Reading Programming

This post was deleted by request of the author.

This post was deleted by request of the author.

RCD,

Do you code? Will you be able to help out with the server and/or clients?

You bet, I do! I can code in Java and C++ for the *nix and windows machines, I have done a lot of web servives programming using Java clients, and .NET servers and clients written in C#, recently. So, I can definitely help in writing the clients on a .NET environment.

Here's a suggestion: We can use SOAP/XML messages to communicate between the clients and the servers. If security is an issue, we can incorporate encryption in the SOAP messages using the latest WS-Security and WS-Transactions standards. If it's only get messages, and the users won't be allowed to modify anything on this host machine, then we probably don't need to care about security for now.

Right now, I am about to start investigating how to pass messages in web services using protocols which are not standards but may be more efficient performance-wise, for example - DIME - a protocol which can be used to send SOAP messages over TCP/IP.

Love is blind, but marriage is a real eye opener.

This post was deleted by request of the author.

This post was edited by frank on Aug 31, 2002.

Hmm, I don't think I'd agree to install Linux on it. :-) It took me long enough to get it up and running on FreeBSD, and I wouldn't want to do all that work again. The box is currently used as a samba, NAT and web server for our in-house LAN. I.e. if it stops working, 3 people are mad at me because they're off the Internet.
I don't see any problems tho, as many Linux applications need minimal changes in order to comile on FreeBSD. Plus FreeBSD can run Linux binaries. :-)
So, unless we're doing very OS-specific, super-tricky stuff, I don't see why we should not be able to use the same code on Linux and FreeBSD.

"*sigh* Some men are really hard to manipulate!" - Orchid

This post was deleted by request of the author.

This post was deleted by request of the author.

...I've managed to contact Jaz in the meantime and asked him for his opinion on this topic. I haven't got a definitive reply yet, but he has asked for more details which I've sent him this evening. I do have the impression tho that he wouldn't like the idea of too many people playing around with NAO's specs or code. :-) I will post more info if/when Jaz tells me more, or maybe he'll write something himself when he and Ettena are back from their holidays.

"*sigh* Some men are really hard to manipulate!" - Orchid

'kay I'm trying to give a short overview here. Just the administrative stuff, don't expect a great story. :-)

NAO integration
As you probably know, our beloved site creator and admin, Jaz, is strolling around in France at the moment. I've had the chance to explain the situation to him, but haven't really got an answer yet (probably his cellphone's battery died, or he's left speechless or something). Since any interaction with the NAO web site must be programmed by him, we can't do very much right now. IMHO it's only fair to wait for him to return anyway.
AFAIK he wouldn't like it too much if the pager started replacing the NAO website, and I agree with him on that. Maybe we can talk him into writing a script that outputs NAO-related data (current users, new posts) in a machine-readable form. I wouldn't give the pager too many functions the web interface already provides.
Other opinions? :-)

The Server (hardware)
Since nobody besides me has offered one, it looks as if we're going to have a FreeBSD server with a (mostly permanent) cable modem connection. I will give away free shell accounts as soon as (if ever) there's a definitive project team.

The Server (software)
So far 3 people (frank, RCD and me) showed interest in (co-)developing the server. I do have some experience on this field, but except for plain C I suck at pretty much any programming language mentioned here.
Question: What programming language should the server be written in?

The Clients (software)
There will most likely be two clients, one for Windows and one for Linux (and probably other *nixes as well). My suggestion is that we build a platform-independent code library which is used by two platform-dependent GUI's. It might be a good idea to use the same programming language for both the server and the client library.
The only GUI stuff I can do is Windows/Delphi. I do own a copy of C++ Builder but have never really gotten into it (C++ looks just too clumsy for me).
Questions:
Who would like to (co-)develop the GUI stuff, and for which platform and programming language?
If there was a Delphi GUI, how easy/hard would it be to port it to Kylix? (I have downloaded Kylix yesterday but can't test it right now due to lack of a Linux box.)

Who would like to be involved, and what about the stuff I have forgotten here?

"*sigh* Some men are really hard to manipulate!" - Orchid

This post was deleted by request of the author.

This post was edited by frank on Sep 04, 2002.

Why not write the server and client in Java?

Because I suck at Java and wouldn't have a say then. ;-)

No, honestly, if everybody wants Java I guess I could live with that. I've just suggested C because it's fast, very cross-platform and I know it well. :-)
The problem is, will everybody (especially Windows users) have a Java runtime environment? Or are you using a Java compiler that produces OS-specific binaries?

I don't know anything about Kylix. What language is that anyway? Pascal???

