I do believe that having an actual emulator would be quite (if not extremely) useful.That would be great indeed but a lot of work too. :-\
As HP looks more open to the community, maybe they would give the interested people some documentation so that a linking program could be coded for Linux, mainly :)I think both Omnimaga and TI-Planet should be eligible as HP beta testers and be given some privileged technical info. Let's hope HP sniffers read these messages and are sensible enough to provide some more useful insight.
Nice! Can you show pics?
Guidebook, firmware and linking software did leak last week on the HP FTP server and were quickly removed:Come on, it can only have been done on purpose, IMHO!
http://tiplanet.org/forum/viewtopic.php?f=55&t=12867&lang=en
Did the "emulator" come with a reset button on the back?nah but there is a reset option in the menus. The thing, though, is that even if it actually worked like on the calc, it would be next to impossible to do the key combination and clicking that option simultaneously.
(the calc simulator software is more annoying to work with - we/I'm not sure exactly when it decides to save stuff...)
That's probably the goal of this topic, as stated by DJ in the first post ?Yeah, this time, i completely with you, weird he? ;) Sooner or later that info will be uncovered, anyway.
The hpwiki could contain such information as well.
Quite possible. Their time is more expensive than ours.Guidebook, firmware and linking software did leak last week on the HP FTP server and were quickly removed:Come on, it can only have been done on purpose, IMHO!
http://tiplanet.org/forum/viewtopic.php?f=55&t=12867&lang=en
BTW, has any you managed to get hold of the "HP-Connectivity Kit"? That would be really interesting. And also the booter code, of course!
Good news, nevertheless! ;)
Sometimes, I keep the save folder open on my computer in case I mess up. If I fear that the calc could crash, requiring a reformat (although so far things are fine) or if I worry about messing up my program, I just copy the hpprgm files into a different folder.(the calc simulator software is more annoying to work with - we/I'm not sure exactly when it decides to save stuff...)
On exit. Everything is only saved when either more memory is needed (and then it will write out loaded objects), or on power down.
Loading is also delayed until needed. Some of the mysterious bits you have trouble interpreting are flags indicating status, location in memory, or similar.
Question: What kind of data is stored into RAM? When a program is executed, for example, is it copied to a certain RAM area, forcing a size per program limit? Are programs always stored in RAM then backed up in Flash when the calc is turned OFF (to prevent them from being lost on a reset)? How large are each GROB memory areas?
OS SDK0.30 + Boot Code V13 + flasher usbtool.
Nothing mentioned about that little orange pcb though. I would love to have a picture of it removed from the calc if possible.Or a higher resolution picture of the motherboard only would also help, I'd say!
Edit: I've just removed the direct link. Tell me if you change your mind. I know it's a difficult decision, especially for this topic.Now we get a feel on how touchy this topic might become in the short term. Maybe the Omni administration could provide some help here, before it's too late and topic becomes locked? :(
The firmware zip file which was available at the time of the leak was not password protected, but it was a different version: SDK0.26 + Boot Code V11.
Edit: I've just removed the direct link. Tell me if you change your mind. I know it's a difficult decision, especially for this topic.
Is this really what I am thinking? O.O
How do you run the BXCBOOT0 bootloader? Judging from the text found in BXCBOOT0.DAT the calculator checks for an SD card. Could this mean the bootloader can load an operating system from an SD card?
EDIT: as far as emulating the Prime is concerned, I keep thinking that we should definitely base an emulator of ours on an existing, versatile, full-featured, GPL'ed framework, instead of reinventing a new emulator from scratch.
Hi, as you might know, HP is in the process of releasing a new graphing calculator, dubbed "Prime", based on a S3C2416 ARM chip, but not running U-Boot + Linux. The hardware is more powerful, and the software is less closed-minded, than those of the TI-Nspire, which is based on undocumented chips.learnt me that:
For the purpose of making a portable community emulator for this platform, QEMU seems an obvious choice, especially since it's easy to find (very old...) forks of QEMU with support for the pretty similar S3C2410 & S3C2440 chips. Writing everything from scratch (which was done for the Nspire emulator) seems a poor way to use our free time.
