187 replies to this topic

[Nightrook]

Can anyone get me a disk image of latest firmware for Asiair pro? I am going to try something.

[billndotnet]

Can anyone get me a disk image of latest firmware for Asiair pro? I am going to try something.

If it's what I think you're going to try, it  "probably" won't activate, since the cpuid for a Seestar "shouldn't" be in the same registry as the ASIair valid cpuids. But given some of the lazy things I've seen, it's possible it might.

The mount interface seems similar to the am3/am5, if not directly derived, it's might actually work.

How are you going to image it?

 

[Taz33]

Can someone who has unlocked and upgraded to Firmware 2.42 (or later) make a dd image of the root partition?  I want to write it to my locked Seestar that is on 2.42.    I have one for firmware 2.30, but not sure if there would be any issues between it and the pi user partition that is on the later firmware.

 

Thanks.

 

Hi Euclid

(+ all other that requested this to me)

 

I'm trying to perform a dump of the Seestar, but it doen't seem to be working..

 

Thanks to the script found in this Github repo: https://github.com/a...in/canOPENER.sh which inspred me.

So praises to the apolloscharge user on GitHub (+ loved the name "canOPENER.sh" for his script !)

 

I switched my device into "maskrom" mode as described in the script to allow USB connection to my computer: here's what I had in the dmesg

[811515.699821] usb 1-1.3: new high-speed USB device number 20 using xhci_hcd
[811516.032539] usb 1-1.3: New USB device found, idVendor=2207, idProduct=110b, bcdDevice= 2.23
[811516.032556] usb 1-1.3: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[811516.032564] usb 1-1.3: Product: USB download gadget
[811516.032570] usb 1-1.3: Manufacturer: Rockchip
[811516.032575] usb 1-1.3: SerialNumber: XXXXXXXXXXXXXXX

I then used the rkdeveloptools from Rockchip GitHub to dump the data partitions from the Seestar into a file.

 

Here's an output of what I've done:

#rkdeveloptool rfi
Flash Info:
Manufacturer: SAMSUNG, value=00
Flash Size: 59648 MB
Flash Size: 122159104 Sectors
Block Size: 512 KB
Page Size: 2 KB
ECC Bits: 0
Access Time: 40
Flash CS: Flash<0>

#rkdeveloptool rcb
Capability:15 00 00 00 00 00 00 00
Direct LBA: enabled
First 4m Access: enabled
Read Com Log: enabled
Read Secure Mode: enabled

#rkdeveloptool ppt
**********Partition Info(GPT)**********
NO  LBA       Name
00  00004000  uboot
01  00006000  misc
02  00008000  boot
03  00028000  recovery
04  00038000  pi
05  00138000  rootfs
06  00B38000  swap
07  00F38000  seestar

#rkdeveloptool rl 0 122159104 Seestar_dump.img
Read LBA to file (100%)

#fdisk -l Seestar_dump.img
The backup GPT table is corrupt, but the primary appears OK, so that will be used.
Disk Seestar_dump.img: 58.25 GiB, 62545461248 bytes, 122159104 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disklabel type: gpt
Disk identifier: 73987B6B-4974-4C94-A3E8-58AB2EB7A946
Device               Start       End   Sectors  Size Type
Seestar_dump.img1    16384     24575      8192    4M unknown
Seestar_dump.img2    24576     32767      8192    4M unknown
Seestar_dump.img3    32768    163839    131072   64M unknown
Seestar_dump.img4   163840    229375     65536   32M unknown
Seestar_dump.img5   229376   1277951   1048576  512M unknown
Seestar_dump.img6  1277952  11763711  10485760    5G unknown
Seestar_dump.img7 11763712  15958015   4194304    2G unknown
Seestar_dump.img8 15958016 122159040 106201025 50.6G unknown

I have a 62Gb image dump as expected but it doen't seem to be "viable" when using kpartx to mount filesystems from the dump file, they just don't work + when gzipping the dump file (as I was hoping to share it with you guys as requested), it only wheights 73Mb which is suspicious... there must be missing data in there !

 

Although kpartx can see expected FS, they don't mount from loop devices created from them:

#kpartx -l Seestar_dump.img
Alternate GPT is invalid, using primary GPT.
loop0p1 : 0 8192 /dev/loop0 16384
loop0p2 : 0 8192 /dev/loop0 24576
loop0p3 : 0 131072 /dev/loop0 32768
loop0p4 : 0 65536 /dev/loop0 163840
loop0p5 : 0 1048576 /dev/loop0 229376
loop0p6 : 0 10485760 /dev/loop0 1277952
loop0p7 : 0 4194304 /dev/loop0 11763712
loop0p8 : 0 106201025 /dev/loop0 15958016

=> I can't see what I've done wrong here... any ideas ?

