Saying hi!

[geckojsc]

Hi everyone! I've been following the Pandora for quite some time, I was one of the last preorderers back in 2009, though I never recieved my device. I'm not worried about that, I know how many issues the whole team went through.

So I recently started at university and found myself in need of a lightweight device for coding and taking notes, like a netbook or something. The top priorities were small size, openness, and long battery life. And that's what brought me back here. I was going to try chasing up my preorder, but realised I would have to pay to upgrade to the 1GHz unit anyway, and that it wasn't fair for ED's company to be losing out when it wasn't them that let me down. So instead, I decided to buy the awesome 1GHz megapack! It arrived already but I got it delivered to London, I won't be able to pick it up until the end of the week.

In the meantime, I've been thinking about development. I already have a few ideas for projects/ports and things (the textadept editor, and the ooc programming language). I just have a couple of questions:

What's the best way (in 2014) to get development tools set up on the device? (cdevtools, the Code::Blocks package, the updated cdevtools2 that was posted recently, or something else?)

The cdevtools readme says I should symlink /usr/local to a directory on the SD card, for custom packages and stuff. Is this safe? Are there any pitfalls I should look out for here?

Anyway, I'm really excited to finally be getting one of these devices after all these years. Looking forward to making and playing awesome stuff on it. See you all around

 

[ekianjo]

What's the best way (in 2014) to get development tools set up on the device? (cdevtools, the Code::Blocks package, the updated cdevtools2 that was posted recently, or something else?)

I'd say Code::blocks since it packs up tons of libraries you would need for compiling about anything out there.  

 

[geckojsc]

I'd say Code::blocks since it packs up tons of libraries you would need for compiling about anything out there.  

Oh, looks like it has SDL2 included? That's quite important for my usage, I'll go with that. Thank you!

 

[ekianjo]

I'd say Code::blocks since it packs up tons of libraries you would need for compiling about anything out there.  

Oh, looks like it has SDL2 included? That's quite important for my usage, I'll go with that. Thank you!

Yeah, it's included It has qt4 and soon qt 5 when PtitSeb releases the next update. 

 

[KickAss]

Hi and welcome!


You better grab your pandora and start the coding. I'm looking forward to your contributions. We can't have enough motivated and capable devs.


The code::blocks pnd is a massive download - my advise would be to not download it with the onboard wifi. Get a dongle instead. Or, of course, download elsewhere and transfer the file via usb.


In general the CLI tools pnd is also very handy.


Be aware tho, that our angstrom based distro and maybe even the pnd system as we know it may not get supported for very much longer, as ppl already started to work on alternatives.


Great to have you onboard!

 

[ekianjo]

Be aware tho, that our angstrom based distro and maybe even the pnd system as we know it may not get supported for very much longer, as ppl already started to work on alternatives.

That's just for packaging, but producing binaries should still remain very much the same process.  

 

[dimag0g]

Hi there,

Just wanted to say that original cdevtools is discontinued, and while I'm trying to do a sequel, it currently has no features compared to Code:blocks apart from size. And I have no plans to add libraries to it, only stuff supported by the firmware.

Concerning linking /usr/local, you only need this if you install extra programs and libraries to your device AND you decide to place them in /usr/local/. Cdevtools PND will use appdata/cdevtools/usr/ for installed packages by default. If you build stuff from sources, it will install to appdata/cdevtools/usr/local by default, and you have total control over the installation process via $PREFIX variable.

The whole thing with /usr/local/ symlinking is about being able to run apps you build with cdevtools outside of cdevtools environment. To achieve it, you need to symlink both /usr/local and appdata/cdevtools/usr/local to the same place (you might as well symlink /usr/local to appdata/cdevtools/usr/local). The security impact depends on what you do exactly (and considering that any PND may execute code from appdata directory I wouldn't be too worried); but keep in mind that you will need to keep the SD card inside the whole time you run apps using /usr/local if you symlink it to SD.

 

[klapse]

I really like cdevtools. Use it for xu4.

 

[geckojsc]

Cdevtools PND will use appdata/cdevtools/usr/ for installed packages by default. If you build stuff from sources, it will install to appdata/cdevtools/usr/local by default, and you have total control over the installation process via $PREFIX variable.

