chinny22 wrote on 2022-12-16, 13:00:

Hmm maybe take the slot 1 system to recreate the true 90's experience, I can't be alone that usually 1/2 of day one of a lan party was spent on troubleshooting rather then actual gameplay?

In fact, even without the Slot1 system, it was very close to what you described.
Basically, everything that could go wrong went wrong 🤣.

I will take some time to write the story this evening. With probably a "do and don't" for people interested in setting up a demo.

Ok, so for the retrospective, a lot of things went wrong.

Just for a little bit of context, my retro collection is quite organized. Everything i have is tested, i take a picture of it, and i save the documentation and drivers in a directory on my website (see my signature, even if it's not 100% up to date).
After that, i generally store motherboard with a CPU + Cooler + Ram in a display case. And the cards (audio, video) in a big box, wrapped in antistatic bags, with paddings between the cards to avoid damages.
I have very few computer cases (probably 4 or 5) because it takes room, and i don't like assembling things in a case (long, painful cable management). Systems that have in cases are "permanent". These are the ones i usually use for games on daily basis. Every other systems (motherboards from the display case) goes on a testing bench, just on demand.

For this event, i had a few period correct systems available. I had a few hesitations, but finally, i settled on using my "permanent" (= with a case) PIII system + a Celeron 300A system on the test bench.

A week before the event, i started to install everything.
I started to build the Celeron system, and immediately, things started to go wrong.
Instabilities, system not starting. I lost a lot of time troubleshooting the issue, even partially re-caping the board at 2AM. Nothing worked.
I stored the board for later and looked for another option. I had two viable options : A Socket 7 Pentium 233 with Voodoo and a Super Socket 7 K6-2 500 with a TNT2 ultra.
Having HalfLife, UT99 and Quake3 deathmatchs in mind, i went for the SS7... but its a system i built recently, i repaired the card a few weeks earlier, and i never really used it until now. So a lot of unknowns things about drivers and compatibility.

Once again, a lot of things went wrong. Among them, i lost a lot of time on drivers issues with the TNT2 (which i got recently too), the AGP configuration, serial ports, issues with IRQ conflicts, "DOS Mode" storage drivers trashing the system performances. I finally ended up with the system working. But i had to custom build a serial port cable because the bracket connector pinout was different from the motherboard pinout. This serial cable adaptor is an important factors of the problems that occurred on D-day.

For the PIII system, i recently swapped its hard drive for a larger one, so i had to install windows again, but i went without issues. It's a system i know well and i have it for a long time. Only UT99 D3D rendered crashed the system, and i lost time to find a solution (just switching to OpenGL did the trick) . As it is my "permanent" W98 system, i use it from my desk, through a 4 way KVM switch with USB. So my usb mouse and keyboards are used on the system through the KVM integrated USB HUB. This, again would cause issues during the event.

For the final setup, i was planing to use a serial null-modem network (either direct serial, for games that supports it, or PPP over serial for games supporting IP).

On the event, i installed everything, booted both systems and immediately had issue with the PIII. USB Keyboard an mouse not recognized.
I tried with a PS/2 mouse and keyboards, nothing worked (hardware related ? i can't tell because i always use usb devices with it). I could go to the bios using the USB keyboard, and i managed to go to windows failsafe mode.
From here, i was able to use the USB mouse. Ok... so it works in failsafe, not in normal mode. This must be a driver issue. And it was probably caused by the fact that the system was used through the KVM. The driver worked well with my keyboard at home, but not the keyboard/mouse i grabbed in the office. This is my error N°1.
I deleted all USB devices and controllers in device manager, and rebooted in normal mode. I worked at first, but once it tried to load the wrong drivers, keyboard and mouse froze. Great, at least, i know which driver is causing the issue. Back to the failsafe mode, i tried to change it by a generic one. Reboot in normal mode. Still freeze. Back to failsafe. Let's disable the device. Reboot. Now Windows freeze before the desktop. Ok... now i started the feel a bit angry. booted to failsafe mode. Device manager, looked for the broken driver files and deleted them. rebooted. Keyboard and mouse worked. Windows asked for a driver. Nah dude, i'll keep it like that ! . One hour lost on site for this keyboard/mouse issue. The next time, i will bring the exact mouse and keyboard i used at home for the setup. Lesson 1 learned.

With both systems running, i started to setup the serial network using a null-modem cable. Cables in place. Hyper terminal started to test the connection... text comes in one way, not the other... Looks like my homemade cable has issues (it worked well at home) I found a loose connection, it was probably damaged during the setup. Fixed the loose cable. No change. Found another loose connection. Fixed it. 3 wired popped out. Great i did not have the paper sheet with the pinout. Now i'm unable to fix it. I had no plan B for network. This is lesson 2. I went home (i live a few min away from the office) Grabbed a PCI Ethernet NIC, a cable, a USB to Serial adapter, new wires and the papersheet with my connector pinout. Went back in the office.

