,
> features in calculators are dictated by people with a mind of lockdown
> and limitation: authorities regulating standardized tests, managers who
> see creative usage of their calculators as a threat (instead of as a
> win-win situation, which it had been for more than ten years before that
> with the TI-Z80 and TI-68k series), etc.
the DSPs, and of course they work very hard to make them useful for
whatever people want to do. That's the point.
> Engineers and enthusiasts, like you and me, have different
> requirements... but they're clearly not (no longer) the driving force in
> the highly captive market you have correctly identified. The people who
> decide about the calculators' features see them as pure teaching tools,
> in a gated environment - and it's sad, to the opinion of a number of us.
calculator could and should remain a serious supplementary math tool
after most engineers and scientists around me turned entirely to the
PC with Matlab and Mathematica, and where they needed a calculator,
the pathetic Windows calculator.
Maybe TI was right. The engineer market faded out. Why bother
competing with the PC applications?
> In the open development community, there would be much fewer negative
> feelings about the Nspire Clickpad / Touchpad / CX (CAS), if it had two
> modes of operation *used at their full potential*:
> * a locked-down mode for tests - which the Nspire series has, under the
> form of the PTT mode;
> * an open mode for usage by enthusiasts, engineers, etc., with official
> documentation and SDK (like the TI-Z80 and TI-68k historically had for
> some time, even if it was significantly incomplete) that helps people to
> use the full potential of the hardware they have bought, and therefore
> own, in the ways they see fit. That's what's missing on the Nspire series.
> Reverse-engineering is a slow process, especially on such large amounts
> of code, and a number of people interested in those aspects have stuck
> to the TI-Z80 and TI-68k series (which haven't been used to their full
> potential yet !) or moved on to other platforms...
hacking the device, or are they just ignorant of the desire of
customers to do this? I mean, is there evidence of hostility, or is
it just an assumption. On some of the forums, I noticed a recent
letter exchange with a TI insider and a calc. hacker, and it seemed
like there was the possibility for constructive dialog. Reading some
other stuff on the hacker forums made me worry that people were taking
an immature attitude and whining that they couldn't have what they
wanted, when it was unclear if they had ever just expressed politely
their wishes. If we attack TI and say "you are waging war against us,
it's not fair, now let us do what we want!" The response is likely to
be different and less constructive than if a different attitude were
used.
For this reason, I have just today employed the assistance of a
colleague at work before sending off an inportant communication to
upper management. I want to make sure my personal feelings aren't
clouding my message so that my message will be heard as mature rather
than whining.
> Regards,
> Lionel.
Good day and thanks again for responding.
Eric Findlay
2) if you want the dot operator method and/or want to shorten a long library name, there is the libshortcut("library","shortcut") function.
Example: libshortcut("myLib", "m") allows m.a2()
--Eric
figureloop
> 1) that is how the libraries work. If you have a properly defined library called myLib.tns with a function called a2, then you can access it from any document via myLib\a2(). You don't even have to use an import command.
>
> 2) if you want the dot operator method and/or want to shorten a long library name, there is the libshortcut("library","shortcut") function.
> Example: libshortcut("myLib", "m") allows m.a2()
>
> --Eric
Jim Fullerenex
>under the impression that calculators with CAS are banned. Is this
>correct, or not?
--
lee kucera
Lionel Debroux
> > Wholeheartedly agreed, but this group is definitely not the place
> > to find a great many people sympathetic to your - and my - views.
> > The open development community is ;)
> I've picked up on the situation. One thing I don't see the point of
> though is fighting it.
The challenge of opening the platform, the motivation of learning things
in the process, honing one's reverse-engineering and engineering skills,
the right to exercise the widely-recognized "freedom to tinker" on the
devices we own, adding lines to the loooong and ever growing list of
closed platforms forced open, etc. ... there are lots of possible
motivations for fighting it ;)
> I mean, if TI and the educators want to have their thing, so I can
> either buy it can accept it for what it is, or buy something else
> that is "open" or has whatever feature I want. Why would I buy a
> machine knowing full well that the manufacturer is trying to keep it
> canned in some way, only to try to fight it, hack it, then get upset
> when they break my hacks with the next OS release?
