71 replies to this topic

[Lyqyd]

Update to nsh today adds Connection Forwarding, for seamless connections when passing through another server.

[Lyqyd]

I've just updated this to add in support for the framebuffer API, which adds a number of nice features:

• Better responsiveness
  
• Less chatty on rednet
  
• Resume previous sessions (if ended with Ctrl-T/R/S at the client)

Also new is tentative support for binary file transfer (using binget and binput files from the github repository), use with caution.

As usual, see first post for more details.

[natedogith1]

I have a bug report for your program. Tested in 1.6pr7 and a quick look at the pastebin to see if it was just an old release. Running nsh with no arguments tells me that I failed to connect to the server, rather than telling me how to actually use the program.

[Lyqyd]

You are correct! I'm away from my computer right now, but the change to correct the issue is to change line 564 from

elseif #args <= 2 then

To this:

elseif #args >= 1 then

Not sure why I had that wrong in there! I'll update the pastebin when I can.

[CaptRanger]

Is this on the list to get working on 1.6X?

[Lyqyd]

Yeah, if it's broken again on 1.6, I'll update it to fix it as soon as I can. I had gotten it working under the 1.6 betas, but I'll check again under the 1.63 release.

[joebodo]

Can you only connect to one remote computer at a time?

I connected to host 59 and when I try to connect to 40 as well (from a different shell), I get the message '59'.

I am using the command 'nsh <host ID>'

Thanks

[Lyqyd]

That should work fine, I will have to do some testing. I assume you're using multishell in the new version of CC?

[joebodo]

Lyqyd, on 12 May 2014 - 07:00 PM, said:

That should work fine, I will have to do some testing. I assume you're using multishell in the new version of CC?

Yes, multishell. I'm using CC 1.63.

Edited by joebodo, 13 May 2014 - 05:43 PM.

[joebodo]

One additional thing - I had to add a type = 'string' check on each message in the rednet_message event. I'm sending some non-strings as well and that was causing nsh to abort.

Not sure if others are having issues (couldn't find any posts in the bugs topic) - but rednet on our server only seems to work if there are at most 2 computers. So, i hacked in rednet from 1.58 and everything is working great.

[RoD]

I can't use the get/put programs (i already downloaded them and placed in the host pc) it frezes the client.

[Lyqyd]

I'm not sure how I missed your post, RoD, but the work I did recently with nsh/LyqydNet/LyqydOS ended up including a fix for the issue. If you re-download nsh and the get/put programs, you should find that they're working fine again.

[lare290]

So I found this program on a floppy, from a dungeon
First I was like "Ooh floppy, it must be cool new game"
Then I was like "Oh damn, it is just something program I dun need"

[Lupus590]

graphical bug report

the underscore is ment to be white (and is on the server, running vncd)

Edited by Lupus590, 13 April 2015 - 12:13 PM.

[LDDestroier]

Hey Lyqyd, I'm making a server on my poweredge (which I just got working, damn RAID), and I put nsh in a resource pack to make it a default program. However I did have a small (fixed, but still) problem with the detection of the framebuffer API. I like the current detection system for it, but could you make it also look in the directory where the current program is running in? I had to add the "/rom/programs/" directory in the table where it looks. Cheers.

[Lyqyd]

I'll look in to adding that. Shouldn't be too difficult. If you were adding a resource pack, why not put framebuffer in /rom/apis, so it would be loaded at boot anyway? Then nsh won't have to look for it at all!

[LDDestroier]

Lyqyd, on 23 April 2015 - 03:51 AM, said:

I'll look in to adding that. Shouldn't be too difficult. If you were adding a resource pack, why not put framebuffer in /rom/apis, so it would be loaded at boot anyway? Then nsh won't have to look for it at all!

...you make a convincing argument.

[Lupus590]

So I tried to do something stupid...

Two pocket computers, both running vncd.
the first connects to the second then (through the first) I connect the second to the first.
The first freezes, the second give the expected error of "can't connect".

Edited by Lupus590, 29 May 2015 - 09:51 PM.

[Lyqyd]

I've pushed an update to vncd that fixes the looped connection issues. Seems I had forgotten to include the nsh API stub so that nsh could detect the looping condition! Whoops. If you had installed nsh & vncd via Packman, I've also just changed the package layout for these tools. It won't change again, but this one time, you'd want to update it via packman remove nsh followed by packman fetch install nsh-get nsh-put vncd. If you didn't want the get and put programs, packman fetch install nsh vncd would work also.

[Lupus590]

The usage printout is a bit unclear in my opinion. Also it looks like the documentation in the OP needs updating.