I'm writing here because I'm not clear about, for instance, 1) the making of a portable UI with clickable buttons interacting with the emulation core, 2) the bootup of something based on neither U-Boot nor Linux, and 3) the creation of memory images / savestates suitable for QEMU.
I've already used QEMU for installing standard Linux distros on emulated x86, x86_64, ARM, PowerPC, Sparc and MIPS - but for those, either there's a BIOS which can consume ISO images, or the user points a Linux kernel and an ISO image, and no need for displaying a graphic, clickable keyboard.
So... could you explain me / point me to the right, up to date, information ? TIA :)
I think most people who have an HP Prime are on HP Museum forums it seems. It's hard to attract HP fans on a TI forum when there is an established HP-only board around, same problem we had with Casio people. Certain people might see TI boards as the "enemy" or something.
But I'll tell you that the Casio ClassPad II fx-CP400 is currently a bad thing (only the main (calculator) application can be rotated, Basic is far slower than on the TI-81/80 from the last centuryIn CP400 defense, however, the line command is much faster than on the PRIZM (although still slower than on the 84+CSE). :P
Well, don't forget the fact that the "HP Prime" not having been released worldwide so far, does clearly have a negative impact on people's enthusiasm. Hacking does require a lot of testing with the real thing. So, i'd say as soon as it becomes available and we can play with the real toy, the hacking interest will certainly increase all of a sudden. So, it's also HP's fault :P. My 2 cents.Yeah I agree. Only Critor and Gilles59 had the calc until recently and it was not necessarily the final release. Kerm only got his sample recently and I don't have one (although I already pre-ordered the calc anyway). The emulator is more like a simulator, so ASM programs most likely won't run on it, and BASIC game speed is innacurate on the emu. Missing the bts rush didn't help either, since most people already got their calcs by now.
Yeah I mean like that one debate topic there about why should there be a Wiki on TI-Planet when there is already wiki4hp that have been around for years, located at http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/forum.cgi?read=249854#249854 . One person in particular wasn't fond of the idea that HP Prime info and fans go on a TI website instead of their own and only HP forum. They probably don't want any other website to budge on their thunder.
If you're sure you don't have a "normal" email account, or if you use an ISP that has the unfortunate tendency to provide both free and paid addresses, then please click here to register by postal mail.They could at least have an email link to contact administration so you could be added manually.
*edit* Just checked out their forums, To say the least, they are really dated. I can't even register unfortunately. I only have hotmail. I don't have access to a non-free email account. Their alternative is a bit ridiculous unfortunately. :/Yeah it's pretty ridiculous.Quote from: The Museum of HP CalculatorsIf you're sure you don't have a "normal" email account, or if you use an ISP that has the unfortunate tendency to provide both free and paid addresses, then please click here to register by postal mail.They could at least have an email link to contact administration so you could be added manually.
I'm not sure how that plan to host a wiki there when many people who might be interested won't even be able to access their forums. :/
Tim Wessman hardly ever post on Omni anymore for unknown reasons.
Tim has to try to avoid participating to HP hacking discussions, since he is working for HP.
I think most people who have an HP Prime are on HP Museum forums it seems. It's hard to attract HP fans on a TI forum when there is an established HP-only board around, same problem we had with Casio people. Certain people might see TI boards as the "enemy" or something.
Then they have some maturity problem.
thanks to the (ab)use of the standard HID class, no additional driver is required.
I am certain a large majority see someone posting "cute" little simple programs that don't really do much doesn't interest a majority of the readers/posters.Indeed.
Were you aware of the wp34s which is a complete from scratch calculator based on a repurposed HP-20/30b? http://commerce.hpcalc.org/34s.php Or how about the user that created a USB connection for said calculator complete with rechargeable li-ion battery that mounts inside the unit? Or the user who created from scratch a brand new replacement board for the HP-41, including every flash module ever sold for it, extensions to the built in system, full compatibility with all HP-IL modules, and so on... http://systemyde.com/hp41/index.html?I knew about the link between WP34S and 30, but not about most of the other, pretty interesting, items that you're mentioning :)
Then you have the team that recreated a mini HP11,12,15,16C calculators and sells them? http://www.swissmicros.com/ The team that is making a complete from scratch hardware calculator to rival anything HP or TI ever made (of the classical style)?