1. The USB -> RS232 cable is not W98 compatible. Plan B failed
2. Nic Installed, ethernet cable in. I could not find my ethernet HUB power adaptor at home, so i had one hope : does the NIC driver have auto MDI/MDIX option. That was not the case. Plan C failed
3. I tried to change the wires from my homemade connector, but i could not repair it (without multimeter for continuity tests, i was blind) -> Plan D failed.

At this point, after 1 hour on USB input problems + more than 1hour on the network issue, the plan E was to abandon the network (and part of my self esteem) and grab a drink 😂

This was not a complete failure, because peoples enjoyed the games in local play. UT/Q3 vs bots. SkiFree and Worms were also very popular.

It was the first time i used retro hardware in an external event, and i can already learn a few things :

1. Bring the exact same system that was tested at home. Including peripherals. Same mouse, same keyboard.
2. If you are in the hurry (like me after my slot 1 MB refused to work) try to prioritize. I lost time trying to repair the motherboard. Time i could have used to better test the final setup instead. The motherboard can be repaired later.
3. Have a "safe option". The serial connection was something i wanted to show. But when my homemade pinout adapter broke, i had nothing else. Having a tested NIC + a switch and 2 cables (or a crossover ethernet cable) would have been great.
4. Have a variety of games. I focused too much on multiplayer games. (but i had storage size constraints too)

I can't define this as a total disaster, but definitely not a great success !
Anyway, with all your advices and good ideas regarding the games, i wanted to share the outcome with you. Even if its not great.

acl wrote on 2022-12-17, 21:45:

Ok, so for the retrospective, a lot of things went wrong. […]
Show full quote

Ok, so for the retrospective, a lot of things went wrong.

Just for a little bit of context, my retro collection is quite organized. Everything i have is tested, i take a picture of it, and i save the documentation and drivers in a directory on my website (see my signature, even if it's not 100% up to date).
After that, i generally store motherboard with a CPU + Cooler + Ram in a display case. And the cards (audio, video) in a big box, wrapped in antistatic bags, with paddings between the cards to avoid damages.
I have very few computer cases (probably 4 or 5) because it takes room, and i don't like assembling things in a case (long, painful cable management). Systems that have in cases are "permanent". These are the ones i usually use for games on daily basis. Every other systems (motherboards from the display case) goes on a testing bench, just on demand.

For this event, i had a few period correct systems available. I had a few hesitations, but finally, i settled on using my "permanent" (= with a case) PIII system + a Celeron 300A system on the test bench.

A week before the event, i started to install everything.
I started to build the Celeron system, and immediately, things started to go wrong.
Instabilities, system not starting. I lost a lot of time troubleshooting the issue, even partially re-caping the board at 2AM. Nothing worked.
I stored the board for later and looked for another option. I had two viable options : A Socket 7 Pentium 233 with Voodoo and a Super Socket 7 K6-2 500 with a TNT2 ultra.
Having HalfLife, UT99 and Quake3 deathmatchs in mind, i went for the SS7... but its a system i built recently, i repaired the card a few weeks earlier, and i never really used it until now. So a lot of unknowns things about drivers and compatibility.

Once again, a lot of things went wrong. Among them, i lost a lot of time on drivers issues with the TNT2 (which i got recently too), the AGP configuration, serial ports, issues with IRQ conflicts, "DOS Mode" storage drivers trashing the system performances. I finally ended up with the system working. But i had to custom build a serial port cable because the bracket connector pinout was different from the motherboard pinout. This serial cable adaptor is an important factors of the problems that occurred on D-day.

For the PIII system, i recently swapped its hard drive for a larger one, so i had to install windows again, but i went without issues. It's a system i know well and i have it for a long time. Only UT99 D3D rendered crashed the system, and i lost time to find a solution (just switching to OpenGL did the trick) . As it is my "permanent" W98 system, i use it from my desk, through a 4 way KVM switch with USB. So my usb mouse and keyboards are used on the system through the KVM integrated USB HUB. This, again would cause issues during the event.

For the final setup, i was planing to use a serial null-modem network (either direct serial, for games that supports it, or PPP over serial for games supporting IP).

On the event, i installed everything, booted both systems and immediately had issue with the PIII. USB Keyboard an mouse not recognized.
I tried with a PS/2 mouse and keyboards, nothing worked (hardware related ? i can't tell because i always use usb devices with it). I could go to the bios using the USB keyboard, and i managed to go to windows failsafe mode.
From here, i was able to use the USB mouse. Ok... so it works in failsafe, not in normal mode. This must be a driver issue. And it was probably caused by the fact that the system was used through the KVM. The driver worked well with my keyboard at home, but not the keyboard/mouse i grabbed in the office. This is my error N°1.
I deleted all USB devices and controllers in device manager, and rebooted in normal mode. I worked at first, but once it tried to load the wrong drivers, keyboard and mouse froze. Great, at least, i know which driver is causing the issue. Back to the failsafe mode, i tried to change it by a generic one. Reboot in normal mode. Still freeze. Back to failsafe. Let's disable the device. Reboot. Now Windows freeze before the desktop. Ok... now i started the feel a bit angry. booted to failsafe mode. Device manager, looked for the broken driver files and deleted them. rebooted. Keyboard and mouse worked. Windows asked for a driver. Nah dude, i'll keep it like that ! . One hour lost on site for this keyboard/mouse issue. The next time, i will bring the exact mouse and keyboard i used at home for the setup. Lesson 1 learned.

