Hi

I'm currently developing a new terminal
software: https://github.com/mkrueger/icy_term/tags

It's written in rust and cross platform. It's specifically written for BBSes. It has a good ansi engine, can copy and paste, supports petscii and atascii.

It is not finished - esp. the UI needs work. I can need feedback & ideas.

Thanks

... Traveler, there is no path. Paths are made by walking.

--- Mystic BBS v1.12 A48 2022/07/15 (Linux/64)
* Origin: 2o fOr beeRS bbs>>>20ForBeers.com:1337 (21:2/150)

Re: New terminal program
By: Omnibrain to All on Mon Nov 14 2022 12:44 pm

I'm currently developing a new terminal
software: https://github.com/mkrueger/icy_term/tags

It is not finished - esp. the UI needs work. I can need feedback & ideas.

Howdy,

Build instructions would be useful - I'm no rust user.

Anyway, I had a go, and got stuck here:

─$ cargo build
Updating crates.io index
error: failed to get `iced` as a dependency of package `icy_term v0.1.6 (/Users/deon/Nextcloud/Downloads/icy_term)`

Caused by:
failed to load source for dependency `iced`

Caused by:
Unable to update /Users/deon/Nextcloud/Downloads/iced

Caused by:
failed to read `/Users/deon/Nextcloud/Downloads/iced/Cargo.toml`

Caused by:
No such file or directory (os error 2)

(I'm building on a Mac M1)


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

I'm currently developing a new terminal
software: https://github.com/mkrueger/icy_term/tags

It's written in rust and cross platform. It's specifically written for BBSes. It has a good ansi engine, can copy and paste, supports petscii
and atascii.

It is not finished - esp. the UI needs work. I can need feedback & ideas.

Thanks

Downloaded the windows version and was able to run it on Win11 x64. 2 major problems I ran into were:
1) The tab key doesn't work (no way to move from entry to entry other than via mouse) while entering data in the phone book.

2) When logging out of my bbs, the bbs screen wouldn't clear. Because the logout screen mostly blocked the program screen, I could tell something was scrolling on the program screen but couldn't read it. As soon as I closed the bbs screen, the program screen closed out as well.

It did run somewhat faster than Netrunner, my default client. I will give it a few updates before I try it again. But I will try it again!

ogg
Sysop, Altair IV BBS
altairiv.ddns.com:2323

... Confucius say: "Man who runs behind car gets exhausted"

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

On 14 Nov 2022, Omnibrain said the following...

Hi

I'm currently developing a new terminal
software: https://github.com/mkrueger/icy_term/tags

i will definately check this out.

... There are three kinds of people: Those who can count, and those who can't

--- Mystic BBS v1.12 A47 2021/12/24 (Linux/64)
* Origin: Clutch BBS * telnet://bbs.clutchbbs.com (21:1/199)

It is not finished - esp. the UI needs work. I can need feedback & ideas.

Excellent to see someone working on things like this :)

I think full cp437 support, full utf8 support, and the ability to switch between the two on the fly would be very cool. Additionally, supporting all 256 colors in xterm would be great. This would enable a lot of games and things not originally intended for BBSes to work as doors.

--- Mystic BBS v1.12 A48 2022/07/11 (Linux/64)
* Origin: m O N T E R E Y b B S . c O M (21:4/173)

Re: New terminal program
By: deon to Omnibrain on Tue Nov 15 2022 09:13 am

Howdy,

It is not finished - esp. the UI needs work. I can need feedback & ideas.

Build instructions would be useful - I'm no rust user.

Ignore me. Didnt realise that you provided binary builds - so I found the MAC M1 version and it works well.

Some great feedback already - would be good to support 256 color ANSI if that was possible, I could see myself playing with that.

Couple of things I've noticed:
* Flash doesnt flash (I use flashing in my AnsiTEX that I've created).
* Also, with my AnsiTEX, I use cursor positioning to get to the bottom line (line 24), and it seems like it's off by 1 char (and thus the first char is on the previous line, column 80, and the 2nd char is on the bottom line, col 1).
* Pressing down arrow on the bottom line (ie: moving around with the cursor) is rendering a dot, as is right arrow in col80 - not sure if that is the editor or the terminal (but not noticed it with syncterm - I'll try that out for sure).

Any chance of supporting viewdata? :)


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

Hi

Some great feedback already - would be good to support 256 color ANSI if that was possible, I could see myself playing with that.

Why 256 color ANSI and which escape sequence?
There is full RGB support:

ESC[ 38;2;r;g;b m Select RGB foreground color
ESC[ 48;2;r;g;b m Select RGB background color

I'll put that on the todo list. If you can point me to the 256 color sequences I'll implement that as well.

* Flash doesnt flash (I use flashing in my AnsiTEX that I've created).
Any chance of supporting viewdata? :)

Ok missed that - only implemented ice colors :).
I didn't know about Viewdata - https://en.wikipedia.org/wiki/Viewdata
Y sounds interesting. I'll look if I can find a BBS for that if I do I'll add support. Did that for Petscii and Atascii.

* Also, with my AnsiTEX, I use cursor positioning to get to the bottom line (line 24), and it seems like it's off by 1 char (and thus the first char is on the previous line, column 80, and the 2nd char is on the
bottom line, col 1).

Can I test that somewhere?

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Re: Re: New terminal program
By: Omnibrain to deon on Tue Nov 15 2022 01:21 am

Howdy,

Why 256 color ANSI and which escape sequence?
There is full RGB support:
ESC[ 38;2;r;g;b m Select RGB foreground color
ESC[ 48;2;r;g;b m Select RGB background color

Ahh, ok - 256 color might be simplier, and I guess its a nice to have. https://en.wikipedia.org/wiki/ANSI_escape_code

ESC[38;5;(n)m Select foreground color
ESC[48;5;(n)m Select background color

I didn't know about Viewdata - https://en.wikipedia.org/wiki/Viewdata
Y sounds interesting. I'll look if I can find a BBS for that if I do I'll add support. Did that for Petscii and Atascii.

You can point to my BBS ansitex.bbs.dege.au, port 516.

I dont have a lot in there yet (this is viewdata, and I'm still developing it) - but page 999a (*999#) is a test page, and page 999b (# again) will show you the characters.

I have a web terminal interface at https://alterant.bbs.dege.au/2018/12/31/oztex/ that you can use to see what it should look like.

AnsiTEX is available ansitex.bbs.dege.au port 23, which is the same usability, but using ANSI instead of Viewdata. (Again dont have much there, as it is still a work in progress when I moments of enthusiasm.)


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

On 14 Nov 2022, Omnibrain said the following...

Hi I'm currently developing a new terminal
software: https://github.com/mkrueger/icy_term/tags
It's written in rust and cross platform. It's specifically written for BBSes. It has a good ansi engine, can copy and paste, supports petscii
and atascii.
It is not finished - esp. the UI needs work. I can need feedback & ideas.

Thanks

Will it run on Steam Deck? Steam deck uses everything in flat packs. This is the only platform I have that I can't BBS from without using the ftelnet. Would like a client that works for deck :D

|23|04Dr|16|12Claw
|16|14Sysop |12Noverdu |14BBS |04(|14Noverdu.com|04)
|10Standard Ports for SSH/Telnet Web/HTTP://|14Noverdu.com:808
|20|15fsxNet/MRC Chat/Registered Doors!/50Nodes/No Time Use! Stay On!|16|07

--- Mystic BBS v1.12 A47 2021/12/24 (Linux/64)
* Origin: Noverdu BBS (21:1/210)

Will it run on Steam Deck? Steam deck uses everything in flat packs. This is the only platform I have that I can't BBS from without using the ftelnet. Would like a client that works for deck :D

I develop using linux (most of the time - macos sometimes). So the only thing missing is a flat pack then. I've not really looked into linux packaging yet just provided a .deb.

But you're right that I should make a flat pack - I put it on the todo list.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Hello Omnibrain,

14 Nov 22 12:44, you wrote to All:

Hi
I'm currently developing a new terminal software: https://github.com/mkrueger/icy_term/tags

It's written in rust and cross platform. It's specifically written for BBSes.
It has a good ansi engine, can copy and paste, supports petscii and atascii.

It is not finished - esp. the UI needs work. I can need feedback &
ideas.

how about RIP graphics? :)

Regards,
--
dp

telnet://bbs.roonsbbs.hu:1212 <<=-

... 11:07pm up 47 days, 10:33:25, load: 79 processes, 283 threads.
--- GoldED/2 1.1.4.7+EMX
* Origin: Roon's BBS - Budapest, HUNGARY (21:4/148)

how about RIP graphics? :)

Thought about it but I don't think it's on the todo list for the near future. Or do you know a nice RIP BBS to call?

ReGIS is another thing I'm searching for. Does anyone use that?

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

*** Quoting Omnibrain from a message to Roon ***

Thought about it but I don't think it's on the todo list for the near future. Or do you know a nice RIP BBS to call?

https://serionbbs.com/rip-bbs

This one is pretty cool.


Jay

... I used to get high on life but lately I've built up a resistance

--- Telegard v3.09.g2-sp4/mL
* Origin: Northern Realms | tg.nrbbs.net | 289-424-5180 (21:3/110.1)

https://serionbbs.com/rip-bbs

Ok will look into it - just finished Viewdata support :)

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

On 15 Nov 2022, Omnibrain said the following...

I develop using linux (most of the time - macos sometimes). So the only thing missing is a flat pack then. I've not really looked into linux packaging yet just provided a .deb.

But you're right that I should make a flat pack - I put it on the todo list.

That would be awesome! Then it will just work on all flavors. I will finally get a proper term on my Steam Deck!

|23|04Dr|16|12Claw
|16|14Sysop |12Noverdu |14BBS |04(|14Noverdu.com|04)
|10Standard Ports for SSH/Telnet Web/HTTP://|14Noverdu.com:808
|20|15fsxNet/MRC Chat/Registered Doors!/50Nodes/No Time Use! Stay On!|16|07

--- Mystic BBS v1.12 A47 2021/12/24 (Linux/64)
* Origin: Noverdu BBS (21:1/210)

Ahh, ok - 256 color might be simplier, and I guess its a nice to have.

Implemented 8bit & 24bit color codes. One of the next things I do is to review for missing escape codes. Everytime I think I do support almost all new ones pop up :).

I just finished the Viewdata support - biggest issue was to find the correct specs :). But I got something working only thing missing are the alternate character sets - didn't find any BBS using them. I'll look on that on request. But so far it works pretty well - next release will have it.

I don't know when I release it - however next week it's out for sure and needs testers.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

1) I know - that's just a placeholder I want to make a much better phonebook. The UI lib I use unfortunately doesn't default to some sane standards.

2) I'm not sure I can follow what you exactly mean here - which BBS?

--- Mystic BBS v1.12 A48 2022/07/15 (Linux/64)
* Origin: 2o fOr beeRS bbs>>>20ForBeers.com:1337 (21:2/150)

Ok scroll down did tell me which bbs :)

Thanks for trying.

... New Opcode #8: IPX - Increase Power and eXplode

--- Mystic BBS v1.12 A48 2022/07/15 (Linux/64)
* Origin: 2o fOr beeRS bbs>>>20ForBeers.com:1337 (21:2/150)

Build instructions would be useful - I'm no rust user.