To many folks like this, hacking software to make games, or even just programming on modern units isn't all that exciting.Easily understood :)
Not sure until the protocol has been analyzed :)Quotethanks to the (ab)use of the standard HID class, no additional driver is required.Abuse? In what way?
Here's just a raw dump of my Prime's data:heya and welcome to the forums. :) Nice to see that HHC2013 is happening by the way. IIRC there were worries about a cancellation due to the recent Colorado floods.
DATA: V0.025.5106 OS(1M):V0.025.5106
ADPT IN : BOOT ARMUboot V13
Samsung K9K2G08U0M 256 MB SLC
So I see you've got the SDK version working! Have you found any differences in function yet?
EDIT: The HHC2013 is going on as we speak. I'm hoping some of our questions about native execution and the like will be answered there...
EDIT 2: I've been messing around with the file formats. The first byte of the custom app's .hpapp files seem to indicate what app it's copying.
heya and welcome to the forums. :) Nice to see that HHC2013 is happening by the way. IIRC there were worries about a cancellation due to the recent Colorado floods.
Also is the calc really sold with OS 0.025.5106? A version number under 1.0.0 often means it's a beta... ???
Where do we usually download released firmwares for HP calculators ?Usually, I'd say somewhere like http://h20000.www2.hp.com/bizsupport/TechSupport/ProductList.jsp?lang=en&cc=us&prodTypeId=215348&prodSeriesId=5367459&taskId=135 (http://h20000.www2.hp.com/bizsupport/TechSupport/ProductList.jsp?lang=en&cc=us&prodTypeId=215348&prodSeriesId=5367459&taskId=135), but it seems both the emulator and the data-link program have firmware-downloading capabilities; however, it seems HP's data server are not working.
Also is the calc really sold with OS 0.025.5106? A version number under 1.0.0 often means it's a beta... ???
some features feel incomplete
Nope. That is a common convention, but really has nothing to do with the state of things...
Examples please?
Ah I see Tim, I worried that the OS was still in beta lol, but if that was the case, I would have been surprised at how stable it is per beta standards (at least, from what I saw in the emu). I hope it's quite stable for myself in its final form too :P.QuoteAlso is the calc really sold with OS 0.025.5106? A version number under 1.0.0 often means it's a beta... ???
Nope. That is a common convention, but really has nothing to do with the state of things. That whole menu is for factory use and they control numbers and versions there. I honestly have no idea what they really are using there or what it is supposed to mean. I suspect it has something to do with bootloader stuff.
some features feel incomplete
Examples please?
QuoteAlso is the calc really sold with OS 0.025.5106? A version number under 1.0.0 often means it's a beta... ???
Nope. That is a common convention, but really has nothing to do with the state of things. That whole menu is for factory use and they control numbers and versions there. I honestly have no idea what they really are using there or what it is supposed to mean. I suspect it has something to do with bootloader stuff.some features feel incomplete
Examples please?
Also, there's no way to select the item on drop-down input boxes without touching the screen. I'd think you could hit ENTER to open up the dropdown, but no.
Also, what is the user function supposed to be used for? I tried looking in the manual and couldn't find anything. It's under shift+help.
Also, what is the user function supposed to be used for? I tried looking in the manual and couldn't find anything. It's under shift+help.
It redefines keys. You make programs with the KEY function instead of the EXPORT function, and the function's name is K_ and then the key's name (numbers are themselves, but more complex key names use This Case). You then press User and then the key you redefined to execute the program. I know; at first, I was used to the 50g notion of adding functions to a userkey list and thus thought the key was useless; as it turns out, there's a different method for userkeys on the Prime.
Is this comparable to anything that TI has?
I've reported both problems to HP-France, but I don't know if they did translate and forward them.
As a general comment here, thanks for the details on things.
When you find a spelling/grammatical issue with the built in help, please email me and I'll be sure it gets resolved in the next revision.
After trying to type the following integral in, I got the weirdest, most undescriptive error that I've ever seen on a calculator. Trying to copy/paste yielded the same result, yet retyping the function gave the correct result.