It depends on who the buyers are.
Nowadays, in some countries with a captive education market, the
populations of buyers has largely moved from individuals to "school
institutions", as a generic term.
So, the situation, for a subset of users, is that they do have the
calculators anyway. And they have the entire right to use it at its full
potential.
However, in markets driven by individuals and word of mouth (or which
have been at some time in the past), I tend to agree with you: in 2000,
I bought a 89 instead of a Casio Graph 100, precisely because I had seen
someone doing advanced stuff on the 89 - a consequence of the greater
openness of the 89. The CAS of the 89 looked faster and more powerful
than that of the Graph 100, too.
> Is it price that drives this behavior? I mean, that to have a
> computing platform at all which one can explore may be beyond the
> means of some and so they only have a $150 calc. to work with?
> Actually I'd have a hard time believing that the skills of the
> remarkable hackers which develop tools, kernels, new languages, etc.
> for these calcs. could possibly have been acquired unless they also
> had access to regular PCs, the internet, etc. So again, why bother
> fighting it? Just program something else which is open, like the PC.
Such skills haven't been acquired purely through calculators indeed.
However, as a matter of fact, dozens of thousands of persons, in today's
IT industry, have started serious, continued programming on calculators,
especially TI-Z80 or TI-68k calculators. Among many people in the open
development community, I did myself.
At the end of the 1990s and the beginning of the 2000s, calculators were
the only portable device that millions of people could bring pretty much
everywhere, and use + program *on the go*, through various forms of
BASIC. Again, many of us did exactly this.
Nowadays, most calculator users in high school and university have, for
instance, smartphones, which are much more powerful than calculators.
They lack on-smartphone programmability, though.
> > > 4. I had hoped the Lua scripting might be capable of tooling the
> > > handheld up for use as a controller/readout for external
> > > instrumentation. I'm not really sure if this is possible yet at
> > > this point, but the overall attitude of "lock-down" surrounding
> > > this machine inclines me toward not bothering to invest serious
> > > effort in using it for these sorts of things.
> > It's possible to output, from Lua, some data (without 0x00
> > bytes...) to the hardly-accessible serial port that is part of
> > the proprietary dock connector with non-standard spacing at the
> > bottom. People from the open development community have been using
> > it as a proof of concept for hooking up a speaker or LEDs.
> > I'm not aware that Lua can read out data from the serial port,
> > though.
> Yes I noticed the use of the serial link from the connector. By the
> sound of things, it's remarkable that this isn't some proprietary
> encrypted link. Maybe I shouldn't give them ideas ;-)
Heh ;)
The many-pin dock connector including the serial port doesn't seem to be
available on the Nspire CM series, though:
http://ti.bank.free.fr/index.php?mod=news&ac=commentaires&id=1278 (in
French).
> Wait a minute, how does the Vernier stuff work?
Older models of lab equipment from Vernier used the "legacy" I/O port
(proprietary flavour of a 2.5mm stereo jack, the tip and ring are
different from a standard 2.5mm stereo jack).
> It's the fact that the calc. was designed to connect to sensor
> instruments that often perked my interest in the possibility of
> building my own instruments and driving them with the calc. There
> must be a data input pathway for the Vernier stuff to work, huh?
Vernier's newest models communicate through USB, with software protocols
(implemented by the Nspire OS >= 3.0) that are unknown to me, but might
be documented somewhere. I haven't looked.
> > Engineers and enthusiasts, like you and me, have different
> > requirements... but they're clearly not (no longer) the driving
> > force in the highly captive market you have correctly identified.
> > The people who decide about the calculators' features see them
> > as pure teaching tools, in a gated environment - and it's sad,
> > to the opinion of a number of us.