Anyway, I had a go, and got stuck here:

─$ cargo build

I recommend using the released version. You need a local iced unfortunately it requires a little patch - one method isn't public I need to be public.
However that'll change in that library.

If you really want to build it - get https://github.com/iced-rs/iced
and apply:
---
a/graphics/src/widget/canvas/geometry.rs+++ b/graphics/src/widget/canvas/geometry.rs
@@ -11,7 +11,9 @@ use crate::Primitive;
pub struct Geometry(Primitive);

impl Geometry {
- pub(crate) fn from_primitive(primitive: Primitive) -> Self {
+ /// TODO
+ pub fn from_primitive(primitive: Primitive) -> Self {

And something more easy - get https://github.com/mkrueger/icy_engine
and put it in the same directory than icy_term/ and iced/.
Then it should build.

When iced gets a proper way to paint images I switch back to the github version and building gets easier.

I splitted out the icy_engine because I need the ansi engine for other projects as well.

Building will change - it's now complex because I needed something for sixel drawing.

--- Mystic BBS v1.12 A48 2022/07/15 (Linux/64)
* Origin: 2o fOr beeRS bbs>>>20ForBeers.com:1337 (21:2/150)

Re: Re: New terminal program
By: Omnibrain to deon on Wed Nov 16 2022 03:33 pm

Howdy,

I just finished the Viewdata support - biggest issue was to find the correct specs :). But I got something working only thing missing are the alternate character sets - didn't find any BBS using them. I'll look on that on request. But so far it works pretty well - next release will have it.

Awesome - I look forward to trying it out.



...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

On 15 Nov 2022, Omnibrain said the following...

I recommend using the released version. You need a local iced unfortunately it requires a little patch - one method isn't public I
need to be public. However that'll change in that library.

Flat pack yet?

|23|04Dr|16|12Claw
|16|14Sysop |12Noverdu |14BBS |04(|14Noverdu.com|04)
|10Standard Ports for SSH/Telnet Web/HTTP://|14Noverdu.com:808
|20|15fsxNet/MRC Chat/Registered Doors!/50Nodes/No Time Use! Stay On!|16|07

--- Mystic BBS v1.12 A47 2021/12/24 (Linux/64)
* Origin: Noverdu BBS (21:1/210)

Awesome - I look forward to trying it out.

Try it out:
https://github.com/mkrueger/icy_term/releases/tag/0.1.7

Some feedback regarding to Viewdata would be nice. This release contains blinking support as well as extended colors.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Flat pack yet?

https://github.com/mkrueger/icy_term/releases/tag/0.1.7

Not sure if it works - it's complex to distribute software. Took ages for me to get the .dmg running. Windows is the easiest thing - the exe just works out ouf the box.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Re: Re: New terminal program
By: Omnibrain to deon on Thu Nov 17 2022 12:32 pm

Awesome - I look forward to trying it out.

Try it out:
https://github.com/mkrueger/icy_term/releases/tag/0.1.7

Some feedback regarding to Viewdata would be nice. This release contains blinking support as well as extended colors.

This is awesome!

I like you work.

I had a quick play, and there are a couple of things that just need to be tweaked.

Is there any off by 1 thing happeing? When I position the cursor at the bottom of the screen I think it is wrapping and appearing on the top line.
(You can see this on my BBS when entering a page with *, the page input should be at on the bottom line, or if you are idle to long, the IDLE DISCONNECT appears on the bottom line.

Did you implement cursor right? The login box, the cursor is not being positioned correctly (its going to the right line, but not moving right to the correct position.

One last request, how do you "scroll back" - on the Mac (in Syncterm), I scroll back looking at history using the scroll wheel - if that was easy to do to, then it would be a great additionl. (Keystrokes is ok too - just wasnt sure what they were...)

Thanks for adding viewdata, it'll help me work on my videotex implementation.



...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

Re: New terminal program
By: deon to Omnibrain on Tue Nov 15 2022 05:53 pm

Howdy,

* Also, with my AnsiTEX, I use cursor positioning to get to the bottom line (line 24), and it seems like it's off by 1 char (and thus the first char is on the previous line, column 80, and the 2nd char is on the bottom line, col 1).

Just though I'd share that with 0.1.7 this is still a challenge.

I use cursor positioning with a synchronet javascript function (which I assume is ultimately calling CSI n;mH, but it is behaving unpredictably (off by 1 sometimes, or centering on the bottom line instead of going to position 0 on line 24).

Happy to provide some screen shots if that helps.


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

On 17 Nov 2022, Omnibrain said the following...

https://github.com/mkrueger/icy_term/releases/tag/0.1.7

Not sure if it works - it's complex to distribute software. Took ages
for me to get the .dmg running. Windows is the easiest thing - the exe just works out ouf the box.

Well it installed but got this error while connecting with quick connect to Noverdu.com. Noverdu.com:22 and Noverdu.com:23

failed to lookup address information: Temporary failure in name resolution

|23|04Dr|16|12Claw
|16|14Sysop |12Noverdu |14BBS |04(|14Noverdu.com|04)
|10Standard Ports for SSH/Telnet Web/HTTP://|14Noverdu.com:808
|20|15fsxNet/MRC Chat/Registered Doors!/50Nodes/No Time Use! Stay On!|16|07

--- Mystic BBS v1.12 A47 2021/12/24 (Linux/64)
* Origin: Noverdu BBS (21:1/210)

Just though I'd share that with 0.1.7 this is still a challenge.
I use cursor positioning with a synchronet javascript function (which I assume is ultimately calling CSI n;mH, but it is behaving unpredictably (off by 1 sometimes, or centering on the bottom line instead of going to position 0 on line 24).

Screenshot won't help I need to look at the whole ANSI page. Can I reproduce that somehow? Which bbs/which use case? Usually n;mH should work.
(Otherwise that editor here would screw up very badly)

So I assume something else interferes with that. I need to look at the circumstances where it's called.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

On Monday, November 14th Omnibrain was heard saying...

I'm currently developing a new terminal software: https://github.com/mkrueger/icy_term/tags
It's written in rust and cross platform. It's specifically written for BBSes. It has a good ansi engine, can copy and paste, supports petscii and atascii.

Very nice, I will be checking this out shortly. The fact that it's in Rust sparks my interest -- accepting PRs?


--
|08 ■ |12NuSkooler |06// |12Xibalba |08- |07"|06The place of fear|07"
|08 ■ |03xibalba|08.|03l33t|08.|03codes |08(|0344510|08/|03telnet|08, |0344511|08/|03ssh|08)
|08 ■ |03ENiGMA 1/2 WHQ |08| |03Phenom |08| |0367 |08| |03iMPURE |08| |03ACiDic
--- ENiGMA 1/2 v0.0.13-beta (linux; x64; 16.16.0)
* Origin: Xibalba -+- xibalba.l33t.codes:44510 (21:1/121)

Re: Re: New terminal program
By: Omnibrain to deon on Sat Nov 19 2022 08:16 am

Howdy,

Screenshot won't help I need to look at the whole ANSI page. Can I reproduce that somehow? Which bbs/which use case? Usually n;mH should work.
(Otherwise that editor here would screw up very badly)

Yup, so my ANSItex BBS - ansitex.bbs.dege.au port 23

After you connect, if:
* You got to a page (eg *999#), pressing the "*" will show it in col 80, when it should be col 1 on the bottom line.

* The word "loading", when it is retrieving a frame (after pressing "#") should replace the page on the bottom line (col 1).

* Go to the login page *0#, then go to *999#, the "*" will be in the middle of the page (and still not on the bottom line)

It appears that cursor movement is not happening (it's being ignored), and any input is continuing on where the cursor would have stopped after rendering a page.

Hope that makes sense...


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

Re: Re: New terminal program
By: deon to Omnibrain on Sun Nov 20 2022 12:08 pm

Howdy,

Screenshot won't help I need to look at the whole ANSI page. Can I reproduce that somehow? Which bbs/which use case? Usually n;mH should work.
(Otherwise that editor here would screw up very badly)

While my last message was for ANSI, I thought I'd give you the steps for the viewdata issue as well.

Some host, but port 516 is viewdata (ansitex.bbs.dege.au).

After you connect, if you go to a page (eg: *999#), the "*" appears on the top line, when it should be on the bottom line (char 1) (like with ANSItex).

If you go to the login screen (*0#), you'll see the "User"/"Password" lines, but the cursor is at char 1 not where it should be (that should be clear).

You can compare how it should work, when using my web console (https://alterant.bbs.dege.au).

Hope this helps - did I mention this is awesome that you implemented it? :)


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

Very nice, I will be checking this out shortly. The fact that it's in
Rust sparks my interest -- accepting PRs?

100% I'll accept PRs. It's one project for me to keep my rust skills not rusting :). And to look at iced - that System76 bases the PopOS! desktop on iced made me curious.

I'm still searching for an UI lib in rust which is capable for my side projects :)'

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Hi

Hope this helps - did I mention this is awesome that you implemented it?

Was mostly my fault I wanted to look at your BBS several times but missed the option to create an account - the QR code never worked for me :(.

Now I managed to hit '0' and saw the issues - y had a bug in the esc[H seqence - fixed.

Viewdata is very old and the spec isn't 100% clear - so I had to guess some things. At least the caret movement should work now - at least better :).

Grab a 0.2.1 release:
https://github.com/mkrueger/icy_term/releases/tag/0.2.1

Implementing Viewdata was fun and gave me the Opportunity to improve my engine code and to learn about - VD was new to me. However I know Teletext so that catched my interest.

If you spot anything else just get back to me - I'll fix it.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Re: Re: New terminal program
By: Omnibrain to deon on Sun Nov 20 2022 02:16 am

Howdy,

Was mostly my fault I wanted to look at your BBS several times but missed the option to create an account - the QR code never worked for me :(.

Do you use SQRL? (The QR Code is for SQRL, and without a client, it wont work...)

Now I managed to hit '0' and saw the issues - y had a bug in the esc[H seqence - fixed.

Viewdata is very old and the spec isn't 100% clear - so I had to guess some things. At least the caret movement should work now - at least better :).

It is working well - thank you. Movement and placement is exactly as I expected.

If you spot anything else just get back to me - I'll fix it.

The only other wierd thing is (with viewdata), the top line looses its color when you accept input on the bottom line.

To see what I mean, after you connect, go to page *999# (you dont need a login to see this page). After the page renders, then press * (as if you were going to another page), and the top line looses its color.

Thanks for what you've done, its awesome.


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

Twas Sunday, November 20th when Omnibrain said...

100% I'll accept PRs. It's one project for me to keep my rust skills not rusting :). And to look at iced - that System76 bases the PopOS! desktop on iced made me curious.

Sweet, I'll try to snag the source soon and start peeking. I gave the latest version a run last night and noted some unsupported ESC sequences, so maybe that's a good spot for me to look initially.


On Sunday, November 20th Omnibrain said...

I'm still searching for an UI lib in rust which is capable for my side projects :)'

What sort of UI are you after?