(http://i.imgur.com/f7WpvUt.png)
Which is not surprising, IMHO. Big corporate companies tend to do that a lot. Blame bureaucracy for most of those occurences. Hence why everyone nicknames TI-Cares support service as TI-Doesnt-Care. The best that can happen is if one TI/Casio/HP employee who is involved in programming the calculator OS is active on forums where the product he worked on is popular, which is the case right now. Otherwise, there is usually some sort of person who filters e-mails and might just delete anything that is slightly negative or just not bother about sending the info to the appropriate team, which is made even worse if it's transmitted from HP France to another HP headquarter or vice-versa.I've reported both problems to HP-France, but I don't know if they did translate and forward them.
Definitely never made it to me...
You know, it does kinda feel like a beta. The documentation isn't very good, some features feel incomplete, you know, the lot. It really feels that they rushed to ship this out. And yet they're still late...Thank you for all your feedback this far, it will certainly be much appreciated by all future Prime users. I think, however, that most of these recent posts hardly have anything todo with the real hacking stuff. :-\
By the way, a Bug reports thread has been made here: http://ourl.ca/19642/361648;topicseen#newWell done!
Sorry to make so many threads, but I think this one deserves its own thread. This was partially mentioned in the EDITMAT thread.
You can get likely your calculator to hard crash and corrupt its memory in the following manner:
1. Create a new app and within the app have new bindings for the default views such as Plot, PlotSetup, etc. Your app should presumably creates new global variables.
2. Run the app and make sure everything is working.
3. Open the program editor with your app currently running. Simply open the source code to your app and close it.
4. Now, press a key corresponding to what your app overrides. For example, if your app has a new Plot() function, press plot.
At this point there is a good chance that your calculator resets, and possibly corrupts memory.
You can use my Graph 3D app, which I have slightly updated in a different thread. Here are the exact steps:
1. Follow the steps on how to install the app. Click here for the source. (http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/forum.cgi?read=251367)
2. Run the app and plot 1/390*(X^3*Y-X*Y^3) using the default window settings. After you see a graph, press [Num] to see a table of Z-values. (This is just to simulate having done stuff on the app.)
3. Now, with Graph 3D as the current app, go to the program editor (press [SHIFT][1]) and open the source to the Graph 3D app. Simply open and then close it with [ESC]. I have sometimes experienced a warmstart at this point.
4. Immediately try to plot using the [Plot] button and the calc will hang and eventually warmstart. Or try pressing [Num]. If it doesn't hang try turning it off... and then if can look at the matrices via [SHIFT][4]. M0 is a copy of Zvals, and I often get a huge matrix that is less than 1 kB in size!
What I think is happening:
After exiting the program editor, the calc tries to compile the source into an executable. The problem, however, is that when it's the currently running app, then function pointers are probably slightly changed and what was once a valid pointer is possibly no longer valid.
Current workaround:
Any time you close the program editor (after editing the currently running app), restart the app (exit to a different app and start your app again).
I don't know if this is exploitable, but there seems to be some reproducible steps to corrput the memory :
the calculator is not running Windows CE.The world is saved. :D
Just out of curiosity : are you using a base RTOS or is the software written from scratch ?Well, almost nothing is written from scratch nowadays even when they claim so. ;)
The kernel is written using GNU C and the GNU toolchain. While itSource : https://github.com/torvalds/linux/blob/master/Documentation/HOWTO
adheres to the ISO C89 standard, it uses a number of extensions that are
not featured in the standard. The kernel is a freestanding C
environment, with no reliance on the standard C library, so some
portions of the C standard are not supported. Arbitrary long long
divisions and floating point are not allowed. It can sometimes be
difficult to understand the assumptions the kernel has on the toolchain
and the extensions that it uses, and unfortunately there is no
definitive reference for them. Please check the gcc info pages (`info
gcc`) for some information on them.
Well, this says otherwise : ...Well, before Linux there was MINIX: http://en.wikipedia.org/wiki/Minix
Does anyone have any clue where that idea came from? I've been unable to track it down myself after some time looking.Yeah I was wondering that too. I didn't hear about the possibility of it using Windows CE until today.
