planning the epochalypse i

I’m slowly making plans for the time around the “Year 2038” problem. I’m sure you’ve heard of it if you are here reading this. The actual date/time is going to be Tuesday, January 19, 2038, at 03:14:07 UTC. We think it’s only going to be a problem for 32-bit software, but how many 64-bit programs out there have 32-bit timestamps being used somewhere? They may write logs using 64-bit timestamps, read them back in to 64-bit variables, but parse them as 32-bit timestamps.

Whatever the case, I know there is going to be some company, somewhere, running some 32-bit software on some VM on some futuristic hardware that is going to have a problem with that date/time. Their phone system is going to go down and no credit card transactions are going to work. Everyone will panic until somebody realizes that little VM, named “Tardis”, has been running their PBX and credit card processing for 30+ years and never got updated because it worked and nobody wanted to mess with it.

They tried to update it once, but the new version of the software had a “maintenance fee” with it and the salesman that sold it to them had left the company. Then the original vendor then went out of business, other companies wanted them to use a different PBX system or something, and they didn’t want to change their credit card processor too. They replaced the hardware and upgraded the OS, but the software stayed the same. Then they put it on a VM and just shuffled it around from server to server, office to office, and it just kept working.

That’s who I want to sell my services to. I want to sell peace of mind to about 5 - 10 of those companies. Chances are nothing is going to happen, but if it does, help identify the problem and fix the issues.

Maybe it’s keycard readers at the remote branches that never got updated, because they had plenty of extra cards and the readers never had a problem before. Maybe it’s a time lock on a safe somewhere. Who knows?

Pay me a few bucks to come in and help you figure out what things might go wrong. What you would be able to do, if any single aspect of your business stopped working, to get you back up and running. Ask me any question any time, and I will help you figure it out. We will make a plan for the worst case scenario, we will figure it out together. What in house systems do you have? What homegrown solutions do you have?

Sell it a disaster recovery plan, sell it a business continuity plan, sell it a “peace of mind” plan. I’m not going to code audit every application in your company, but I will help you identify systems and processes that may be affected and help you make a plan to address possible issues. What happens if your partners, affiliates, customers, vendors, or suppliers have problems?

I think it would be possible to sell this service to a few companies. Only cover last quarter of 2037 through end of Q1 - first quarter of 2038, and then be done and out. Maybe start earlier, I’m not sure. After end of Q1, accounting will be done, and any issues will have been identified and addressed in the wild by then.

A few questionnaires, a few meetings, a few phone calls, a few emails, a few on-site visits - most of the burden would be on the client to gather the information if it exists. A fair amount of the initial kick offs and brainstorming, can be done with multiple clients at once, letting them all share ideas and resources. There won’t be any proprietary information or trade secrets being shared - it’s going to be identifying systems and processes that could be affected and why. Is data being stored correctly? Are there any issues that you have already identified?

If something does go wrong, we can figure it out together. Where is it going to go wrong? What is going to go wrong? Did you reboot it?

I’m thinking of a flat fee for the service, maybe about 100 hours of a pricey consultant’s time, for a reasonable amount of “unlimited” hours. Getting the price point right will be difficult.

The difficult part will be marketing to the right parties. Depending on the business, there may be a hard freeze on anything (hardware, software, people, money, etc.) from December to January. Getting started in January may be too late, to get any real work or planning done. Maybe give January people a quick once over for a discount with no “day of” support.

If I can get some interest in the idea, I might be able to get a few people together in different markets and industries to form a small support network. Sharing ideas, sharing information, sharing resources, sharing contacts.

Anyway, that’s just my planning for the epochalypse so far.