--
|08 ■ |12NuSkooler |06// |12Xibalba |08- |07"|06The place of fear|07"
|08 ■ |03xibalba|08.|03l33t|08.|03codes |08(|0344510|08/|03telnet|08, |0344511|08/|03ssh|08)
|08 ■ |03ENiGMA 1/2 WHQ |08| |03Phenom |08| |0367 |08| |03iMPURE |08| |03ACiDic
--- ENiGMA 1/2 v0.0.13-beta (linux; x64; 16.16.0)
* Origin: Xibalba -+- xibalba.l33t.codes:44510 (21:1/121)

Sweet, I'll try to snag the source soon and start peeking. I gave the latest version a run last night and noted some unsupported ESC
sequences, so maybe that's a good spot for me to look initially.

There are some missing - for example ansi music completely. I haven't looked into ways playing midi music x-platform yet.
The problem with ESC sequences is that there is not really a standard anywhere and there are some junk sequences out there.

I mostly use:
https://vt100.net/docs/vt510-rm/chapter4.html https://invisible-island.net/xterm/ctlseqs/ctlseqs.html https://en.wikipedia.org/wiki/ANSI_escape_code

for a reference.

What sort of UI are you after?

Initially I wanted to create an ansi drawing tool. The terminal is just a test project :-) - and the engine was made for a drawing tool.

So I need a feature rich control set. I need a high performance custom control drawing way - iced is too slow. And on the drawing grid I need more features than what iced offers - for example I can't mix drawings and images - need to hack the lib for that.

ATM I try to get something done in egui - that allows me to make controls that use shaders directly. That would be awesome. And it has way more controls - let's say they have a context menu - iced doesn't.

Iced isn't bad but just doesn't offer enough to realize a real application. I've already used EGUI - downside there is that the UI code can become quite messy at some point.

A plus for an UI lib would be that it runs in the browser, mobile etc. minimum is cross platform.
I started the ansi drawing tool ~1y ago and used gtk4 bindings - looked awesome but it wasn't really cross platform so I stopped.

Unfortunately there isn't the #1 goto lib for rust yet - I explored druid/egui and iced for my needs and still can't really decide. As said atm I explore egui and try to get something done with fragment shaders. But I've 0 xp with that so it'll take a while.
And that doesn't mean I stop developing the terminal or the backend. Still some things to do.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Do you use SQRL? (The QR Code is for SQRL, and without a client, it wont work...)

Ok guessed my Iphone camera app can do all that - I'll look at it :).

The only other wierd thing is (with viewdata), the top line looses its color when you accept input on the bottom line.

I'll look at it. Viewdata can contain some issues - it's important to get feedback for me.

Thanks for what you've done, its awesome.

TY at least it's an area not many ppl are interested in :).

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

On 11/14/22 13:44, Omnibrain wrote:

I'm currently developing a new terminal
software: https://github.com/mkrueger/icy_term/tags

Sweet... any consideration on getting the main rendering engine working
stand alone, such as a webassembly+canvas client?
--
Michael J. Ryan - tracker1@roughneckbbs.com
--- SBBSecho 3.15-Linux
* Origin: Roughneck BBS - roughneckbbs.com (21:3/149)

The only other wierd thing is (with viewdata), the top line looses its color when you accept input on the bottom line.

Fixed the issue - viewdata seems to fill colors in the current line on caret home and new line - but only when column > 0 - that was the bug.

I didn't find any good spec about viewdata - as said - so I had to guess some things. Fortunately it's not a very complex encoding so bugs are limited.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Sweet... any consideration on getting the main rendering engine working stand alone, such as a webassembly+canvas client?

Would be a side effect if I'm switching to egui. So y it's on the table. But will take some while. I'm building a new computer these days, x-mas etc. can't promise anything in the near future. But so far it looks good with egui - there I can get to the speed level I want.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Just looked at this today..

your ESC[K should clear to the end of the line (with the current
background colour, not black - unless black is the current background
colour)

Also, there is 2 different ways of doing 24bit colour, I assume you're
doing it the correct way, but there is another way that is supported in
BBSes, I found no specs for it, but it looks like this:

ESC[A;XX;YY;ZZt

Where A = 0 if background, or 1 if fore ground

XX is The red component
YY is the green component
ZZ is the blue component

Usually you see this format in > 16 colour ansi art, the reason is,
usually the art will have a standard colour change code, followed by a 24
bit colour code, that way terminals that don't support the 24bit colour
scheme fallback to displaying the standard 16 colours.

I didn't learn about this until someone mentioned my terminal didn't
support 24bit colour, when it did.

there is a good document in the synchronet repository, called cterm.txt
it explains lots of weird ANSI that BBS systems seem to use.

Andrew




--- Talisman v0.46-dev (Windows/x64)
* Origin: Smuggler's Cove - Private BBS (21:1/182)

your ESC[K should clear to the end of the line (with the current

Will look at that I suppose that may affect other clear to end commands as well - will look at the specs.

ESC[A;XX;YY;ZZt

Where A = 0 if background, or 1 if fore ground

Will look at it - do you have any ANSI art using that - then I can look at other tools for output comparison.

TY that helps me a lot.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

The only other wierd thing is (with viewdata), the top line looses its color when you accept input on the bottom line.

Made a mini release:
https://github.com/mkrueger/icy_term/releases/tag/0.2.2

There were some issues on the #999* page which I tried to fix - should now be displayed correctly. If you spot anything just report back.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Fixed the clear bug (was a bug in all clear operations) & added support for these colors.Do you know if they're in the range 0-255? Sometimes it's 0-63.

Will be in the next release (0.2.3 or newer).

Thanks - good finding

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Re: Re: New terminal program
By: Omnibrain to deon on Mon Nov 21 2022 09:36 am

Made a mini release:
https://github.com/mkrueger/icy_term/releases/tag/0.2.2

There were some issues on the #999* page which I tried to fix - should now be displayed correctly. If you spot anything just report back.

Howdy,

This release fixes the top line (line 1) from loosing color. :)

I've noticed now, that line 2 looses color when submitting a form.

After you connect, if you go to the login page, use any username/password, when the status line (bottom line) shows PRESS 1..., the second line looses it's color.

Thanks...


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

I've noticed now, that line 2 looses color when submitting a form.

Now I've kind of a problem.
This sequence clears the attribute:
\x1E\x09\n\x1E

x1E is home home sets the attribute in the line if col > 0, x09 is caret right. The question is how \n is interpreted. if its col=0, line++ the the next x1E doesn't set the attributes in line 2.

When I change \n to col=0 I break the login name/pw positioning. Name is positioned with:
\x1E\x09\x09\x09\x09\x09\x09\x09\x09\x09\x09\x09\x09\x09\x09\x09\n\n\n\n\n\n\n\ n\n\n\nNAME

So name/pw would end up in col 0 - that's why I changed \n to be just line++ in the first place.

Now is the question which behavior is correct? Unfortunately I don't have a spec just a table with commands and this table leaves much room for interpretation.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Re: Re: New terminal program
By: Omnibrain to deon on Tue Nov 22 2022 12:16 am

Howdy,

Now I've kind of a problem.
This sequence clears the attribute:
\x1E\x09\n\x1E

Now is the question which behavior is correct? Unfortunately I don't have a spec just a table with commands and this table leaves much room for interpretation.

All these are cursor movement (not characters are written - or overwritten).

\x1e is "home", takes the cursor to 1,1
\x09 is "right", moves the cursor to the right (but doesnt overrite anything) \x0a is "down", moves the cursor down (some horizontal position)

And just while I'm on this topic:
\x08 is "left"
\x0b is "up" (same horizontal position)

The cursor wraps when it gets to its boundary, ie:
* If you are at position 40 and "right", you are on the next line at position 1 * If you are on the bottom line (24) and down, you wrap to the top line (same col).

And thus, if the cursor is at 40,24 and "right", it wraps to 1,1.


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

All these are cursor movement (not characters are written - or overwritten).

y I know the issue is that 1E fills the attribute to eol like \r does. When I remove that it's fixed as well. I think I do that - didn't find any frame that requires the caret home attr fill.

The only thing I'm failing at is the nxtel drop shadow test.

<esc>T<esc>]<esc>GDrop Shadow<esc>Tk<esc>V<esc>\<7F><7F>

I interpret the esc codes as:
T: blue bg(graphics)
]: set bg =fg
G: Alpha white
T: blue bg(graphics)
k: will be interpreted as sextant
V: will set cyan graphics fg
\: will set black background

'k' is printed for me with blue bg and blue fg. should by cyan fg and black bg but these sequences follow up. If I mess around with setting attrs to the last printed char I mess up everything. Do you know what is wrong here?

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Sent too early

<esc>T<esc>]<esc>GDrop Shadow<esc>Tk<esc>V<esc>\<7F><7F>...

T: blue fg(graphics)
]: set bg =fg
G: Alpha white"Drop Shadow": Text (white fg, blue bg)
T: blue fg(graphics)
k: will be interpreted as sextant - blue bg/fg
V: will set cyan graphics fg
\: will set black background

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Ok figured it out - was somehwere else (as usual) - the hold graphics behavior.

Release:
https://github.com/mkrueger/icy_term/releases/tag/0.2.3

I hope that I now can continue to work on a rendering engine. But I can't tolerate bugs in the parser engine :).

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

On Sunday, November 20th Omnibrain was heard saying...

There are some missing - for example ansi music completely. I haven't looked into ways playing midi music x-platform yet. The problem with ESC sequences is that there is not really a standard anywhere and there are some junk sequences out there.

I can probably help in this area. I have a huge a lot of information on what specs were implemnted by BBSs and related terminals of the era + modern terminals. Hopefully will have time to hopen up a PR fairly shortly-ish (maybe after turkey day weekend) for you to peek at.

What's yoru timeline for TLS/SSL support?

Thanks for the info on the UI want/needs... I'll poke around, but I think you've broken down the ones I really know already :D






--
|08 ■ |12NuSkooler |06// |12Xibalba |08- |07"|06The place of fear|07"
|08 ■ |03xibalba|08.|03l33t|08.|03codes |08(|0344510|08/|03telnet|08, |0344511|08/|03ssh|08)
|08 ■ |03ENiGMA 1/2 WHQ |08| |03Phenom |08| |0367 |08| |03iMPURE |08| |03ACiDic
--- ENiGMA 1/2 v0.0.13-beta (linux; x64; 16.16.0)
* Origin: Xibalba -+- xibalba.l33t.codes:44510 (21:1/121)

Re: Re: New terminal program
By: Omnibrain to deon on Tue Nov 22 2022 03:28 pm

Howdy,

Ok figured it out - was somehwere else (as usual) - the hold graphics behavior.

Release:
https://github.com/mkrueger/icy_term/releases/tag/0.2.3

Awesome, this works great.

I know you are working on other things, so consider this minor and if you dont get a chance to look at it, ok.

Is there a chance you broke/changed "cursor off" (0x14)? I've noticed that the cursor is on now, and remains flashing at the end of what was sent. (Cursor on is 0x11).

And I've noticed an incompatibility. If I overwrite a color change, technically everything after the color change, should be that color unless there is another color change, or the end of the line, whichever occurs first. (I'm about to write an editor, and this will be impacted by this.)

You can see this in action, by connecting and without logging on, type *1# you should get a "red" ACCESS DENIED on the bottom line, then press "*" (to enter a page number). Technically everything after the * should be GREEN, but it remains RED.

Anyway, these are minor, thank you.


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

Hi

Is there a chance you broke/changed "cursor off" (0x14)? I've noticed
that the cursor is on now, and remains flashing at the end of what was sent. (Cursor on is 0x11).