Linux was not based on MINIX's code. I have heard of it.The Linux kernel has indeed been written from scratch, based on theories learnt from Minix.
That is interesting. Regarding such OS change, however, would this affect compatibility with existing HP PPL programs?
You use Home variables to store
data that is important to you, such as matrices, lists, real
numbers, etc.
...This seems to me as a very reasonable guess, indeed. They (HP) wouldn't spend time and money reinventing the wheel (as someone already has put it). If this guess happens to be true, then we should be able to get it replaced with our own 'custom' bootloader. That would a major step towards starting the hacking process. Unfortunately(?) this requires somebody with some knowledge and above all with some spare time to dig into it. Let's hope more hackers do arrive at omnimaga! ;)
It looks like the bootloader is a general purpose one that is capable of booting both PE and ELF format binaries from at least yaffs2 and fat filesystems. It seems to have support for the lcd, usb and serial.
I wouldn't be surprised if it provided an interface via the uart - but am not going to try myself since I need to resolve a minor hardware defecrt with hp first.
We wouldn't need to replace the bootloader if it takes standard ELFs. Just load linux with it.Yeah, maybe so?! I'm just playing on the safe side. We never know what's hidden inside a bootloader until we try to replace it.
The issue though is that if the calculator gets popular next school year when it comes out in brick and mortar stores (if ever), then many people who get or make programs will not ever know that there is a modified OS that no longer supports variables the way the official OS did and then share their programs. At the same time, some hardcore programmers who use the OS will release their own programs, flooding the web with programs compatible with only one OS at once. It doesn't matter if there aren't many HP PPL programs right now, as there could be many in the future. Basically, people will have to constantly install back and forth the unnoficial OS and official one in order to run everything they want.QuoteThat is interesting. Regarding such OS change, however, would this affect compatibility with existing HP PPL programs?
As there hardly are any the more important question first: "would this affect compatibility with programs preinstalled?"
Except for the Graphing variables and the settings variables I don't see why any are predefined. As far as I can tell they serve no purpose whatsoever. This is what the Manual has to say:QuoteYou use Home variables to store
data that is important to you, such as matrices, lists, real
numbers, etc.
I don't get why I would want to do that.
Would it affect PPL programs? Yes. If the program used the variables while assuming they are of their respective types. For example an assignment L1:=1 now results in a list containing one element: L1={1}. This behaviour would change. If the type of L1 was initially undefined the same assignment would result in a variable of type real (I think) L1=1
Imho this is the more often expected behaviour.
Of course renaming all those home variables to chinese characters would make them rather useless (which imho they already are). So if someone from HP is reading this: How about instead add a button to create new variables in the VARS view? So that I can define my own variables "that are important to me"?
Also: Why do only variables created in the home view appear in the VARS view (some don't - something about that dialogue popping up), while some can only be found in the mem view?
I am personally making large programs for the HP Prime right now and I definitively don't want to have to re-write everything in one year when custom OSes start coming out. I already have one game out, HP Prime Tunnel, that uses L1 and A-Z variables. There are already around 20 programs on Hpcalc.org.
I think it would be best if there was some sort of emulator released that changes the behavior of the HP PPL interpreter or compiler or at least a program allowing you to dual boot. Even better: An OS that supports both syntaxes/type of vars.
is it possible to create permanent variables and lists that won't get deleted on program exit?
In a compiled language like C, all this work is done before the program ever launches
HP: you've stolen my variables. By restricting where I can use lower or upper case variables, you've DESTROYED my work-flow. Equations require a combination of upper and lower case variables... what were you thinking?
Wow, if PPL is interpreted, then I guess that might explain why it's so fast lol.
Advanced users have no problems with "i am in charge of creating a variable". However, we have very strong feedback from lower end users and educators that having a set of predefined variables is very important and useful to them for many reasons.
Prime in its first release is not targeted directly at high end users. What was and wasn't in the first release was not targeted at the top 1% of users. If that messes up "your workflow" as was posted in that thread, and even the fact that someone is posting about calculators, pretty much by definition means they fall in that tiny power user minority.
8+3*i=:A Error:Bad argument typeWhy the hell can't I assign?