It's supposed to be the Linux version of Delphi, which basically is a RAD tool for Object Pascal. (Before flaming me, please note that the Delphi compiler currently produces the fastest compiled code for Windows. :-) ) I'm currently trying to install Linux on my main machine and try it out, but my unusual disk partitioning is still causing me trouble. I'll let you know when (if) I can give Kylix a try.

"*sigh* Some men are really hard to manipulate!" - Orchid

The problem is, will everybody (especially Windows users) have a Java runtime environment? Or are you using a Java compiler that produces OS-specific binaries?

JRE on windows runs perfectly fine. The version that one can download from SUN's website is pretty stable. For development purposes, I have always used JBuilder for windows, and never had any problems with it's compiler or it's runtime environment.

Java compiler always converts code to OS-independent byte-codes. And, the JRE which interprets the byte-code is OS-dependent. So, the same byte-code can be executed by any JRE.

Love is blind, but marriage is a real eye opener.

This post was deleted by request of the author.

A couple benefits that Java would give us is that the server could run on a FreeBSD machine or a Linux machine or a Windows machine. One client would also run on each of these platforms out of the box.

Just keep in mind that we're talking about a P75 with 16Megs of RAM. If you want to run a Java app on it, we either need another server or some serious upgrading. I'm hoping to upgrade to 64MB soon, but Java isn't exactly known for economic use of CPU resources either, so...

"*sigh* Some men are really hard to manipulate!" - Orchid

This post was deleted by request of the author.

However keep in mind that a server written in Java will mean that we may not need your P75 with 16 megs of RAM.

Hmm, one can also write pretty platform-independent ANSI-C code. :-)
But don't worry, I just keep telling that because so far nobody besides me has offered a server.
I could do the C stuff or even explain you some things. OTOH it's probably better to stick to one single language for the server. Everything else might quickly get complicated. When the majority decide that this be Java, that's perfectly okay for me. After all it's a chance for me to dissect your code and learn something. :-)

I'll see what I can do to upgrade the server. Maybe I can even get my hands on another, more powerful machine. Most of the CPU-hogging stuff will be done on the clients anyway.
Or does anybody else have a more powerful machine with a permanent 'net connection?

"*sigh* Some men are really hard to manipulate!" - Orchid

This post was deleted by request of the author.

I will upgrade my server ASAP then. :-)

Is it too early to create shell accounts for you?

"*sigh* Some men are really hard to manipulate!" - Orchid

This post was deleted by request of the author.

we are talking about a p75, right ?
and you are planning to do the server in java ?

i mean .. the 'i need lots of resources' - java ? the 'i create ~20 threads to do some simple http-serving' - java ?

i recommend using c and a really simple protocol.
some udp messaging would be the protocol of my choice.
as soon as the way of communication is defined, publish the specs so anybody can write his own client :)

Note that, u. o. u., s. is not a computer program and his answers might not always be appropriate.

I think that it is nice to have both the server and the clients in the same language but it is not an absolute necessity. We can explore the possibilities of writing the server in the language which is suitable for the job, and can write the clients in any language for any platform that we wish for.

The server side
Lets itemize the problems that the server is going to solve. Then, based on that we can decide which language will be the best. Here is my initial try to list the things that the server will do:-

Functionality 1

Communicate with the NAO server to get data (new posts, other statistics)

Publish it all the clients that have subscribed to its service.

We might have multiple types of subscriptions, some client might subscribe to all the types of messages, some client might subscibe to only broadcast/important messages (special messages/NAO updates from the NAO admin (Jaz)).

Functionality 2

Allow multiple users (clients) to communicate with each other (pager functionality) via this service.

Keep a history of such communication per user if the user desires so.

Enable/Allow voice communication between NAO messenger users

Allow the users to share applications/send files.

For now, it looks like we need a pretty simple model - an application that should be able to send and receive messages in a standard format (some protocol that we have decided upon), and store a history of such messages per user.

So, the current problem can be solved by any of the following languages C++, perl, python, C, Java and .NET.

The Client side
The client side needs to have a proper GUI to display the messages, and should be able to probe the server periodically to receive messages and/or the server should be able to send messages to these clients. So, the clients should be written in a language which has good GUI support (Java, and .NET come to my mind immediately).

Which language shall we use for the server?
Okay, that still does not answer the question as to which language shall be used. For the client side, I think we have more or less unanimously chosen it to be Java. However, since we are going to make it an open-source, and will release the specifications, anyone is allowed to build the client in whatever language they desire. It is the server that we have to decide now!

If null is able to upgrade the server it would be great. If we use the current P75, then we should go with C++ or C. I feel so helpless. I have permanent DSL connection at my home, and my home computer is on all the time, my wife uses it to communicate with her relatives and friends in India and other places. Apart from that we dont use the machine ever. It is a Pentium 4, 1.7GHz, 256MB DDR SDRAM, 40GB hard disk. The only sad part is that it runs Windows XP Home edition, and since we are under the current warranties of Dell, if we change the OS, the warranties on the system will become void. So, it looks like I can not do anything in spite of having such a machine at home :(.