> I guess I'm strange in having held on to the notion that the
> calculator could and should remain a serious supplementary math
> tool after most engineers and scientists around me turned entirely
> to the PC with Matlab and Mathematica, and where they needed a
> calculator, the pathetic Windows calculator.
>
> Maybe TI was right. The engineer market faded out. Why bother
> competing with the PC applications?
The thing is, they could have focused on education (it's their entire
right) without slamming hard the door in the face of the enthusiast user
base that had helped the popularity of their products (in some markets,
at least) for more than 10 years :)
Calculator makers are in a unique position for giving youngsters the
opportunity of programming on the go: many high-schoolers don't have
netbooks, and smartphones aren't programmable directly on the smartphone...
And I'd say that the opportunity to program on the go is important for
hobbyists.
> > In the open development community, there would be much fewer
> > negative feelings about the Nspire Clickpad / Touchpad / CX (CAS),
> > if it had two modes of operation *used at their full potential*:
> > * a locked-down mode for tests - which the Nspire series has,
> > under the form of the PTT mode;
> > * an open mode for usage by enthusiasts, engineers, etc., with
> > official documentation and SDK (like the TI-Z80 and TI-68k
> > historically had for some time, even if it was significantly
> > incomplete) that helps people to use the full potential of the
> > hardware they have bought, and therefore own, in the ways they
> > see fit. That's what's missing on the Nspire series.
> > Reverse-engineering is a slow process, especially on such large
> > amounts of code, and a number of people interested in those
> > aspects have stuck to the TI-Z80 and TI-68k series (which haven't
> > been used to their full potential yet !) or moved on to other
> > platforms...
>
> Is it that TI product development managers are hostile to people
> hacking the device, or are they just ignorant of the desire of
> customers to do this?
Well, if they are ignorant of the desire of customers to do this, I
think that they're blind and incompetent: the versatility and potential
of fiddling their older series provided, participated to their success
in the marketplace (in some markets, at least) :)
But after a period where it was explicitly allowed by the management at
TI Education (it it wasn't, how to explain official ASM support for
years in multiple models of the TI-Z80 and TI-68k series ?), some people
in the management at TI Education _have_ grown hostile in the past few
years:
> I mean, is there evidence of hostility, or is it just an assumption.
Well, in which category would you put the DMCA takedown notices
(completely illegal ones, at that - the counter-notices available from
the links below show how TI severely misused the DMCA) sent to several
persons and web sites who published the factorization of the RSA public
signing keys for OS and FlashApps, used in TI-Z80 and TI-68k calculators
? ;-)
https://www.eff.org/press/archives/2009/10/13
https://www.eff.org/deeplinks/2009/10/texas-instruments-stop-digging-holes
Not only this, but from real-life discussions between several members of
the open development community on the one side, and people at high
levels of responsibility in TI on the other side, we also know that
there are some people, in TI Education, who consider usage of the
calculators for purposes not originally intended by TI as a threat.
Let me emphasize on the "some" people - we know that not everybody in
the management at TI Education follows such a line of thinking, and
fortunately so :)
> On some of the forums, I noticed a recent letter exchange with a
> TI insider and a calc. hacker, and it seemed like there was the
> possibility for constructive dialog.
Some constructive dialog seems to have restarted lately, after a period
where communication was performed through lawyers.
I'd like to think that it's a good thing because it's in everybody's
best interests, plain and simple :)
Perhaps some people have realized that TI Education wouldn't want to
copy, for the Nspire, the dangerous path that Sony followed for the PS3:
* from clamping down on the freedom to tinker (which added "hitting back
at Sony in the wallet" to the list of motivations for reverse-engineering)
* to legal attacks against those who destroyed the faulty protection as
a result of the clampdown on the freedom to tinker (whatever the
motivations were)
they ended up motivating enough anger that their business model was
attacked to the core, through the thorough penetration of the
PlayStation Network. Days of downtime, dozens of millions of dollars of
losses, as the result of bone-headedness of the management and legal
department. Good job Sony :)
That said, some people at TI still haven't realized a more fundamental
fact (because they're still trying - and miserably failing - with the
84+ Pocket.fr and the 1.03 boot code for the 84+, for instance): locking
down is ultimately futile. Sooner or later, the platform will be forced
open.