8+3*i=:L1 {8+3*i}Ah great now it works. I wonder why. And where do those curly brackets come from? There was something about different variables in the manual. I guess I'll store that in one of the complexes.
8+3*i=:C1 Error:Invalid input:banghead:
It makes sense to have a system set of variables that can be transferred from program to program, but it would also be nice to have local-program variables. It seems like having both should be an option.
...8+3*i=:C1 Error:Invalid input
At least I can take comfort in knowing that I actually got heard out by someone that could do something about it. I actually do - thanks for that.
Heya and welcome to the forums! :)
Unfortunately, as far as HP Prime hacking goes, it's pretty much dead. Nobody will ever hack the HP Prime because nobody seems interested in doing it for a platform that has less than a million users in the world :(
Even TI-Nspire hacking is dead, although there are a few games still coming out every now and then. No more Ndless development nor anything else, and that's with a much more popular (although more closed) platform. It took three years until ASM arrives on the Nspire too. Same for the Casio PRIZM, but again, hacking it is no longer as needed, since Casio has done nothing to block third-party ASM/C code.
So if you're reading that because goole lead you here: Try the simulator and make up your own mind.
...Well, IMHO it mostly depends on your luck. One of these days a prestigious hacker knocks at the Omninaga's door offering his/her services for free. Until then, keep your hopes as high as possible. Yeah, on this matter I agree with Jim why should we be so negative? After all there are great (TI-Nspire) contributions on this site and some and still rather recent! Come on DJ_O cheer up a bit, brighter days are coming! :)
-Have enough time and motivation
-Has interest in the platform
-Has the knowledge to hack it. ExtendeD had 9 years of ASM experience when he got involved into Ndless.
...
Well, IMHO it mostly depends on your luck. One of these days a prestigious hacker knocks at the Omninaga's door offering his/her services for free. Until then, keep your hopes as high as possible. Yeah, on this matter I agree with Jim why should we be so negative? After all there are great (TI-Nspire) contributions on this site and some and still rather recent! Come on DJ_O cheer up a bit, brighter days are coming!Reminds me of nLaunch (CX).. but that was probably only because of TI's madness.
Heya and welcome to the forums! :)
Unfortunately, as far as HP Prime hacking goes, it's pretty much dead. Nobody will ever hack the HP Prime because nobody seems interested in doing it for a platform that has less than a million users in the world :(
Even TI-Nspire hacking is dead, although there are a few games still coming out every now and then. No more Ndless development nor anything else, and that's with a much more popular (although more closed) platform. It took three years until ASM arrives on the Nspire too. Same for the Casio PRIZM, but again, hacking it is no longer as needed, since Casio has done nothing to block third-party ASM/C code.
For the next few years, we'll have to stick to HP PPL language (although that's already quite fast enough for the time being, since it's compiled).
On a better note, apparently it was discovered not so long ago that the HP Prime pretty much lacks any viable protection against third-party OSes. In other words, someone could disassemble the HP Prime OS then modify it to add a LaunchASM() command that runs ARM assembly or C programs. He could even write his own OS or make such mod avaialble as IPS patch to avoid getting in trouble by releasing an OS containing copyrighted content.
Besides that, if you or anyone else knows enough ARM assembly, hacking and stuff, they are well welcome to help finding other alternatives. Of course, keep in mind that in the future, HP might release OS updates that could unintentionally (or if they're like TI, intentionally) block ASM exploits.
It breaks file transfer between the emulator and calc,
Well I bought the Prime because I was a huge fan of my two 48 that I used to program in asm (I loved Saturn asm, Jazz was so cool). But I lost them some years ago when a fire started in my flat. I don't have the real machine yet (should be here in the next few days), but I played a while with the simulator. At first, I must say I was a little disappointed. A lot of cool feature of the 48 are gone, RPN is not half as practical as it was, and RPL miss me a lot. The filesystem (sort of) that used to manage objects and variables too. It was just excellent.