Could be - I basically refactored everything just to fix a bug that was elsewhere. Will add it to the unit test suite - could be that the flags on that level are missing.

And I've noticed an incompatibility. If I overwrite a color change, technically everything after the color change, should be that color
unless there is another color change, or the end of the line, whichever occurs first. (I'm about to write an editor, and this will be impacted
by this.)

You can see this in action, by connecting and without logging on, type
*1# you should get a "red" ACCESS DENIED on the bottom line, then press "*" (to enter a page number). Technically everything after the * should
be GREEN, but it remains RED.

I'll look at it. ty - I need such cases. Doesn't matter if minor or not. But could be that I don't make an extra release - I think I'll look at ansi music as well.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Is there a chance you broke/changed "cursor off" (0x14)? I've noticed
that the cursor is on now, and remains flashing at the end of what was sent. (Cursor on is 0x11).

\x14\n\n\nHelloWorld turns the caret off
\x11 on - from my basic test it works. Do you've longer test string I can try out?

You can see this in action, by connecting and without logging on, type
*1# you should get a "red" ACCESS DENIED on the bottom line, then press "*" (to enter a page number). Technically everything after the * should

Fixed - y took out attribute fill on home. Seems I had a bug in one unit test.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Re: Re: New terminal program
By: Omnibrain to deon on Wed Nov 23 2022 08:29 am

\x14\n\n\nHelloWorld turns the caret off
\x11 on - from my basic test it works. Do you've longer test string I can try out?

Yes, if you connect to my Viewdata BBS, when the connect frame is presented, the cursor is flashing on the bottom line.

Page *999#, the cursor is flashing on the bottom right corner, and 999b (its the top left corner).

In all cases it shouldnt be showing when the frame is rendered, only when input is being asked.

Perhaps it is being shown accidently as a result of other control codes being used?


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

Yes, if you connect to my Viewdata BBS, when the connect frame is presented, the cursor is flashing on the bottom line.

TY - need the exact case. Seems that FF doesn't reset the cursor visible state in Viewdata.

Made a release:
https://github.com/mkrueger/icy_term/releases/tag/0.2.4

Mac builds follow tomorrow. I really need to get the flatpak working. I use Linux as dev machine and I can't really distribute it for linux :(.

I've been working on ANSI music - still have issues finding a good output sound library but it's implemented (but deactivated until 0.3.0).

If you find anything - just report back.

Thanks

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Re: Re: New terminal program
By: Omnibrain to deon on Thu Nov 24 2022 03:00 pm

Howdy,

Made a release:
https://github.com/mkrueger/icy_term/releases/tag/0.2.4

Mac builds follow tomorrow. I really need to get the flatpak working. I use Linux as dev machine and I can't really distribute it for linux :(.

Just checking on the MAC builds - I dont see them on github?


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

Just checking on the MAC builds - I dont see them on github?

My fault maybe didn't commit the edit release. Uploaded a build today. Contains 1-2 mini fixes. I've been working on ANSI music - parsing works quite well but haven't found the perfect playback library yet so it's deactivated by default.

(but can be enabled in the config files and can actually play)

ATM I rewrite the UI in egui - I created a much better terminal canvas with good speed. It'll be awesome when finished.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Re: Re: New terminal program
By: Omnibrain to deon on Mon Nov 28 2022 09:52 am

Howdy,

My fault maybe didn't commit the edit release. Uploaded a build today. Contains 1-2 mini fixes. I've been working on ANSI music - parsing works quite well but haven't found the perfect playback library yet so it's deactivated by default.

Bummer, there are a few issues with this build.

The color on the 1st line is changing again, and on the login page, and the cursor is appearing after a while. Also with the test frame (999a & 999b) three presses of "*" causes the terminal to go into a weird input mode, but I havent debugged if that is just the terminal or the backend. (Testing with my web client doesnt exhibit the same behavour.)

ATM I rewrite the UI in egui - I created a much better terminal canvas with good speed. It'll be awesome when finished.

Look forward to see that though - it already is fast (almost too fast :)


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

The color on the 1st line is changing again, and on the login page, and the cursor is appearing after a while. Also with the test frame (999a & 999b) three presses of "*" causes the terminal to go into a weird input mode, but I havent debugged if that is just the terminal or the backend.

No surprise because I changed the home cmd again to reset attributes in line.

The sequence that does this is: \x1E\x09\n\x11\x1E

x1E is home 09 is right \n is down \x11 is caret off now caret is at (1,1) the next \x1E resets the attribute in line.
Is there a special handling of set attributes in that case?

I have no idea what "*" could change on terminal side but will look into it.

Look forward to see that though - it already is fast (almost too fast :)

No comparison to the fragment shader one - ATM working on that maybe ~1 week and I can make the 1st release :).

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Re: Re: New terminal program
By: Omnibrain to deon on Tue Nov 29 2022 10:13 am

Howdy,

No surprise because I changed the home cmd again to reset attributes in line.

The sequence that does this is: \x1E\x09\n\x11\x1E

These are all cursor movements, and shouldnt change any attributes.

Attributes should only change when printed (and should affect the rendering of the rest of the line).

IE: When you change the current character to green \x1b\x52 (which advances the cursor 1 position - but if hold is in place it repeats the last character) - that should change everything to the right to green and stop when it encovers another color or the end of the line (whichever first).

...deon

...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

These are all cursor movements, and shouldnt change any attributes.

Attributes should only change when printed (and should affect the rendering of the rest of the line).

IE: When you change the current character to green \x1b\x52 (which advances the cursor 1 position - but if hold is in place it repeats the last character) - that should change everything to the right to green
and stop when it encovers another color or the end of the line
(whichever first).

Thought I'm a bit faster with the new UI libary and rendering engine but it takes longer than I hoped so I made a 0.2.5 release.

I changed the behavior to what you suggested seems to work - at least it doesn't break all test pages.

0.3.x will need 1-2 weeks I suppose I really messed up everything this time - but it'll be great.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

It is not finished - esp. the UI needs work. I can need feedback & ideas.

A good selection of transfer protocols.

--- Mystic BBS v1.12 A47 2021/09/29 (Windows/64)
* Origin: Legends of Yesteryear (furmenservices.net:23322) (21:4/102)

Geri Atricks wrote to Omnibrain <=-

It is not finished - esp. the UI needs work. I can need feedback & ideas.

A good selection of transfer protocols.

Most offer at least X/Y/Zmodem and Ymodem-G. Which others do you want to
see? Back in the day, I remember WXmodem, Jmodem, SEAlink, Lynx/Puma, and a few others.



... Computer Hacker wanted. Must have own axe.
--- MultiMail/DOS
* Origin: possumso.fsxnet.nz * SSH:2122/telnet:24/ftelnet:80 (21:4/134)

A good selection of transfer protocols.

Which ones do you miss?

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

On Sunday, December 4th Blue White said...

Most offer at least X/Y/Zmodem and Ymodem-G. Which others do you want to see? Back in the day, I remember WXmodem, Jmodem, SEAlink, Lynx/Puma, and a few others.

HSLink (I believe it was) rocked. Y is OKish for the web, but in generall all of these are pretty horrible for going over telnet / TCP/IP.

We really need a 'modern' protocol (nothing to invent here really) that boards that can't natively support can just drop in (think "httpxfer.exe" or w/e). One of the issues is it really needs to run from a differnet socket connection though (ie: port) as Telnet itself gets in the way of decent transfers.


--
|08 ■ |12NuSkooler |06// |12Xibalba |08- |07"|06The place of fear|07"
|08 ■ |03xibalba|08.|03l33t|08.|03codes |08(|0344510|08/|03telnet|08, |0344511|08/|03ssh|08)
|08 ■ |03ENiGMA 1/2 WHQ |08| |03Phenom |08| |0367 |08| |03iMPURE |08| |03ACiDic
--- ENiGMA 1/2 v0.0.13-beta (linux; x64; 16.16.0)
* Origin: Xibalba -+- xibalba.l33t.codes:44510 (21:1/121)

HSLink (I believe it was) rocked. Y is OKish for the web, but in generall all of these are pretty horrible for going over telnet / TCP/IP.

HSLink was the Torrent protocol for BBSing back in the day. One of my favorites.

Nothing like setting up batches of files to upload and download in your queue. Then selecting HSLink to have them upload/download at the same time, maxing both directions of the connection at the same time.

- Mark
   
--- WWIVToss v.1.52
* Origin: http://www.weather-station.org * Bel Air, MD -USA (21:1/132.0)

HSLink (I believe it was) rocked. Y is OKish for the web, but in genera all of these are pretty horrible for going over telnet / TCP/IP.

HSLink was the Torrent protocol for BBSing back in the day. One of my favorites.

Nothing like setting up batches of files to upload and download in your qu Then selecting HSLink to have them upload/download at the same time, maxin both directions of the connection at the same time.

I was using Bimodem on RemoteAccess which was revelation for us at the time. Upload, Download AND chatting with my users. Woop!

oP!

... THE fIRST sTEP iS tO tAKE oFF tHE cAPS lOCK

--- Mystic BBS v1.12 A47 2021/12/24 (Linux/64)
* Origin: TheForze - bbs.theforze.eu:23 (21:3/126)

HSLink (I believe it was) rocked. Y is OKish for the web, but in generall all of these are pretty horrible for going over telnet / TCP/IP.

We really need a 'modern' protocol (nothing to invent here really) that boards that can't natively support can just drop in (think "httpxfer.exe" or w/e). One of the issues is it really needs to run from a differnet socket connection though (ie: port) as Telnet itself gets in the way of decent transfers.

HAve to concur with the viability of traditional transfers over TCP. Not
sure how ancient software is going to cope with swapping sockets.. the most obvious would be some kind of FTP implementation, in this day and age it
ought to be reasonably trivial.

Spec


*** THE READER V4.50 [freeware]
--- SuperBBS v1.17-3 (Eval)
* Origin: Good Luck and drive offensively! (21:3/101)

Most offer at least X/Y/Zmodem and Ymodem-G. Which others do you want to see? Back in the day, I remember WXmodem, Jmodem, SEAlink, Lynx/Puma,
and a few others.

I think X/Y/Zmodem(-G) are really all that are needed these days.

--- Mystic BBS v1.12 A48 2022/07/11 (Linux/64)
* Origin: m O N T E R E Y b B S . c O M (21:4/173)

HAve to concur with the viability of traditional transfers over TCP. Not sure how ancient software is going to cope with swapping sockets.. the most obvious would be some kind of FTP implementation, in this day and
age it ought to be reasonably trivial.

Yeah - one thing some BBSes do is offer a link to a web download which works well, but that takes a handful of manual steps.

--- Mystic BBS v1.12 A48 2022/07/11 (Linux/64)
* Origin: m O N T E R E Y b B S . c O M (21:4/173)

*** Quoting NuSkooler from a message to Blue White ***

We really need a 'modern' protocol (nothing to invent here really)
that boards that can't natively support can just drop in (think "httpxfer.exe" or w/e). One of the issues is it really needs to run
from a differnet socket connection though (ie: port) as Telnet itself
gets in the way of decent transfers.

I've seen a Mystic board out there that gave you a URL to download your file(s). That worked pretty well and was certanly faster & more reliable
than any (x|y|z)modem protocol over telnet.