Edited by Taz33, 04 August 2024 - 12:08 PM.

[mconsidine]

I think back in this thread there was a mention that there might be a flag on the boot sectors that generate all 0s when it's attempted to be read.  I think this is a pi/rockchip/whatever feature.  "LBA flag" rings a bell, but I'm not near my computer.

[Nightrook]

Managed to crack ASIAIR and install it on a vanilla raspberry PI 4B.
 

 

 

You can see that SN 01F43F82 is not present on their website (https://asiair.en.zwo.biz/#/license) as a valid Serial.

Of course the power related things don't work.

 

 

I don't have ZWO hardware yet, so I don't know if it really works for real. Currently trying to make it work with non-zwo hardware.

Edited by Nightrook, 09 August 2024 - 01:59 PM.

[AstroPotamus]

This is a screenshot without context.  You mention the serial number but don't show it in the screenshot.  I'm not saying you didn't do what you said you did, but neither is this proof that you did.

[Nightrook]

This is a screenshot without context.  You mention the serial number but don't show it in the screenshot.  I'm not saying you didn't do what you said you did, but neither is this proof that you did.

SN is found on the top right corner of first image.
Disk image will be released as soon as I am sure it works.

[AstroPotamus]

Woah. I'm so sorry!  I only saw the second image when I was looking at it - the first one didn't load for some reason.  So sorry!!  Good job!! 

[Nightrook]

Woah. I'm so sorry!  I only saw the second image when I was looking at it - the first one didn't load for some reason.  So sorry!!  Good job!! 

No problem 

I am thinking of releasing the disk image so that you peeps can determine whether it works.

Wonder how fast ZWO will patch the loophole 

[billndotnet]

Managed to crack ASIAIR and install it on a vanilla raspberry PI 4B.
 

 

 

 

You can see that SN 01F43F82 is not present on their website (https://asiair.en.zwo.biz/#/license) as a valid Serial.

Of course the power related things don't work.

 

 

 

I don't have ZWO hardware yet, so I don't know if it really works for real. Currently trying to make it work with non-zwo hardware.

I can save you some angst here.

The imager binary is the core piece that the tablet app talks to, it does all the work. It loads libgphoto2 to support non-ZWO DSLR cameras, and then proprietary drivers for ZWO cameras. That's it. It invokes a modified INDI server that redirects both camera and focuser calls to ports on the imager binary, so side loading INDI drivers for non-ZWO cameras simply doesn't work. Replacing it with a main branch INDI server will break the workflow. The version of indiserver code that ZWO posted to their public git repo is a malicious-compliance grade joke, it's missing some essentials to get it built, and I don't believe it matches what they deploy.

I've got some detailed teardowns over on the indilib forums if you want more details about the internals.

[Nightrook]

I can save you some angst here.

The imager binary is the core piece that the tablet app talks to, it does all the work. It loads libgphoto2 to support non-ZWO DSLR cameras, and then proprietary drivers for ZWO cameras. That's it. It invokes a modified INDI server that redirects both camera and focuser calls to ports on the imager binary, so side loading INDI drivers for non-ZWO cameras simply doesn't work. Replacing it with a main branch INDI server will break the workflow. The version of indiserver code that ZWO posted to their public git repo is a malicious-compliance grade joke, it's missing some essentials to get it built, and I don't believe it matches what they deploy.

I've got some detailed teardowns over on the indilib forums if you want more details about the internals.

Ahhh... Thanks for the tip but I was thinking of a different approach. Can you point me to your findings though? Might save me some time, between work and life, there is not much time to spent on these things.

 

I have to say, ZWO is being a major criminal here.

Edited by Nightrook, 13 August 2024 - 12:17 PM.

[pupak]

libgphoto2 is free software. Maybe he should modify it so that it forms the interface between the non-ZWO camera and the AsiAir.
He would think that he is communicating with the CANON 6D and would then control the IMX6200. Is it realistic?

[AstroPotamus]

Ideally, there should be a penalty for non compliance with OSSW and ASI should be the subject of a law suit.

[tjay]

Ideally, there should be a penalty for non compliance with OSSW and ASI should be the subject of a law suit.

There sort of is, but it requires the owner of the Open Source Software to attempt to enforce their license terms, which requires somebody pay a lawyer to do something.

[billndotnet]

Ahhh... Thanks for the tip but I was thinking of a different approach. Can you point me to your findings though? Might save me some time, between work and life, there is not much time to spent on these things.

