Interviews

Interview with Maxim Masiutin Of RIT Labs
By Frank Vest
Edited by Frank Vest

Q: Tell us a little about yourself. Just who are you? :)

I'm the vice president and co-founder of RIT LABS, software company specializing on communication software. Our flagship products are Dos Navigator, Argus, The Bat! and SecureBat!

Besides my vice-president functions, I also do a lot of programming.

I'm 23 years old, married, and we have a 4-month nice daughter!

I like the communication with people. I also like music :-) In 1996, when I started Argus development, I liked rock, mostly so-called classic rock ('60s, '70s and '80s), but now it seems that I do not like aggressive music any more. Now I like blues, jazz, and new age music. My favorite jazz guitarist is Enver Ismailov. I met him on a concerto and he is very charming.

I live in Chisinau, Republic of Moldova. We have a great opera theater here and we used to visit operas with my wife.

Q: What got you first interested in writing Argus and/or other software?

The reason is simple! I am a fidonet fan and I like programming! We were running the biggest node in the city! I was astounded by the programmers and sysops who were experts in fidonet technology, why should I stay and not write a mailer that the people and I would like?

Q: How long have you been programming?

I'm in programming for ten years. My first computer was a MSX/2, very neat computer!

Q: What do you see as the strong points of your programs? Give us a quick review.

The strongest point is that it works with dial-up and TCP/IP nodes simultaneously, and it supports multiple lines in a single process.

In the past, we ran the biggest node in the city, it has three lines, 2CM and one night. The node was operating under OS/2. We had three instances of Xenia Mailer for OS/2, one Xenia for each line. Then we had connected to the Internet via a leased serial line and run some more instances of Xenia to handle VModem (TCP/IP) connections. Under OS/2, the mailers could use the awesome Serial Input Output driver by Ray Gwinn, that provided VModem, telnet and raw TCP/IP connections. I asked myself questions: why Xenia, being an OS/2 application, couldn't handle multiple connections, modem and TCP/IP having only one system process of xenia.exe, one instance of the program, running? Why should I have six processes consuming a couple of megabytes of RAM for each? We had 16MB of RAM under an Intel LX server based on an i486DX2/66 processor. This was a rather powerful computer for that time, but the multiple instances of Xenia consumed a lot of memory and made this computer slow responding to user commands under Presentation Manager.

When the users of OS/2 had a facility of transferring their mail bundles via TCP/IP by means of SIO by Ray Gwinn, there was nothing for Windows95.

We decided to write Argus. My colleague, programmer, Stefan Tanurkov, who was working with Intershop Limited on a joint project, TERMINATOR 2, the full fido solution for DOS, and he was helping me a lot at the beginning of Argus development. There were also a lot of experienced sysops around me who were helping me with advice. Argus was also influenced by Xenia Mailer.

First version of Argus was written in October 1996, it only supported dial-up connections (no TCP/IP). We used this version internally on our node. We left Xenia for TCP/IP only, while Argus was serving three dial-up nodes on a separate machine under NT 3.51. Our OS/2 machine began to work faster, while there was no speed degradation on NT machine. Argus was using the same amount of memory for three dial-up lines as one instance of Xenia.

First version of Argus was not a stable version yet, and a couple of sysops in our network has started to fight against Argus, but the majority of sysops in the network were friendly to the new weak mailer, they didn't support those who were against Argus, they even started installing Argus to help us, and the project has survived.

Three month later, in January 1997, Argus has started supporting WinSocket, and has supported Raw TCP/IP, Telnet and BinkP protocol.

As a goal, Argus could serve an almost unlimited number of dial-up and TCP/IP connections simultaneously, with small memory footprint.

BinkP by Dima Maloff was the new revolutionary protocol for transferring fidonet mail bundles over the internet. While VModem, Raw TCP/IP or Telnet did only provide a TCP/IP layer for old handshakes and transfer protocols (EMSI, Hydra) that were implementing their own mechanisms of recovery of errors that could occur over a serial line, BinkP was relying on error correction provided by TCP/IP and didn't do unnecessary error protection and synchronization steps. Such new technique gave a fantastic speed increase over the TCP/IP connection with delay, e.g. satellite connections.

I would like to say thank you again to Ray Gwinn and Dima Maloff, the major contributors of using Internet as a transport for transferring Fidonet bundles.

Q: Where do you see BBS systems going in the future.

Fidonet won't die. The transport and the network topology will be evolving, but the community, the people, and the spirit of the network of friends will remain. I believe that fidonet is the network of the friends. I do not agree with those who get upset when looking on the nodelist that shrinks week after week. Even if there will remain no fidonet nodes, the people who were at the beginning of fidonet will remain and will keep communicating with each other.

Maxim Masiutin
Vice President, Rit labs S.R.L.
http://www.rit labs.com/