Jay

... 3.141592653589793238462643383279502884197169399375105820974944592307816406

--- Telegard v3.09.g2-sp4/mL
* Origin: Northern Realms | tg.nrbbs.net | 289-424-5180 (21:3/110.1)

Re: Re: New terminal program
By: Omnibrain to deon on Fri Dec 02 2022 03:16 pm

Howdy,

Thought I'm a bit faster with the new UI libary and rendering engine but it takes longer than I hoped so I made a 0.2.5 release.

I changed the behavior to what you suggested seems to work - at least it doesn't break all test pages.

Had a quick look at 0.2.5 today - and the rending works really well. I think you got everything covered. Thank you..

There is something minor going on (which I need to check and make sure it's not side) - when you navigate with *...# (where .. is a number), just before clearing of the page to render the requested page, there is garbage printed on the bottom line. It doesnt happen everytime, but it happens at least once every 3-4 times.

I also have that going into strange input mode issue that I've commented on before (especially after rendering the test frame 999a), pressing * * * will make the terminal go into a non responsive state. That still could be in my side, so I'll investigate and let you know if I uncover anything.

Anyway, well done - I think 0.2.5 is good for viewdata :)


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

A good selection of transfer protocols.

Which ones do you miss?

SZModem. You could play tetris with someone at the other end, or chat while the file was transferring.

--- Mystic BBS v1.12 A47 2021/12/24 (Linux/64)
* Origin: Agency BBS | Dunedin, New Zealand | agency.bbs.nz (21:1/101)

SZModem. You could play tetris with someone at the other end, or chat while the file was transferring.

I still have RIPScript on the todo list but I'm surely not going that route including games :). ZModem can be pretty fast - I got DL speed to my line limit - ~5-10MB/s I don't see much of a bottleneck there.

However I limited speed a bit for upload - I broke some BBS connections. Telnet doesn't have much overhead.

Now to Tetris - get Steam or the Epic Games store there are plenty of games out there :). If there is a real transfer protocol missing that can be used somewhere I'll put it on the todo list.

ATM I'm uncertain about the Kermit protocol - but that's the only one. Haven't seen any BBS that supports something beyond xyz & Kermit yet.

General update:
The next thing I do is to make a really, really fast terminal control - I'm currently writing this message with it. It's really awesome (on Linux - need to check other platforms). All I need to do is to improve the shader a bit then I'll release it - switched the UI API, reworked the communication layer - faster than ever and the lightning fast UI control on top. I hope it'll be ready this week. 0.30.0 may contain more bugs (not in the engine itself) - feedback would be nice. Esp. if the direction where it's going is a good or bad one.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

I was using Bimodem on RemoteAccess which was revelation for us at the
time. Upload, Download AND chatting with my users. Woop!

Cool, I don't think I ever tried using that one with WWIV. HSLink worked fantastic, though and I tried to get more people using it to save time on xfers.

Back then it was not easy keeping that balance of being a message board and also file transfer system. I used to have more generous time limits for xfers in the middle of the night so the system could be used more for messages during the day.

I still remember seeing both send/receive modem lights on solid while doing HSLink transfers. It was a huge step forward for transfers back in the day.

- Mark
   
--- WWIVToss v.1.52
* Origin: http://www.weather-station.org * Bel Air, MD -USA (21:1/132.0)

Re: Re: New terminal program
By: Geri Atricks to Omnibrain on Mon Nov 14 2022 03:18 pm

It is not finished - esp. the UI needs work. I can need feedback &
ideas.

A good selection of transfer protocols.

I liked having different file transfer protocols in the 90s using dialup, but these days, file transfer protocols seem largely redundant when using BBSes over telnet & similar. With TCP/IP, thare are already mechanisms built in for data transfers, and there are things like FTP and SCP for file transfers over a network. I've heard that older file transfer protocols like Zmodem and such can actually work worse over an internet connection than they did on dialup.

Nightfox

I liked having different file transfer protocols in the 90s using
dialup, but these days, file transfer protocols seem largely redundant when using BBSes over telnet & similar. With TCP/IP, thare are already mechanisms built in for data transfers, and there are things like FTP
and SCP for file transfers over a network. I've heard that older file transfer protocols like Zmodem and such can actually work worse over an internet connection than they did on dialup.

I believe this is correct - we're doing redundant error correction which I believe can get out of sync between sender/receiver and cause a transfer to fail.

This is a highly reductive explanation but I think is pretty close.

--- Mystic BBS v1.12 A48 2022/07/11 (Linux/64)
* Origin: m O N T E R E Y b B S . c O M (21:4/173)

Re: Re: New terminal program
By: Omnibrain to boraxman on Mon Dec 05 2022 11:24 am

Howdy,

General update:
The next thing I do is to make a really, really fast terminal control - I'm currently writing this message with it. It's really awesome (on Linux - need to check other platforms). All I need to do is to improve the shader a bit then I'll release it - switched the UI API, reworked the communication layer - faster than ever and the lightning fast UI control on top. I hope it'll be ready this week. 0.30.0 may contain more bugs (not in the engine itself) - feedback would be nice. Esp. if the direction where it's going is a good or bad one.

I think there are two things that would be super useful:

* Scroll back - sometimes things fly by and it would be nice to scroll back to see it. Most of us have a GUI, so scrolling with the mouse wheel is easiest, vs a CTRL KEY sequence

* Cut and paste - sometimes its nice to cut text and use it (like a URL), and sometimes its useful to paste in (like output from a command when helping somebody). I think I worked out "cut", but couldnt work out "paste"..

But look forward to seeing what you are creating...



...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

* Scroll back - sometimes things fly by and it would be nice to scroll back to see it. Most of us have a GUI, so scrolling with the mouse wheel is easiest, vs a CTRL KEY sequence

Scroll back should work with the mouse wheel. However next version will have
a scrollbar :). The UI lib I use now is more flexible in scrollbar usage.

* Cut and paste - sometimes its nice to cut text and use it (like a
URL), and sometimes its useful to paste in (like output from a command when helping somebody). I think I worked out "cut", but couldnt work out "paste"..

It's a bit tricky - select text - right click is copy and middle click is paste. The UI lib I used didn't have context menus. I switched the UI lib the next one will have copy & paste in the context menu.
However paste will still work on middle click.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

and SCP for file transfers over a network. I've heard that older fil transfer protocols like Zmodem and such can actually work worse over internet connection than they did on dialup.

I believe this is correct - we're doing redundant error correction which
I believe can get out of sync between sender/receiver and cause a
transfer to fail.

This is a highly reductive explanation but I think is pretty close.

I recently implemented a simple telnet and z-modem transfer protocol. And these old protocols aren't bad. All protocols make data transfer larger. FTP as well. Telnet roughly ~1/256 with Z-Modem it's more complex. It really depends on the implementation and the dialect used.

It's true that the error correction is redundant but calculating a crc takes almost no time. The problem with Z-Modem is that there is no REAL Z-Modem protocol - it's a language with many variants. The implementor has some options how data can be transferred.
There are several header types that can be used, package length can vary and packages can be transferred continuosly or need feedback.
Packages get roughly 9/256 times bigger and the header overhead depends on the package size. But with continuosly transfer it's basically nothing. Longer files ->less header overhead.

I don't know the overhead of FTP but ZModem can be good - it really depends on the implementation and the speed of the system you're connecting to. I can get roughly my line speed ~5MB/s with ZModem. But in general I don't really like/recommend ZModem. I don't have any data yet but YModem-1k-G should be faster than ZModem for internet connections. The data blocks don't get bigger - however the header overhead is 6/1024 - and that's it.

The only benefit ZModem has over YModem is that transfer can be continued.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

This is a highly reductive explanation but I think is pretty close.

One thing about the zmodem in icy term. (esp. in the next version where I optimized my transfer layer) I limited the upload speed there because old systems simply broke if it's too fast. So it's not a good idea for a bbs terminal to provide unlimited zmodem upload speed :).

I'll surely add that to the options at some point. Modern BBS systems shouldn't be limited there - will be interesting to get some benchmarks.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

file transfers over a network. I've heard that older file transfer protocols like Zmodem and such can actually work worse over an internet connection than they did on dialup.

Or worse, not function at all.

Spec


*** THE READER V4.50 [freeware]
--- SuperBBS v1.17-3 (Eval)
* Origin: Good Luck and drive offensively! (21:3/101)

I believe this is correct - we're doing redundant error correction which I believe can get out of sync between sender/receiver and cause a transfer to fail.

You've also got basic differences too, TCP/IP will allow for packets out of sequence, Z-modem cannae handle that..

Spec


*** THE READER V4.50 [freeware]
--- SuperBBS v1.17-3 (Eval)
* Origin: Good Luck and drive offensively! (21:3/101)

SZModem. You could play tetris with someone at the other end, or cha while the file was transferring.

I still have RIPScript on the todo list but I'm surely not going that route including games :). ZModem can be pretty fast - I got DL speed to
my line limit - ~5-10MB/s I don't see much of a bottleneck there.

However I limited speed a bit for upload - I broke some BBS connections. Telnet doesn't have much overhead.

Now to Tetris - get Steam or the Epic Games store there are plenty of games out there :). If there is a real transfer protocol missing that
can be used somewhere I'll put it on the todo list.

ATM I'm uncertain about the Kermit protocol - but that's the only one. Haven't seen any BBS that supports something beyond xyz & Kermit yet.

General update:
The next thing I do is to make a really, really fast terminal control - I'm currently writing this message with it. It's really awesome (on
Linux - need to check other platforms). All I need to do is to improve
the shader a bit then I'll release it - switched the UI API, reworked
the communication layer - faster than ever and the lightning fast UI control on top. I hope it'll be ready this week. 0.30.0 may contain more bugs (not in the engine itself) - feedback would be nice. Esp. if the direction where it's going is a good or bad one.

ZModem is quite good, I even use it time to time when I SSH into another computer. An easy way to do a quick file send from one machine to another. But I have to make sure to escape all control characters, as a large transfer may fail.

For BBS's, sometimes it just doesn't work. IT may pause, give errors, or fail completely, or not even start.

--- Mystic BBS v1.12 A47 2021/12/24 (Linux/64)
* Origin: Agency BBS | Dunedin, New Zealand | agency.bbs.nz (21:1/101)

ZModem is quite good, I even use it time to time when I SSH into another computer. An easy way to do a quick file send from one machine to another. But I have to make sure to escape all control characters, as a large transfer may fail.

At least it's not that bad as some people think. If we would be in 1992 I would vote to replace ZModem with something more efficient - as said YModem-1kG should beat ZModem. And ZModem has many flaws. Like a buggy crc16 implementation and the ability to execute arbitrary commands on the remote computer (didn't implement that for icy_term btw.) and a too complex communication system.

That leads to:

For BBS's, sometimes it just doesn't work. IT may pause, give errors,
or fail completely, or not even start.

Implementing ZModem isn't easy - it's easy to break and as said it's too complex. That's why so many variants exists like 4k/8k, PDZModem, FastZModem or LeechZModem etc. a good protocol should be as simple as possible (but not simpler) IMO.
I don't think it's a suitable protocol for the 90' or now -
however it's working and it's not as bad as ppl think. However it's widespread and anyone can use that. The killer feature back in the days - for me - was that I could continue transfer with that.

I agree that these days the best way for BBS would be to allow http(s) downloads. However a good ZModem implementation can get ok speeds.
But the benefit for the http downloads is that they're on a different channel and users can do BBSing and downloading at the same time. So it would be really nice to have that in these days.