 

I have to say, ZWO is being a major criminal here.

Most of my teardowns are here:
https://www.indilib....e-licences.html

[ASJet]

Managed to crack ASIAIR and install it on a vanilla raspberry PI 4B.
 

 

 

 

You can see that SN 01F43F82 is not present on their website (https://asiair.en.zwo.biz/#/license) as a valid Serial.

Of course the power related things don't work.

 

 

 

I don't have ZWO hardware yet, so I don't know if it really works for real. Currently trying to make it work with non-zwo hardware.

Nice work,I wonder when can you release the open source version of the asiair image,I'm happy to be your first tester.

[cucubits]

I have to say, ZWO is being a major criminal here.

Couldn't agree more. 

 

Just stumbled across this thread. I've been running a cracked asiair image on a RPI4 for a while now. I'm handy with unix, here to help if you need a hand. 

[Xilman]

Just bought a Seestar and unfortunately found this forum too late as I  had updated the firmware upon startup..

 

Tried the root script and no luck. Seestar responds with a verbal "update of firmware failed". Tried SSH into pi with password raspberry and it did not allow a connection.

 

Darryl

Bought one a few days ago with the same result.#

 

A few minutes ago I found this https://github.com/c...seestar-s50_img

 

Not yet tried it but it seems that removing the battery and powering through the USB port, together with mucking about with resets, may be more successful. The instructions given indicate that a Linux system may be required to control everything.

 

As always, this is not advice that you try it. You may brick your device.

 

Paul
 

[AstroPotamus]

I managed to jailbreak my SeeStar and install some SSH keys on it and poked around looking at things.  Then I stopped.  I just use it like a normal person and use the app plus the incredible https://github.com/s...rld/seestar_run Python to do mosaics and stuff.  You really don't gain anything other than a sense of accomplishment by cracking into it.  It's good at what it does, so I've decided to let it be what it is.

 

I assume that software upgrades in the future will add things like internal mosaic capabilities, perhaps scheduling, or at least target selection and imaging times.

[billndotnet]

I managed to jailbreak my SeeStar and install some SSH keys on it and poked around looking at things.  Then I stopped.  I just use it like a normal person and use the app plus the incredible https://github.com/s...rld/seestar_run Python to do mosaics and stuff.  You really don't gain anything other than a sense of accomplishment by cracking into it.  It's good at what it does, so I've decided to let it be what it is.

 

I assume that software upgrades in the future will add things like internal mosaic capabilities, perhaps scheduling, or at least target selection and imaging times.

My motivation to root mine was to examine the software fork from the ASIair and see what other software licenses they're violating, after last year's brouhaha. They doubled down and use even more GPL covered code in the Seestar, on top of what they're still using in the ASIair. They put all of their customers at risk of having an unsupportable device with these shenanigans.

[astro2024]

Couldn't agree more. 

 

Just stumbled across this thread. I've been running a cracked asiair image on a RPI4 for a while now. I'm handy with unix, here to help if you need a hand. 

I would like to try your cracked asiair iso image I have an rpi4 and an asiair v1 which is outdated that would be great from you

[kepler64]

how can I get the seestarOS.zip file needed to restore the emmc? The google drive link and torrent mentioned in the github repository do not work for me. Unfortunately I was too brave updating my rooted Seestar yesterday....

 

Thanks a lot.

[kepler64]

OK, got the image, did the restore, everything is fine now! Thank you again.

btw the name of the image has to be dump_seestar.img

 

CS, Markus

[kaicyung]

To whoever jailbroke into a Seestar S50, then tried to make it asi-air compatible, but messed up, and cause firmware to fail communication with any client, then returned the unit back to ZWO....


well, I have your unit. ZWO sold it to me as a new unit. It took me this long afternoon to figure all this out. You messed up the xml file.

 

To ZWO, it's not cool to ship a refurb unit to new and unsuspecting customers. Especially not reflashing the firmware back to factory settings.

[billndotnet]

To whoever jailbroke into a Seestar S50, then tried to make it asi-air compatible, but messed up, and cause firmware to fail communication with any client, then returned the unit back to ZWO....


well, I have your unit. ZWO sold it to me as a new unit. It took me this long afternoon to figure all this out. You messed up the xml file.

 

To ZWO, it's not cool to ship a refurb unit to new and unsuspecting customers. Especially not reflashing the firmware back to factory settings.

holy smokes, that means it wasn't tested at all. The least they could have done was re-image it. Was it maybe just reboxed by the retailer?