With both systems running, i started to setup the serial network using a null-modem cable. Cables in place. Hyper terminal started to test the connection... text comes in one way, not the other... Looks like my homemade cable has issues (it worked well at home) I found a loose connection, it was probably damaged during the setup. Fixed the loose cable. No change. Found another loose connection. Fixed it. 3 wired popped out. Great i did not have the paper sheet with the pinout. Now i'm unable to fix it. I had no plan B for network. This is lesson 2. I went home (i live a few min away from the office) Grabbed a PCI Ethernet NIC, a cable, a USB to Serial adapter, new wires and the papersheet with my connector pinout. Went back in the office.

1. The USB -> RS232 cable is not W98 compatible. Plan B failed
2. Nic Installed, ethernet cable in. I could not find my ethernet HUB power adaptor at home, so i had one hope : does the NIC driver have auto MDI/MDIX option. That was not the case. Plan C failed
3. I tried to change the wires from my homemade connector, but i could not repair it (without multimeter for continuity tests, i was blind) -> Plan D failed.

At this point, after 1 hour on USB input problems + more than 1hour on the network issue, the plan E was to abandon the network (and part of my self esteem) and grab a drink 😂

This was not a complete failure, because peoples enjoyed the games in local play. UT/Q3 vs bots. SkiFree and Worms were also very popular.

It was the first time i used retro hardware in an external event, and i can already learn a few things :

1. Bring the exact same system that was tested at home. Including peripherals. Same mouse, same keyboard.
2. If you are in the hurry (like me after my slot 1 MB refused to work) try to prioritize. I lost time trying to repair the motherboard. Time i could have used to better test the final setup instead. The motherboard can be repaired later.
3. Have a "safe option". The serial connection was something i wanted to show. But when my homemade pinout adapter broke, i had nothing else. Having a tested NIC + a switch and 2 cables (or a crossover ethernet cable) would have been great.
4. Have a variety of games. I focused too much on multiplayer games. (but i had storage size constraints too)

I can't define this as a total disaster, but definitely not a great success !
Anyway, with all your advices and good ideas regarding the games, i wanted to share the outcome with you. Even if its not great.

IMG_20221217_191320.jpg

Great story!

Honestly, if you only lost a couple hours on-site to issues like those, I think you did quite well. I was thinking you were going to say that it was a total catastrophe... your serial cable caught fire, damaged both PCs, fire alarm was set off, sprinklers soaked your gear and all the party-goers... 🤣

But no... people got to use the computers, and that was the point.

And you came away with a small list of things you learned. That's arguably the best result of any endeavor.

Also, when you described how you should have just focused on testing the end result rather than trying to fix a motherboard... I could feel it. That would totally be me. I'll have a vision of what I'm going to do and if there are any obstacles to that vision I will obsess over fixing them rather than changing the vision slightly. If I keep pushing and working on the obstacle it often works out and in the end it is clear to me (possibly no one else) that a lot of planning went into that highly refined "vision" of the outcome... but when it simply doesn't work out it is incredibly frustrating and I lose a lot of time... and the lost time brings even more frustration. 😐

Good job being able to adapt and to know when to just let it be. No one is going to be as picky about this as you are. I doubt anyone will remember the lack of multiplayer. 😀

I think it went pretty good as well. I think we can all tell worse stories from LAN parties.😉
Thanks for the conclusion.👍

Ozzuneoj wrote on 2022-12-18, 05:05:

Honestly, if you only lost a couple hours on-site to issues like those, I think you did quite well. I was thinking you were going to say that it was a total catastrophe... your serial cable caught fire, damaged both PCs, fire alarm was set off, sprinklers soaked your gear and all the party-goers... 🤣

That would have been worse indeed ! 😀

Ozzuneoj wrote on 2022-12-18, 05:05:

Great story! [...] But no... people got to use the computers, and that was the point. And you came away with a small list of thi […]
Show full quote

Great story!
[...]
But no... people got to use the computers, and that was the point.
And you came away with a small list of things you learned. That's arguably the best result of any endeavor.

Doornkaat wrote on 2022-12-18, 09:46:

I think it went pretty good as well. I think we can all tell worse stories from LAN parties.😉
Thanks for the conclusion.👍

Thank you very much.
In fact, some people asked if I'm considering bringing computers again. I will definitely think about it, but without the pressure of a fixed event date. And without the 90's theme i will have a broader choice of 2000-2003 systems (which is my personal period of interest)

People don't know what happend behind the scenes. They had fun so I would regard this as a success. Even more so since they asked to repeat it. Don't be to harsh to yourself.
I would add another point to your list for the next time: Bring everything twice (if possible). I mean a replacement mouse, keyboard, all cables. Even if it's untested it's better to have something to swap out if anything breaks or just for troubleshooting.

Thanks for sharing the story. Good job!