But if there are BBSes out there that use for example the HYDRA file transfer protocol (FSC-0072) I would be tempted to implement that and try it out.
(I don't know much about that protocol - don't ask). Or JModem or one of the bi directional ones. Probably get some real world data.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

But if there are BBSes out there that use for example the HYDRA file transfer protocol (FSC-0072) I would be tempted to implement that and
try it out. (I don't know much about that protocol - don't ask). Or
JModem or one of the bi directional ones. Probably get some real world data.

I recently re-implemented Hydra in daydream but actually had no reasonable way to test it. Let me know if this is of interest.

Honestly rereading through a lot of the file xfer protocol thread, it seems like ZModem is tough due to implementation variability. Perhaps a simplistic protocol with tcp/ip frame optimization and a from-scratch, simple implementation standard, is what's needed.

--- Mystic BBS v1.12 A48 2022/07/11 (Linux/64)
* Origin: m O N T E R E Y b B S . c O M (21:4/173)

I recently re-implemented Hydra in daydream but actually had no
reasonable way to test it. Let me know if this is of interest.

The project is for implementing such old things :). But y I'm interested do you've a test bbs for that?

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

But look forward to seeing what you are creating...

I just released the 0.3.0 version: https://github.com/mkrueger/icy_term/releases/tag/0.3.0

Rewrote the UI layer - it's way faster than before and has some features it hadn't before. Scrollbars/Context menu/Output with scanlines.

Thought I need more time to get the shader code to that level but managed today the biggest features. There is still much to do but I want feedback if I'm doing going in a good direction.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

The project is for implementing such old things :). But y I'm interested do you've a test bbs for that?

I don't but can walk you through setting it up if you like.

https://github.com/ryanfantus/daydream if you'd like to look for now.

--- Mystic BBS v1.12 A48 2022/07/11 (Linux/64)
* Origin: m O N T E R E Y b B S . c O M (21:4/173)

Re: Re: New terminal program
By: Omnibrain to deon on Tue Dec 06 2022 12:14 am

Howdy,

Scroll back should work with the mouse wheel. However next version will have a scrollbar :). The UI lib I use now is more flexible in scrollbar usage.

Hmm it doesnt for me - perhaps because I'm on a MAC?

It's a bit tricky - select text - right click is copy and middle click is paste. The UI lib I used didn't have context menus. I switched the UI lib the next one will have copy & paste in the context menu.
However paste will still work on middle click.

MAC's dont have a "middle click" - is there another key?


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

Re: Re: New terminal program
By: Omnibrain to deon on Tue Dec 06 2022 03:56 pm

Rewrote the UI layer - it's way faster than before and has some features it hadn't before. Scrollbars/Context menu/Output with scanlines.

Howdy, just tried this - and can confirm copy/paste work well :)

One thing I noticed, is that "flashing" only works on mouse movement... If the mouse stops, then the flashing stops (and if it is off at the time, you cant see it).


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

I don't but can walk you through setting it up if you like.

Ok found a daydream BBS supporting hydra and smodem - interesting. I may look at it. Thanks.

--- Mystic BBS v1.12 A48 2022/07/15 (Linux/64)
* Origin: 2o fOr beeRS bbs>>>20ForBeers.com:1337 (21:2/150)

NuSkooler wrote to Blue White <=-

HSLink (I believe it was) rocked. Y is OKish for the web, but in
generall all of these are pretty horrible for going over telnet /
TCP/IP.

Yes, HSLink did rock. It was supposed to do what BiModem was supposed to do, but was a lot easier to set up!



... How do you tell when you're out of invisible ink?
--- MultiMail/DOS
* Origin: possumso.fsxnet.nz * SSH:2122/telnet:24/ftelnet:80 (21:4/134)

On Sunday, December 4th Weatherman muttered...

HSLink was the Torrent protocol for BBSing back in the day. One of my favorites.
Nothing like setting up batches of files to upload and download in your queue. Then selecting HSLink to have them upload/download at the same time, maxing both directions of the connection at the same time.

Yep, this thing blew my mind when I was a youth hehe





--
|08 ■ |12NuSkooler |06// |12Xibalba |08- |07"|06The place of fear|07"
|08 ■ |03xibalba|08.|03l33t|08.|03codes |08(|0344510|08/|03telnet|08, |0344511|08/|03ssh|08)
|08 ■ |03ENiGMA 1/2 WHQ |08| |03Phenom |08| |0367 |08| |03iMPURE |08| |03ACiDic
--- ENiGMA 1/2 v0.0.13-beta (linux; x64; 16.16.0)
* Origin: Xibalba -+- xibalba.l33t.codes:44510 (21:1/121)

On Sunday, December 4th Warpslide said...

I've seen a Mystic board out there that gave you a URL to download your file(s). That worked pretty well and was certanly faster & more reliable
than any (x|y|z)modem protocol over telnet.

Yep, ENiGMA1/2 the software I maintain also allows for this. You can even batch and the system will add all the files you select to a single streamed zip file. I'd love to be able to do it in terminal though, maybe just for nostalgia :D



--
|08 ■ |12NuSkooler |06// |12Xibalba |08- |07"|06The place of fear|07"
|08 ■ |03xibalba|08.|03l33t|08.|03codes |08(|0344510|08/|03telnet|08, |0344511|08/|03ssh|08)
|08 ■ |03ENiGMA 1/2 WHQ |08| |03Phenom |08| |0367 |08| |03iMPURE |08| |03ACiDic
--- ENiGMA 1/2 v0.0.13-beta (linux; x64; 16.16.0)
* Origin: Xibalba -+- xibalba.l33t.codes:44510 (21:1/121)

One thing I noticed, is that "flashing" only works on mouse movement...
If the mouse stops, then the flashing stops (and if it is off at the
time, you cant see it).

Y can confirm that the update call is broken. Damn used to work may've to do with my sixel updates :(. Unfortunately I can't cover that with a unit test.

btw. I test on mac as well but I use a logitech mouse so I've had a middle click there. But the context menu should fix that.

--- Mystic BBS v1.12 A48 2022/07/15 (Linux/64)
* Origin: 2o fOr beeRS bbs>>>20ForBeers.com:1337 (21:2/150)

One thing I noticed, is that "flashing" only works on mouse movement...
If the mouse stops, then the flashing stops (and if it is off at the
time, you cant see it).

I consider that big enough for a 0.3.1 bugfix release:

https://github.com/mkrueger/icy_term/releases/tag/0.3.1

You're basically my Q&A team :).

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

Re: Re: New terminal program
By: Omnibrain to deon on Tue Dec 06 2022 11:02 pm

I consider that big enough for a 0.3.1 bugfix release:

https://github.com/mkrueger/icy_term/releases/tag/0.3.1

You're basically my Q&A team :).

Working well - awesome.

Hey with the scroll back thing, it looks like the scroll back buffer is emptied when a clear screen is sent. Any chance of changing that, or making it a configuration item?

I see the scrolling is working when there is a long message, but not when a clear screen was sent. Sometimes when I scan new and fat finger a key, I loose a message (the most time I use scroll back), but if the next message displays (and it cleared the screen), I've lost it :(

Also, when replying to this message (in SBBS using iCY), the screen was all messed up (I use the FSEditor in SBBS), and at first display (which is the quote display), many things were missing. I had to leave the quote display and re-enter it in order to see things properly.


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

Re: Re: New terminal program
By: Omnibrain to deon on Tue Dec 06 2022 03:56 pm

Howdy,

Rewrote the UI layer - it's way faster than before and has some features it hadn't before. Scrollbars/Context menu/Output with scanlines.

This new version 0.3.x the font is certainly "less clear" than 0.2.x which was slightly less clear than SyncTerm.

In fact the 0.3.x the font colors are almost blurry (and I'm not using the "CRT Effect") - so wondering if there is anything you can do for that?

I'm guessing its because the font is being scaled to a larger screen (I have a 27" monitor)?


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

Hey with the scroll back thing, it looks like the scroll back buffer is emptied when a clear screen is sent. Any chance of changing that, or making it a configuration item?

Need to think about this how this can be done properly. It's sometimes not easy because it can get very messy when using some ansi features like scrolling - would potentially mess up the buffer. It's something where I need to do some experimenting. Clearing everything on clear screen is certainly safe.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

This new version 0.3.x the font is certainly "less clear" than 0.2.x
which was slightly less clear than SyncTerm.

I'll certainly add options for that. ATM I experiment around with a CRT like mode (see the checkbox on top) and maybe adding some different shader effects.

But I'll bring back a sharper/cleaner look back because I personally prefer that. SyncTerm cheats a bit they're scaling 1x, 2x, 4x so they don't need any kind of interpolation :).

I'm now using opengl & shaders directly for rendering the terminal and there are many features to play around with. At some point I need to do terminal settings and there I'll throw that in.

--- Mystic BBS v1.12 A48 2022/07/15 (Windows/64)
* Origin: Altair IV BBS (21:2/147)

I'm now using opengl & shaders directly for rendering the terminal and there are many features to play around with. At some point I need to do terminal settings and there I'll throw that in.

Separately, how tough would it be to have linux builds? I can probably build myself but just asking if you have any ci/cd build type thing :)

--- Mystic BBS v1.12 A48 2022/07/11 (Linux/64)
* Origin: m O N T E R E Y b B S . c O M (21:4/173)

At least it's not that bad as some people think. If we would be in 1992
I would vote to replace ZModem with something more efficient - as said YModem-1kG should beat ZModem. And ZModem has many flaws. Like a buggy crc16 implementation and the ability to execute arbitrary commands on
the remote computer (didn't implement that for icy_term btw.) and a too complex communication system.

That leads to:

For BBS's, sometimes it just doesn't work. IT may pause, give errors or fail completely, or not even start.

Implementing ZModem isn't easy - it's easy to break and as said it's too complex. That's why so many variants exists like 4k/8k, PDZModem, FastZModem or LeechZModem etc. a good protocol should be as simple as possible (but not simpler) IMO.
I don't think it's a suitable protocol for the 90' or now -
however it's working and it's not as bad as ppl think. However it's widespread and anyone can use that. The killer feature back in the days
- for me - was that I could continue transfer with that.

I agree that these days the best way for BBS would be to allow http(s) downloads. However a good ZModem implementation can get ok speeds.
But the benefit for the http downloads is that they're on a different channel and users can do BBSing and downloading at the same time. So it would be really nice to have that in these days.

But if there are BBSes out there that use for example the HYDRA file transfer protocol (FSC-0072) I would be tempted to implement that and
try it out. (I don't know much about that protocol - don't ask). Or
JModem or one of the bi directional ones. Probably get some real world data.

I don't use the HTTP downloads because copy and paste in syncterm isn't that great. If you are willing to get up telnet capability, you can have a client which can download and still browse a BBS at the same time.

--- Mystic BBS v1.12 A47 2021/12/24 (Linux/64)
* Origin: Agency BBS | Dunedin, New Zealand | agency.bbs.nz (21:1/101)

Separately, how tough would it be to have linux builds? I can probably build myself but just asking if you have any ci/cd build type thing :)

It's extremely complex to get things done on linux. I had deb builds but no one could use them. I can make a flatpak which nobody can install.