Ouch, that sucks about the fire. >.<
And yeah, the reason why a lot of 48 features are gone is because the HP Prime was aimed at a completely different userbase than the older models. Instead of engineers/professionals, it was aimed towards students for the most part, and in the past, I often heard younger students complain about how confusing and hard RPN/RPL are, so forcing RPN/SysRPL upon them with no option to use algebraic mode and a Basic/Pascal-like language would have turned them away. The same thing happened with the 83+/84+ when FastRPL language came out: It showed big promises for programmers who wanted to program fast games without having to learn ASM, but to know FastRPL, you needed some ASM knowledge (the stack, for example). Now look at how popular Axe Parser is.
If you liked the HP48 and want a faster machine with a larger display and much more expandability, why not consider an HP50G? You can program in RPL, SystemRPL, Saturn Assembly, ARM, or C (via HPGCC). As for Jazz, yeah it was very cool -- and still is. I updated it to the HP50G and it is much more robust now.
http://users.ju.edu/hduong/jazz
The 50g OS adapted for the screen (res and colors), installable as an app on the Prime, and a good C/asm dk for the ARM, you have the ultimatest pro/students/hackers calc ever :pUltimatest for hackers for sure but for students i doubt it. Schools would never embark on such an adventurous trip.
.. Let's hope they have a Prime now ^^I wonder, are they still working for HP? ;)
Quote.. Let's hope they have a Prime now ^^I wonder, are they still working for HP? ;)
@Han: wow thanks for such a comprehensive answer!
That would be nice indeed, but still, the problem with the 50g is that the OS is basically a Saturn CPU emulator on which an improved HP 48 OS is running. Even with a 200+ MHz ARM processor, the fact the OS is running on a Saturn emulation layer makes it way slower than the HP 39gII and Prime OSes.Actually stock, it runs 75MHz. You need to run another program to "uncap" it's speed.
I wish that HP actually ported the 50g OS to ARM ASM rather than just sticking to Saturn ASM and using an emulator, the OS wouyld have ran much faster.
The HP Prime app idea is nice by the way.
Oh right, I forgot about the app. And yeah having rewritten the old OS would have been a great idea (perhaps better, IMHO, providing that they also include HP PPL and make the calc more user-friendly for younger audiences). It's just a shame that this wasn't done with the 50g, since a lot of this platform hardware potential (regardless of it it runs at 75 or 200 MHz) is being wasted, as well as battery power, via slow emulation.Thanks, just decided to finally join since it seems that the Prime has generated interest in HP calcs. I know that Casios were kinda popular, but the fx-cg10 just didn't really grab my attention, and the fx-cp400, well all I can say is: Are they SERIOUS?! (This is a POS @ $140, maybe $60 it MIGHT compete w/a 50g which you can get for $80 on Amazon nowadays.)
Welcome to the forums by the way! :)
...Hmm, you just might be right!? If you don't know it yet, please have a look here Maxima On Android (http://ourl.ca/18114) for a little more.
...was also thinking, a pretty much complete maxima install on Android is ~100MB, and I hate to say it, but it's probably generally a better CAS than xCAS(IIRC that's the real name of the 49/50/Prime CAS... I think that Cyril de Brebbison is still working with HP at least for the CAS and some other parts of the "apps"...)
Yeah as a long time TI fan, I was still happy to see people here take interest in other brand products and I myself took interest in them in the past as well, especially the Prime. As for the cg10, my main issue is the slow Basic language and I don't want to learn Lua to make faster stuff when I can already do it in Basic/PPL on the 84+/Prime. The FX-cp400 price tag is inexcusable considering even maths are slow. The hardware might be better, but the software is just inexplicably slow. Someone needs to teach Casio optimizing tricks.Oh right, I forgot about the app. And yeah having rewritten the old OS would have been a great idea (perhaps better, IMHO, providing that they also include HP PPL and make the calc more user-friendly for younger audiences). It's just a shame that this wasn't done with the 50g, since a lot of this platform hardware potential (regardless of it it runs at 75 or 200 MHz) is being wasted, as well as battery power, via slow emulation.Thanks, just decided to finally join since it seems that the Prime has generated interest in HP calcs. I know that Casios were kinda popular, but the fx-cg10 just didn't really grab my attention, and the fx-cp400, well all I can say is: Are they SERIOUS?! (This is a POS @ $140, maybe $60 it MIGHT compete w/a 50g which you can get for $80 on Amazon nowadays.)