Ah thanks! I guess it's the same for the codeblocks interface too.

One more thing - is it worth having a separate card/partition and formatting it to ext2? Or if I'm using codeblocks where virtually every library is included, is that not necessary?

On the other hand, is there any reason why I shouldn't just format one card with ext2/3/4 and use it for all my apps? I'm a bit of a noob when it comes to filesystems.

 

[ekianjo]

One more thing - is it worth having a separate card/partition and formatting it to ext2? Or if I'm using codeblocks where virtually every library is included, is that not necessary?

You need an ext2 for the card where you use code::blocks. For dev purposes a linux format is a must. so ext2 is recommended. 

On the other hand, is there any reason why I shouldn't just format one card with ext2/3/4 and use it for all my apps? I'm a bit of a noob when it comes to filesystems.

If you don't intend for the card to be readable on Windows systems or other devices, yeah, you can use ext2 and live with it no problem.  

 

[KickAss]

On ext2 you can edit the files in your .pnd directly and don't need to re-package after.


Saves lots of time and nerves while debugging!

 

[ekianjo]

On ext2 you can edit the files in your .pnd directly and don't need to re-package after.

What do you mean exactly by "edit the files in your .pnd directly" ?

Is this something I am not aware of ?  

 

[KickAss]

you would mount your pnd using the pnd_run.sh script like so:


../pnd_run.sh -p /path/to/your.pnd -m
it will be mounted to /mnt/utmp/yourpndname1

you can access and edit your files there and once you're done, you'd just unmount your pnd again, like so:

../pnd_run.sh -p /path/to/your.pnd -u

 

[canseco]

More devs are more than welcome,

If you don't intend for the card to be readable on Windows systems or other devices, yeah, you can use ext2 and live with it no problem.

You can read ext2/3/4 with this program on Windows:

http://sourceforge.net/projects/ext2read/

 

[Orkekum]

Looking forward to see what you create!

 

[ekianjo]

More devs are more than welcome,

If you don't intend for the card to be readable on Windows systems or other devices, yeah, you can use ext2 and live with it no problem.

You can read ext2/3/4 with this program on Windows:

http://sourceforge.net/projects/ext2read/

Nope, I tried and it never worked for me. On different computers, even.

you would mount your pnd using the pnd_run.sh script like so:


../pnd_run.sh -p /path/to/your.pnd -m
it will be mounted to /mnt/utmp/yourpndname1

you can access and edit your files there and once you're done, you'd just unmount your pnd again, like so:


../pnd_run.sh -p /path/to/your.pnd -u

Thanks, I was not aware it was possible. So when you unmount the PND you basically get a PND with new content ?

How about the appdata thing then ? does it save the data in appdata even if your PND is not in the pandora folder?

 

[geckojsc]

you would mount your pnd using the pnd_run.sh script like so:


../pnd_run.sh -p /path/to/your.pnd -m
it will be mounted to /mnt/utmp/yourpndname1

you can access and edit your files there and once you're done, you'd just unmount your pnd again, like so:


../pnd_run.sh -p /path/to/your.pnd -u

Sounds awesome, so that would mean that when I run the pnd, it gets the contents from the mounted folder instead of the archive itself?

If you don't intend for the card to be readable on Windows systems or other devices, yeah, you can use ext2 and live with it no problem.

You can read ext2/3/4 with this program on Windows:

http://sourceforge.net/projects/ext2read/

That's okay, I can always boot to Linux or use FTP or something whenever I need to access the card

Looking forward to see what you create!

Thanks!

 

[Granitehead]

That reminds me: Are you guys aware you don't even need to "package" a .pnd to run it? You can just make a folder in one of the usual search locations and put the files into it, it should be recognized by the system.

No mounting or anything necessary.

 

[KickAss]

Srsly?


Why aint that info documented?


I always thought libpnd and pndnotifyd kinda rely on actual .pnd files.


Anyways, thats also good to know

 

[ekianjo]

That reminds me: Are you guys aware you don't even need to "package" a .pnd to run it? You can just make a folder in one of the usual search locations and put the files into it, it should be recognized by the system.


No mounting or anything necessary.

Wow, interesting...