Love is blind, but marriage is a real eye opener.

This post was deleted by request of the author.

This post was deleted by request of the author.

The only problem I see would be that you're the only one able to run/debug the server. If you don't mind this (and probably even got a static IP), I don't see why it wouldn't work.

...Provided you're willing to let us use your computer ofcourse. :-) But I assume you wouldn't have offered it if you didn't.

What do you think?

"*sigh* Some men are really hard to manipulate!" - Orchid

This post was deleted by request of the author.

NAO Integration
I totally agree on the NAO Integration section. We definitely do not want to give many functionalities to the messenger and would definitely not like it to replace the NAO website. Its purpose is to just integrate/link the NAO user to the website "all" the time. So, by instantly sending messages to the user about a new post being added to the website in a particular forum, the user is prompted in a way to "go" to the website to read the post.

The pager in the NAO messenger would merely replace/complement the functionality currently existing on the NAO website, and if NAO messenger is successful, the current NAO pager can just make calls to the new pager, thus, shifting the load to the server where the NAO messenger is hosted.

Server (hardware)
Can't really comment on this because my hardware knowledge is nil.

Server and Client (software)
If everyone desires that the client and the server should be written in the same language then I too support Java. The only worry I have is performance of Java clients on a windows machine. I mean there won't be any problems of porting the clients on a .Net language (C#, ASP .NET) and the performance would increase 10-fold. We can investigate this later on. For now, I am with Frank that we should use Java.

Even though all my expertise is in C++, I wouldn't suggest using that, because after using C++ for almost 8 years now, I feel that it is not suited for complex or business applications. It is very good for learning the programming concepts, but when it comes to handling real-time problems, the focus of the programmer shifts from solving the problem to using the C++ concepts. This takes more time, and still leaves many holes as no one can write the "perfect" program. Just my 2 cents on this one.

Love is blind, but marriage is a real eye opener.

This post was deleted by request of the author.

Sorry for not replying earlier, there was a lot to be taken care of. Null sent me little SMS reports about this while I was in France, although I'm afraid half of them got gulped by the French GSM network.

"Statement":

- You want to use a seperate server for your project, but still the data has to be pushed or pulled from NAO to this very server. This amounts to an awful lot of traffic, which is a PITA when you've got to move servers. I still remember how much of a PITA it was when I had to find a new hosting company some years ago and came up with the 150$/month hosting plan everywhere since the Coder's Knowledge Base with all it's traffic was still being hosted on NAO.

- As I said, I don't want to provide an API. So if there's communication between the pager and NAO, it would only go one way.

- Maybe I could provide data feeds for some things, but since I'm about to move to another city and start college I'm a little busy, and I have so many ideas for new, funky stuff for Netalive.org. I'd much rather like to implement these things in the spare time I have left, as they would probably be more useful to most users than the pager (which probably only the coding crew would use for some time) - It's really cool stuff you know :) Short, I don't know if I can give you the level of dedication that you might need.

That's it for now... in my opinion you should consider Frank's opinion that the project should not be 100% dependent on NAO. If you really pull this off, you would have invested too much work in vain if NAO stops playing with the pager some time in the future.

'Yeah, That's what Jesus would do. Jesus would bomb Afghanistan. Yeah.' - snowlion

- As I said, I don't want to provide an API. So if there's communication between the pager and NAO, it would only go one way.

'kay. Not a big surprise anyway. :-)

in my opinion you should consider Frank's opinion that the project should not be 100% dependent on NAO.

I support that. As I said before (? If I didn't forget it anyway :-) ), the whole thing should basically work by itself and just provide NAO access as an add-on.

I don't know what will happen when the pager finally works. I guess it depends on how many people we can talk into using it. :-) For me (I can't speack for the others) the primary motivation is that I've thought about building a pager for years, and finally there might be a chance to do it. I will definitely use it as long as I'm not the only one. :-)

"*sigh* Some men are really hard to manipulate!" - Orchid

This post was deleted by request of the author.

So where do we stand if Jaz does not want to make any interfaces to the website for us to use?

I guess we'll just have to wait until he's back tomorrow. :-) Altho he'll probably have a lot to read then, so it might take some days...

"*sigh* Some men are really hard to manipulate!" - Orchid

This post was deleted by request of the author.

Dont forget, Frank, you got only one more week of the life as you know it! After that... forget about coding, its sleepless nights for the next at least 5 years...you will be happy about each second that you have on your own, apart from all the joy, your life definitley will change to degree you cannopt imagine right now! ;-) Wrong time to start a project like this right now...I think! Just a guess....
;-)

After decades of construction my website is finally up an running: www.kkds.de

This post was deleted by request of the author.