The history of computing is shock full of platforms that the
manufacturer tried - and royally failed - to lock down, and the list is
growing ever and ever...
> Reading some other stuff on the hacker forums made me worry that
> people were taking an immature attitude and whining that they
> couldn't have what they wanted, when it was unclear if they had
> ever just expressed politely their wishes.
Over more than 10 years, not everybody was fully polite, of course - but
people did express wishes to TI :)
For example, a list of bugs was created (
http://www.ibiblio.org/technicalc/buglist/bugs.pdf ), and TI _does_ know
about it (or at least, was contacted about its contents): some of those
bugs were fixed on TI-68k/AMS, and some more, as far as I could see, on
Nspire/Phoenix.
Let's also mention the analysis and explicit reporting, complete with
the bugfix, to TI-Cares, of an important bug in OS 2.53MP for the
84+(SE). This bug was not fixed in 2.55MP, and neither were a number of
other bugs (including data losses) that have been well analyzed - and
patched - by the community...
And I'm positive that TI has received a number of feature request in
math functions.
The persistence of unfixed severe bugs participated in TI and TI-Cares
earning the "TI-doesn't-care" moniker:
http://www.ticalc.org/archives/news/articles/4/46/46578.html
http://www.omnimaga.org/index.php?topic=6767.0
http://www.unitedti.org/forum/index.php?showtopic=9589
http://www.casiocalc.org/?showtopic=5504&st=160
Let's also mention the wishes have been indirectly posted over the
years, but had obviously lesser chance of being granted (giving that TI
isn't monitoring the community much, AFAWCT...), under the form of
third-parties reimplementing various bits of core OS functionality in a
much faster way, either embedded in ASM programs or patching the OS.
"Let's work around the limitations ourselves because they get in the way
of what we want to do", if you will.
For TI-68k calculators, let's mention four items:
* ~3x for "next_expression_index", the routine at the very core of the
CAS (one of the most widely called routines in the entire OS, and
therefore a critical hot spot for performance), without loss of
functionality IIRC;
* number to string conversion;
* on AMS 2.xx and 3.xx, ~12x faster (!) drawing characters and strings.
Part of that speedup can be attributed to loss of genericity over buffer
sizes, but the slowdown between AMS 1.xx and 2.xx can be entirely
attributed to the implementation not having a fast path for the
overwhelmingly common use case of fonts not being redefined;
* on AMS 2.xx and 3.xx, support for non-English languages slows down a
lot, even if no language localization is active. Again, this is because
there's no fast path for the common use case of English language. This
use case is common even in non-English-speaking countries, because BASIC
programs not written for English wreck a lot of havoc, so third parties
often recommend not to use the language localizations.
> If we attack TI and say "you are waging war against us, it's not
> fair, now let us do what we want!" The response is likely to be
> different and less constructive than if a different attitude were
> used.
Well, as far as the Nspire was concerned, in the mindset of many of us
in the community, the one side starting being non-constructive and
declaring war was TI ;)
At the time of OS 1.1, in the minds of a number of old-timer
TI-Z80/TI-68k users, individual buyers of their calculator(s), it was
like "Er, a TI calculator that not only provides no access to native
code, not only has obfuscated software upgrades, but doesn't even have
BASIC ?! WTF are they smoking ?? They're negating what made a part of
their success in the marketplace for more than 10 years. Seriously, WTH
are they trying to do ??".
And that's why Lua, which gets the programmability of the Nspire series
past the stage of "The 20-year-old TI-81 can do better in some areas",
is so welcome. Even if it occurs the eternity (in computer time) of 4
years after the introduction of the Nspire platform.
Oh yeah, this was a very long e-mail :D
Regards,
Lionel.
figureloop
> Hi,
[edit a lot]
Lionel, thanks for that. This is a very "educational" overview of the
history. I agree with most of your perspectives. Best wishes!