Also, if there are binary build steps, you may want to consider using an ec2 instance for the build. (https://docs.aws.amazon.com/lambda/latest/dg/lambda-runtimes...)

All users of terraform will look the same, it's to get a general idea of "how many terraform using instances are there".

    curl https://eblong.com/infocom/ | grep href | grep -v http | cut -d '"' -f2 | awk '{print "curl -O https://eblong.com/infocom/" $0}' | bash -x

I usually do these sorts of things in a loop, sometimes like:

    <something> | while read f
    do
      sleep 10
      <do-something-with> "$f"
    done

I usually use this:

  <something> | tr '\n' '\000' | xargs -0 -n1 command

Just commenting because this was a bugbear of mine when getting started in bash.

- Never do a fixed bid. Not even when the job looks small or you are desperate for work. Any time a fixed bid goes wrong, not only do you lose money and time, but it's unbelievably demoralizing.

- Hire people. If you do it all by yourself, you'll end up doing all the jobs. Fun for about 1 month and then exhausting.

- Be honest. This is the one thing I feel like I got right. When I didn't know how hard something was, I would just tell the prospect or customer that while simultaneously telling them how we were going to figure it out. This approach always got me customers that I could work with and who were patient with me when I was consulting.

- Net 30 or better terms. Cashflow on Net 60 or Net 90 is brutal.

1. You do not fully understand the scope of the project, or think you do but actually don’t. In these cases a fixed bid will be catastrophic.

2. The client themselves do not fully understand the scope of the project, or think they do but actually don’t. This is even more of a disaster for a fixed bid, because even if you give them exactly what they asked for it isn’t going to be what they needed, and that’s what they’ll measure success against.

The advice for fixed bids deals more with the question of value capture. If by hard work, study, and practice, you have a way to deliver $1,000,000 of value to a client, it makes sense to charge (at least) $100,000 for it, even it takes only an hour. This is technically a fixed bid, but not in the traditional sense. Here you understand exactly what it is your client needs, how you’re going to do it, and what value they will derive from it, and so charge a percentage of that value instead of an hourly rate.

Also write down in your SOW that anything you send to client for review or as a deliverable, is considered accepted by client if they don't say anything about it for 2 days.

A little more nuanced view of this: if you know the client, their business and their systems well, there can be nothing more profitable than fixed bids. I have quadrupled my normal rate this way and ended up with happy clients. If you don't know them well, then, yes, fixed bids can be very, very risky unless you figure out a way to corner the risk.

This. Because you're not charging for your time, but for the value you add to the client's business.

Again, this was not some cheap and poor startup, a major company that just raised a ton of money with lots of tech media coverage.

In the end, it's my fault for not going through the exact UI logic on each of the designs, but it's easy to miss when there are 100 designs.

If there is a clear scope laid out in the initial contract it's more clear who is on the hook for anything outside the scope, and easier to recoup losses when the client fails to deliver the requirements properly.

On top that, offer a 5% discount if they pay within 30 days. Helps with a lot of accounting departments.

Charge another $5/10 an hour if you want to compensate.

I once onboarded as a vendor for a large, inflexible client and had to choose between two possible payment terms.

something like: EITHER Net 90 OR a schedule of discounts for any payment earlier than that (e.g 1.5% for 30 days and 3% for 15 days). and I wondered how many of their vendors simply jacked up their rates or fees to compensate, thus saving the inflexible client nothing. I guess this is very common.

Invoicing an engineering manager who needs things to happen tends to be the least friction from my side.

Even if it costs more on the purchase, inflexibility does help streamline business operations. Any business who can just mandate "this is how this process is going to work" and make it stick is going to benefit from not needing to make bespoke agreements with every little vendor they co-operate with. I believe WalMart is the classic example for this case.

You need to learn how to do fixed bids and then have processes to support it.

Think of it this way, you have 40 hours * sizeof(yourteam) to sell per week. There's a ceiling on $/hr or you price yourself out of the market and you don't have a time machine to create more hours. You can either add people to the team or detach the price from time with a fixed bid.

If you don't want hundreds, if not thousands, of consultants in your company then fixed bid is your only path forward.

Even better... get paid in advance with a retainer, if you can swing it.

Isn't the first rule of consulting literally that you should be billing by value, not by hours?

a) I can do this in 4-6 weeks, but we can also hire a consultant who can probably do this in 6-8 weeks and it is going to cost roughly this much (in this particular example I had intimate knowledge of the system and was proficient with the framework/tools).

b) Look, we don't have anyone on the team particularly suited for this project. We can either train ourselves, but the project will take about two months, or we can hire an expert who can probably do that in a month, and this is going to cost roughly this much.

Of course, be very conservative with your estimates and think hard, if someone from outside of the company can do the project faster (green-field project with unknown tech) or slower (upgrading internal system, for which the original developer is still around, but busy earning money for the company)

> Consequently, powering a full-size bus would require a battery pack that is unacceptably large and heavy, as well as too costly to make a battery-electric bus commercially competitive.

An interesting alternative is modern In-Motion Charging (IMC) buses: stick reasonably big batteries on a trolleybus, when overhead wires are available the trolley uses them and recharges the batteries; and it can run for some time on batteries so it's not as bound to OHL as traditional trolleys.