I use linux for development but how to deploy software on linux remains a complete mystery for me. The easiest deployment is windows. There is an exe which runs. Getting MacOS build and signed was a nightmare but it works now.

Linux hates me and doesn't want that I deploy software there :(. Which is sad because it's my daily driver.

--- Mystic BBS v1.12 A48 2022/07/15 (Linux/64)
* Origin: 2o fOr beeRS bbs>>>20ForBeers.com:1337 (21:2/150)

I don't use the HTTP downloads because copy and paste in syncterm isn't that great. If you are willing to get up telnet capability, you can
have a client which can download and still browse a BBS at the same time.

I've no issues with copy&paste in icy_term :). I've implemented only a very, very simple telnet protocol and I don't think I'll use any telnet features. It's simply not the right layer for that. Downloads should work with ssh/raw/telnet connections.

However at some point I'll look at protocols again. I now just want to finish what I already have. ATM that would be ANSI music where I need a good output library. The other stuff is just an extra. But I value any feedback and try to get things into it.

But I want to start an ANSI drawing tool - I made one in the 90' (MysticDraw) the last update was 2003 or so where I translated it to SDL. Now 20 years later I'll update it to rust. Never got feedback for the old one and don't think I get feedback for the new one but I maintain it :).

However I may rename it to IcyDraw - didn't know about MysticBBS back then.

--- Mystic BBS v1.12 A48 2022/07/15 (Linux/64)
* Origin: 2o fOr beeRS bbs>>>20ForBeers.com:1337 (21:2/150)

A good selection of transfer protocols.

Which ones do you miss?

Back in the 'analog' days I also had Smodem, stable protcol and up/down/chat
at the same time. Liked it very much!

gTX!
s

--- Mystic BBS v1.12 A48 2022/07/15 (Linux/64)
* Origin: .+Sound&Demos+. - <sndgarden.bbsfreak.nl:*hidden*> (21:1/116)

Hey with the scroll back thing, it looks like the scroll back buffer emptied when a clear screen is sent. Any chance of changing that, or making it a configuration item?

Need to think about this how this can be done properly. It's sometimes
not easy because it can get very messy when using some ansi features
like scrolling - would potentially mess up the buffer. It's something where I need to do some experimenting. Clearing everything on clear
screen is certainly safe.

Maybe a stupid idea, but what about a feature to 'log' it to a video format ?

gtX!

--- Mystic BBS v1.12 A48 2022/07/15 (Linux/64)
* Origin: .+Sound&Demos+. - <sndgarden.bbsfreak.nl:*hidden*> (21:1/116)

Maybe a stupid idea, but what about a feature to 'log' it to a video format ?

I can't imagine to do things better in that area than obs could. Having the scrollback buffer or buffers has the opportunity to copy text or graphics out of them.

A plain video wouldn't have any benefit over obs. It would just add size to the executable, loading speed etc. without much benefit.

At least atm I don't see any benefit but I may be wrong.

--- Mystic BBS v1.12 A48 2022/07/15 (Linux/64)
* Origin: 2o fOr beeRS bbs>>>20ForBeers.com:1337 (21:2/150)

Linux hates me and doesn't want that I deploy software there :(. Which
is sad because it's my daily driver.

Hm, that's unfortunate. Are there simple build instructions to just make a local binary? Maybe I can help if so.

--- Mystic BBS v1.12 A48 2022/07/11 (Linux/64)
* Origin: m O N T E R E Y b B S . c O M (21:4/173)

Hm, that's unfortunate. Are there simple build instructions to just make
a local binary? Maybe I can help if so.

Install the rust toolchain:
https://www.rust-lang.org/tools/install
Get the source code from github:
https://github.com/mkrueger/icy_term

cargo run --release
Or if you just want to build
cargo build --release

Building rust applications usually is easy if they don't have local references exectubale is in:
target/release/icy_term

In build/create_flatpak.sh is a script where I've tried to create a flatpak. But installing that is almost impossible - I tried it.

Creating a .deb is doable too.
"cargo install cargo-bundle"
And then use "cargo bundle --release"
will create a .deb file target/release/bundle/deb/icy_term_0.3.1_amd64.deb

But so far no one could install that. But my testgroup was n=1. So I don't bother with building linux stuff anymore. It's not that this kind of application attracts many users for testing stuff.

But I've tried to get something build for linux users.

--- Mystic BBS v1.12 A48 2022/07/15 (Linux/64)
* Origin: 2o fOr beeRS bbs>>>20ForBeers.com:1337 (21:2/150)

maybe you need to switch the toolchain to nightly:

rustup toolchain install nightly
rustup default nightly

--- ENiGMA 1/2 v0.0.13-beta (linux; x64; 16.16.0)
* Origin: Xibalba -+- xibalba.l33t.codes:44510 (21:1/121)

Omnibrain wrote to boraxman <=-

I don't use the HTTP downloads because copy and paste in syncterm isn't that great. If you are willing to get up telnet capability, you can
have a client which can download and still browse a BBS at the same time.

I've no issues with copy&paste in icy_term :). I've implemented only a very, very simple telnet protocol and I don't think I'll use any telnet features. It's simply not the right layer for that. Downloads should
work with ssh/raw/telnet connections.

However at some point I'll look at protocols again. I now just want to finish what I already have. ATM that would be ANSI music where I need a good output library. The other stuff is just an extra. But I value any feedback and try to get things into it.

But I want to start an ANSI drawing tool - I made one in the 90' (MysticDraw) the last update was 2003 or so where I translated it to
SDL. Now 20 years later I'll update it to rust. Never got feedback for
the old one and don't think I get feedback for the new one but I
maintain it :).

However I may rename it to IcyDraw - didn't know about MysticBBS back then.

Is there a Linux binary? I've downloaded the source code, but not sure how to compile. There is a script to build a flatpak, but I don't have flatpak installed.

___ MultiMail/Linux v0.52

--- Mystic BBS/QWK v1.12 A47 2021/12/24 (Linux/64)
* Origin: Agency BBS | Dunedin, New Zealand | agency.bbs.nz (21:1/101)

Separately, how tough would it be to have linux builds? I can probably build myself but just asking if you have any ci/cd build type thing :)

It's extremely complex to get things done on linux. I had deb builds
but no one could use them. I can make a flatpak which nobody can
install.

I use linux for development but how to deploy software on linux remains
a complete mystery for me. The easiest deployment is windows. There is
an exe which runs. Getting MacOS build and signed was a nightmare but
it works now.

Linux hates me and doesn't want that I deploy software there :(. Which
is sad because it's my daily driver.

That is one of the things that kind of sucks about Linux, though it isn't really specifically a Linux thing. It's more the toolchains and languages. I've written some programs in C and C++ and its not too difficult because of the build system used and because the dependencies are handled by the distros package manager. With Rust and languages where they have their own package management system, things get far more complex.

I could give building a DEB a try if you provide the compiled binary and package description.

___ MultiMail/Linux v0.52

--- Mystic BBS/QWK v1.12 A47 2021/12/24 (Linux/64)
* Origin: Agency BBS | Dunedin, New Zealand | agency.bbs.nz (21:1/101)

maybe you need to switch the toolchain to nightly:

rustup toolchain install nightly
rustup default nightly

Ok, cool, may give this a shot after trying th elatest build here.

--- Mystic BBS v1.12 A48 2022/07/11 (Linux/64)
* Origin: m O N T E R E Y b B S . c O M (21:4/173)

On Saturday, December 10th boraxman was heard saying...

Is there a Linux binary? I've downloaded the source code, but not sure how to compile. There is a script to build a flatpak, but I don't have flatpak installed. ___ MultiMail/Linux v0.52

I added some build instructions here:
https://github.com/mkrueger/icy_term

I've published a .deb:
https://github.com/mkrueger/icy_term/releases/tag/0.3.2

Some feedback if it works would be nice. One user said in an older version that it doesn't work for some reason. I suspected a different glibc version but didn't hear back.

--- ENiGMA 1/2 v0.0.13-beta (linux; x64; 16.16.0)
* Origin: Xibalba -+- xibalba.l33t.codes:44510 (21:1/121)

I added some build instructions here:
https://github.com/mkrueger/icy_term

Hey Omni, I tried building last night and haven't done any troubleshooting yet but figured I'd let you know how it played out on a rather bog standard Arch install (which means probably a bit bleeding edge): https://pastebin.com/1cPEbdms

--- Mystic BBS v1.12 A48 2022/07/11 (Linux/64)
* Origin: m O N T E R E Y b B S . c O M (21:4/173)

On Saturday, December 10th boraxman was heard saying...

Is there a Linux binary? I've downloaded the source code, but not su how to compile. There is a script to build a flatpak, but I don't ha flatpak installed. ___ MultiMail/Linux v0.52

I added some build instructions here:
https://github.com/mkrueger/icy_term

I've published a .deb: https://github.com/mkrueger/icy_term/releases/tag/0.3.2

Some feedback if it works would be nice. One user said in an older
version that it doesn't work for some reason. I suspected a different glibc version but didn't hear back.

Thank you. I got it compiled and it worked. I like the CRT emulation, I think this is perhaps the biggest feature for me.

I'll keep trying it out. I did get a little confused with the icons at the top when connected. There were no tooltips or text to say what they were. Personally, I prefer text menus over trying to figure out what the icon means.

Also, I'm having trouble connecting to minds eye, I'll troubleshoot further and let you know. It's not Minds Eye BBS as I can connect with Syncterm.

The choice if rust was an interesting choice. The source tree after compilation was 1.6GB!!! That is not including files in the ~/.cargo directory that were also added (its 550MB). The binary itself after stripping was 29MB, which is OK.

Thanks!

--- Mystic BBS v1.12 A47 2021/12/24 (Linux/64)
* Origin: Agency BBS | Dunedin, New Zealand | agency.bbs.nz (21:1/101)

Hey Omni, I tried building last night and haven't done any
troubleshooting yet but figured I'd let you know how it played out on a rather bog standard Arch install (which means probably a bit bleeding edge): https://pastebin.com/1cPEbdms

"git pull" should fix that issue.
I recommend doing a "cargo update" as well but that bug should be fixed by updating the source.

... I do not fear computers. I fear the lack of them.

--- Mystic BBS v1.12 A48 2022/07/15 (Linux/64)
* Origin: 2o fOr beeRS bbs>>>20ForBeers.com:1337 (21:2/150)

On Sunday, December 11th boraxman was heard saying...

I'll keep trying it out. I did get a little confused with the icons at the top when connected. There were no tooltips or text to say what they were. Personally, I prefer text menus over trying to figure out what the icon means.

I can add tooltips. Icons have the benefit of beeing easier to recognize. Once I get a crate for i18n I'll may switch to text.


boraxman around Sunday, December 11th...

Also, I'm having trouble connecting to minds eye, I'll troubleshoot further and let you know. It's not Minds Eye BBS as I can connect with Syncterm.

oops... I don't add :23 anymore - I suspect that's missing in your URL. Add that and it should work. I'll fix that.


boraxman around Sunday, December 11th...

The choice if rust was an interesting choice. The source tree after compilation was 1.6GB!!! That is not including files in the ~/.cargo directory that were also added (its 550MB). The binary itself after stripping was 29MB, which is OK.

Rust binaries are not well known for beeing small. But for beeing secure and fast. The release version on my system is ~11MB fyi.