I'd still like to contribute. :-)

I think what we need to do next is agree on a means to communicate. There are so many things to discuss. I have a pretty detailed idea how the pager should work, and I bet you have the same ideas, just a bit different from mine. :-) So what I'd like to do next is agree on some standards (protocols and the such), and maybe already assign some work.

Frank, do you think we should postpone part of the project for some time? I understand you're soon going to be pretty busy and suffer from acute lack of sleep. :-)

What we absolutely need is some place for extensive discussions. I see these media...

NAO (pro: everybody can follow the wicked paths of our work. Con: probably not everybody wants to.)

Mailing list

Shared, 'net-accessible document directory (for sources and specs)

IRC (for brainstorming and stuff)

Build some other, developer-specific discussion forum.

All of the above work without any special software. There are many other means to communicate, just add the tool of your choice to the list.

NAO, mailing list and IRC are not a prob. If a 'net-accessible directory is needed, I could offer FTP/HTTP and SMB (<--Windows network) access to ~4Gigs of space.

So: When and where shall we meet?

"*sigh* Some men are really hard to manipulate!" - Orchid

This post was deleted by request of the author.

This post was deleted by request of the author.

This post was deleted by request of the author.

This post was deleted by request of the author.

Well, if the idea is to get everyone on netalive to be able to talk to each other, then maybe we should lean towards a simpler solution as Frank suggested. What if we just made a channel on irc called NAO where we can just talk?

It is free, and it gives everyone a chance to talk.

"If I die of a heart attack eating bacon, I'll be a happy man." -My father

This, my friends, is a great idea. :) What about following it?

Un bacio è un'apostrofo rosa scritto tra le parole "ti amo".

We can talk right now, too.

"Sie wollen nichts anderes. Sie wollen kämpfen! Sie sind Soldaten! Fucking Wahnsinnige!" - Noel G.

This post was deleted by request of the author.

I'm still interested. Just don't have many more ideas to contribute and am currently waiting for something to happens. :-)

"*sigh* Some men are really hard to manipulate!" - Orchid

This post was deleted by request of the author.

Well, the mail I've sent you some two weeks ago contains everything I have to contribute for the moment. Also, all of us seem to be quite busy with personal things. Maybe response times will improve again in the near future.

An IRC channel is IMHO a nice idea, but it doesn't cover my personal communications needs... i.e. face-to-face chats, notifications of all kinds (user logged in, new post, ...), and - last but not least - being notified of events instead of having to watch some status window all the time. (Not to forget the coolness factor of something we've made ourselves! :-P ) Thus, I'd still support the idea with the pager.

My personal problem is that, since
1. everything is gonna be programmed in Java,
2. I suck at Java, and
3. RCD's machine is going to be the server (?),
there's not much I can do at the moment. I've already sent you my proposal for the technical details, and I'm willing to discuss/extend these and (try to) produce some Java code. I'd need some basic Java code tho, just to know where to start.

I'm a bit busy this week (moving/cinema/various favours). Hopefully I will have more time next week. Just tell me when you've got the time to work on it. :-)

"*sigh* Some men are really hard to manipulate!" - Orchid

What is a face-to-face chat???

"Sie wollen nichts anderes. Sie wollen kämpfen! Sie sind Soldaten! Fucking Wahnsinnige!" - Noel G.

What is a face-to-face chat???

In this case: a private, two(or more)-person chat (as opposed to IRC, where everybody sees everything you write).

"*sigh* Some men are really hard to manipulate!" - Orchid

Like ICQ...

"Sie wollen nichts anderes. Sie wollen kämpfen! Sie sind Soldaten! Fucking Wahnsinnige!" - Noel G.

Like ICQ...

Yeah, like ICQ.
Except that it's not so insecure, not so bloated and is supposed to have some NAO-specific functions. :-)

"*sigh* Some men are really hard to manipulate!" - Orchid

bloated???

"Sie wollen nichts anderes. Sie wollen kämpfen! Sie sind Soldaten! Fucking Wahnsinnige!" - Noel G.

bloated???

Yeah, "bloated" like in "this software has about a billion 'features' you'll never need in your life". Plus not everybody enjoys those pop-up ad info windows all the time. (Sure, you can turn most of them off, but they're still there, introducing bugs and eating your system resources.)

"*sigh* Some men are really hard to manipulate!" - Orchid

This post was deleted by request of the author.

This post was edited by frank on Oct 08, 2002.

...i am also occupied with a few changes at my end. have bought a new house, moved in yesterday, will reply in details later on...

Love is blind, but marriage is a real eye opener.