Welcome to the forums by the way! :)
...was also thinking, a pretty much complete maxima install on Android is ~100MB, and I hate to say it, but it's probably generally a better CAS than xCAS(IIRC that's the real name of the 49/50/Prime CAS... I think that Cyril de Brebbison is still working with HP at least for the CAS and some other parts of the "apps"...)
...I dare to suggest they try to hire you! :P
As for HP, they just need to learn marketing better, so their calcs come out after most bad bugs get fixed, not before, and learn to promote their product outside the internet.
...
.., unless you mean for marketing,...^Exactly(!), that's what I mean. ;)
...Yeeeah, easier said than done. We still lack the person to do the hard work, though. What a shame. :P
I like the idea about cross-compatible programs between the Nspire and the Prime, though. O.O
Thanks to bb010g for posting this link on IRC http://ried.cl/en/franqueando-los-secretos-del-hardware-mediante-ingenieria-inversa/
Maybe that could be useful for Prime hacking?
The main issue, though, is that making a new OS is incredibly hard. Even on the 83+, after 15 years there are still no third-party OS other than games and unfinished OSes with limited functionality. http://www.ticalc.org/pub/83plus/os/
Thanks to bb010g for posting this link on IRC http://ried.cl/en/franqueando-los-secretos-del-hardware-mediante-ingenieria-inversa/
Maybe that could be useful for Prime hacking?
WOW. This could be the breakthrough. For one, we could watch the USB port when, say, a new firmware is being installed...
And this article also seems to imply that HP ignores CRCs when transferring. If they ignore checking with firmware, our job could be a lot simpler.
I probably won't try to hack it since I am not into low level and hardware stuff but if I find anything interesting I'll share it.I'm working on it...school & finals...grumble grumble grumble
Also if you get a prime I hope you do code a bit for it too, unlike certain other Omni users :P
Hello,Oh great !
Yes, from HP.
>Perhaps the uart is only used for flashing the bootloader for example and not for debugging.
nope, bootloader, when we need to flash it (which has never happend to me) is done through JTAG...
I can't help you hack Prime application, sorry, but I can help you do prety much anything else...
As you stated earlier, the application is a .elf file and you can replace the application by anything else that you compile and it should be started by the OS... then you are free to do whatever you want... but this transforms prime into a single purpose machine...
cyrille
Perhaps it's a place for us to put our own software on :3 (it's a stretch)Or maybe a swap partition?
Or maybe a swap partition?I doubt it. The HP Prime's RAM is around 32mb mDDR and I doubt that they would make a swap partition that small if anything, but I could be wrong.
Could be used internally for developing calculator software? Since it's accessible from recovery mode. Just throwing ideas out there.If so, it could be useful for hooking Linux in.
When the HP Calculator is set to recovery mode, there is a partition available named "HP CALC DEV"(around 247.12mb is still free)
I loaded the .BIN file from the old SDK OS into a ARM dissassembler (armU, to be exact), and I found that I can emulate it's basic operations from there. I discovered what memory addresses correspond to the splash screen bootup and the main input loop.Interesting, let us know :). What I still wonder is if it's possible to run native code directly from the official OSes or if it really has to be done via an hacked/modified OS. It would also be interesting to see if the community would be able to help HP fix some of the OS bugs.
I'll post more specifics later.
I didn't think about that idea before actually. I am unsure if it would work, but if it did, then it would be kinda the opposite of Z80 calcs, in the way that instead of running a shell from an OS, you would run the OS from a shell, right?Exactly.
The thing with modifying the OS directly is that it would be more work for the user to patch it and there would be high risks of bricking their calcs if they use the wrong IPS patch or something. Plus would IPS Patches even work with such file format anyway? And to find an exploit inside the OS itself it will be much more work, not to mention it might end up getting patched by HP if it was the result of a glitch. So your idea could be something worth investigating.You can't really brick the calc as long as the recovery firmware works afaict.
My idea is still much easier to implement and won't require us to update the patches for each new HP OS. As long as we don't mess with exam mode we shouldn't have to fear HP.The bad thing about a shell is that you can't see whether it's active. For ndless, you have a way to remove it certainly (filesystem format) but is there a similiar way on the HP?