--- ENiGMA 1/2 v0.0.13-beta (linux; x64; 16.16.0)
* Origin: Xibalba -+- xibalba.l33t.codes:44510 (21:1/121)

"git pull" should fix that issue.
I recommend doing a "cargo update" as well but that bug should be fixed
by updating the source.

Confirmed! Thanks.

--- Mystic BBS v1.12 A48 2022/07/11 (Linux/64)
* Origin: m O N T E R E Y b B S . c O M (21:4/173)

I just released the 0.3.0 version: https://github.com/mkrueger/icy_term/releases/tag/0.3.0

Rewrote the UI layer - it's way faster than before and has some features it hadn't before. Scrollbars/Context menu/Output with scanlines.

Every release is better and better! Really enjoying your work and Icy!

Thank you for this!

... A SQL query walks into a bar and sees two tables. Asks: 'Can I join you?'

--- Mystic BBS v1.12 A48 2022/07/15 (Raspberry Pi/32)
* Origin: CaNerDUH BBS (21:4/151)

I'll keep trying it out. I did get a little confused with the icons the top when connected. There were no tooltips or text to say what t were. Personally, I prefer text menus over trying to figure out what icon means.

I can add tooltips. Icons have the benefit of beeing easier to
recognize. Once I get a crate for i18n I'll may switch to text.

My personal opinion is that icons are more confusing, unless a very tried and true icon us used (floppy disk for save, scissors for cut, etc). In theory they are supposed to be easier, in practice, not. I hate having to over over an icon to find what it is for.

Also, I'm having trouble connecting to minds eye, I'll troubleshoot further and let you know. It's not Minds Eye BBS as I can connect wi Syncterm.

oops... I don't add :23 anymore - I suspect that's missing in your URL. Add that and it should work. I'll fix that.

boraxman around Sunday, December 11th...

Yep, that fixed it.

The choice if rust was an interesting choice. The source tree after compilation was 1.6GB!!! That is not including files in the ~/.cargo directory that were also added (its 550MB). The binary itself after stripping was 29MB, which is OK.

Rust binaries are not well known for beeing small. But for beeing secure and fast. The release version on my system is ~11MB fyi.

Yeah, I find rust seems to have a lot of zealots who think that rewriting things in rust is the way to go.

I briefly looked at the language, and it seemed to have promise, but the issues with it, such as those I mentioned led me to stick with C++.

I think 11MB for a program which has CRT emulation is decent.

--- Mystic BBS v1.12 A47 2021/12/24 (Linux/64)
* Origin: Agency BBS | Dunedin, New Zealand | agency.bbs.nz (21:1/101)

On Monday, December 12th boraxman muttered...

My personal opinion is that icons are more confusing, unless a very tried and true icon us used (floppy disk for save, scissors for cut, etc). In theory they are supposed to be easier, in practice, not. I hate having to over over an icon to find what it is for.

I added hover tooltips now. I thought the icons are self explaining - but maybe they're not. I'll think about that & investigate more. Maybe try out text again.



Twas Monday, December 12th when boraxman said...

oops... I don't add :23 anymore - I suspect that's missing in your

URL.

Add that and it should work. I'll fix that.

Yep, that fixed it.

Fixed in newly released 0.4.0 - broke that in one of my refactorings :/. I need to put more effort in the unit tests atm they're badly broken because of all the rewrites.

Yeah, I find rust seems to have a lot of zealots who think that rewriting things in rust is the way to go.

Difficult - but I think it's the way to go too if the project is worth it. Most projects don't need that but in some cases it can make sense. Almost all C libraries have plenty vulnerabilities that they won't when written in rust. However rust is no magic bullet but it helps. Personally I won't go back to c/c++ again.
(Except for very small ardunio projects)

In a bigger Desktop application for example GIMP I don't see much sense in that. But for let's say web browsers or security related software it can make sense. I can understand ppl that want to rewrite everything. We would surely live in a better world however the cost/gain ratio should be considered.
And in 5y a new language may pop up that's better than rust - do we rewrite again?

On Monday, December 12th boraxman said...

I briefly looked at the language, and it seemed to have promise, but the issues with it, such as those I mentioned led me to stick with C++.

I wanted to learn it for a long time. I started this year and now I regret not having leaned it sooner. It's a very hard language to learn. I'm surprised that it's so widespread. But y it has weak spots and issues but it's the best language for system programming I know so far.





--- ENiGMA 1/2 v0.0.13-beta (linux; x64; 16.16.0)
* Origin: Xibalba -+- xibalba.l33t.codes:44510 (21:1/121)

On Saturday, December 10th blade was heard saying...

Every release is better and better! Really enjoying your work and Icy!

Thank you for this!

Thanks :). I just released 0.4.0

https://github.com/mkrueger/icy_term/releases/tag/0.4.0

Still some work to do but I can see 1.0.0 :).

--- ENiGMA 1/2 v0.0.13-beta (linux; x64; 16.16.0)
* Origin: Xibalba -+- xibalba.l33t.codes:44510 (21:1/121)

My personal opinion is that icons are more confusing, unless a very t and true icon us used (floppy disk for save, scissors for cut, etc). theory they are supposed to be easier, in practice, not. I hate havi to over over an icon to find what it is for.

I added hover tooltips now. I thought the icons are self explaining -
but maybe they're not. I'll think about that & investigate more. Maybe
try out text again.

It could just be me being daft. Some were obvious, the upload/download ones I *suspected* were upload and download. The hangup one, I *suspected* could be to close but wasn't sure until I pressed it.

But I'm a CLU/TUI kind of person, I don't use many modern "apps". For telnet I'll either use Telnet itself, syncterm for BBSs or occasionally Qodem, all of which are text based. As the terminal itself is text based I feel that a text based interface makes more sense. I use Emacs, a terminal, more older GUI programs which have menus. I don't use phone apps much, so what might be a convention elsewhere is something I'm not familiar with.

--- Mystic BBS v1.12 A47 2021/12/24 (Linux/64)
* Origin: Agency BBS | Dunedin, New Zealand | agency.bbs.nz (21:1/101)

It is not finished - esp. the UI needs work. I can need feedback & id

A good selection of transfer protocols.

Most offer at least X/Y/Zmodem and Ymodem-G. Which others do you want to see? Back in the day, I remember WXmodem, Jmodem, SEAlink, Lynx/Puma,
and a few others.

I think that pretty much covers it. In fact, I don't even remember a WXmodem, but I do remember seeing those others on a few boards.

--- Mystic BBS v1.12 A47 2021/09/29 (Windows/64)
* Origin: Legends of Yesteryear (furmenservices.net:23322) (21:4/102)

Geri Atricks wrote to Blue White <=-

A good selection of transfer protocols.

Most offer at least X/Y/Zmodem and Ymodem-G. Which others do you want to see? Back in the day, I remember WXmodem, Jmodem, SEAlink, Lynx/Puma,
and a few others.

I think that pretty much covers it. In fact, I don't even remember a WXmodem, but I do remember seeing those others on a few boards.

WXmodem stood for Windowed Xmodem. It may have been exclusive to GT Power BBSes.


... Heisenberg may have slept here.
--- MultiMail/DOS
* Origin: possumso.fsxnet.nz * SSH:2122/telnet:24/ftelnet:80 (21:4/134)

On 12/8/22 07:46, Omnibrain wrote:

Hm, that's unfortunate. Are there simple build instructions to just make
a local binary? Maybe I can help if so.

Install the rust toolchain:
https://www.rust-lang.org/tools/install
Get the source code from github:
https://github.com/mkrueger/icy_term

Very cool.. noticing eframe supports web embedding too... don't suppose
you have instructions for a web build/embed?

Would be very cool to be able to use this with a wss connection with
rlogin straight into doors.
--
Michael J. Ryan - tracker1@roughneckbbs.com
--- SBBSecho 3.15-Linux
* Origin: Roughneck BBS - roughneckbbs.com (21:3/149)

Very cool.. noticing eframe supports web embedding too... don't suppose you have instructions for a web build/embed?

Would be very cool to be able to use this with a wss connection with rlogin straight into doors.

Not yet - I've tried to build it for wasm but one of the deps didn't work haven't looked deeply into it. But y in theory it should run.

--- Mystic BBS v1.12 A48 2022/07/15 (Linux/64)
* Origin: 2o fOr beeRS bbs>>>20ForBeers.com:1337 (21:2/150)

Re: Re: New terminal program
By: Omnibrain to boraxman on Wed Dec 07 2022 08:08 am

But I want to start an ANSI drawing tool - I made one in the 90' (MysticDraw) the last update was 2003 or so where I translated it to SDL. Now 20 years later I'll update it to rust. Never got feedback for the old one and don't think I get feedback for the new one but I maintain it :).

Not sure if you're aware, but there's a fork of MysticDraw here: https://gitlab.synchro.net/main/sbbs/-/tree/master/src/syncdraw

Someone recently asked in one of the Synchronet-support conferences about the lack of certain line-drawing character/function-key combos too (coincidentally):
https://web.synchro.net/?page=001-forum.ssjs&sub=sync&thread=48757
--
digital man (rob)

Synchronet/BBS Terminology Definition #86:
UART = Universal Asynchronous Receiver/Transmitter
Norco, CA WX: 50.5°F, 85.0% humidity, 0 mph NNE wind, 0.00 inches rain/24hrs --- SBBSecho 3.20-Linux
* Origin: Vertrauen - [vert/cvs/bbs].synchro.net (21:1/183)

Re: Re: New terminal program
By: Omnibrain to boraxman on Sun Dec 11 2022 06:35 am

Howdy,

Fixed in newly released 0.4.0 - broke that in one of my refactorings :/. I

Just wondering if there is any more releases coming? ;)

I'm liking 0.4.0, however, the screen display is "small", ie: the terminal is in about a quarter of the window. Resizing the window still has the terminal in about a quarter of it :(

(This is on an Intel Mac - havent tried my M1.)


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)

Hi

Just wondering if there is any more releases coming? ;)

Sure, if I find urgent bugs. The next thing I'll do is working on an ansi drawing tool :)

I'm liking 0.4.0, however, the screen display is "small", ie: the
terminal is in about a quarter of the window. Resizing the window still has the terminal in about a quarter of it :(

Ok that's "urgent" can you post such things in github under issues - best with screenshots. On my PC and Mac resizing works.

--- Mystic BBS v1.12 A48 (Linux/64)
* Origin: 2o fOr beeRS bbs>>>20ForBeers.com:1337 (21:2/150)

https://gitlab.synchro.net/main/sbbs/-/tree/master/src/syncdraw

Haven't seen it - thanks for the info. Will look at it. I'm still planing to work on the ansi drawing tool but I kind of busted my PC by buying an AMD card - hopefully in april the linux drivers are stable enough to continue.

--- Mystic BBS v1.12 A48 (Linux/64)
* Origin: 2o fOr beeRS bbs>>>20ForBeers.com:1337 (21:2/150)

Re: Re: New terminal program
By: Omnibrain to deon on Sat Feb 18 2023 08:38 am

Ok that's "urgent" can you post such things in github under issues - best with screenshots. On my PC and Mac resizing works.

Done, I opened an issue.

Thanks :)


...δεσ∩
--- SBBSecho 3.15-Linux
* Origin: I'm playing with ANSI+videotex - wanna play too? (21:2/116)