There are three ways into this house if you are a cat. The front door, which is large, old, and involves a human. The back door, which is usually ajar in summer. And the flap, which is small, modern, and the only one Tuga actually uses.

All three lead to the same kitchen. They differ in how much effort they cost, who else can use them, and how modern the fitting is.

Bitcoin addresses work the same way, and the difference shows up on your fee bill every time you spend.

1. Reading the Address

You can identify the type at a glance by the first characters:

• 1... — Legacy (P2PKH). The original format from 2009. Works everywhere. Costs the most in fees because the data is bulkier.
• 3... — Script / wrapped SegWit (P2SH). A transitional format. Cheaper than legacy, and accepted by almost everything. Also used for multisig setups.
• bc1q... — Native SegWit (bech32). The mainstream modern format. Noticeably cheaper than legacy. Supported by every serious wallet and exchange today.
• bc1p... — Taproot (bech32m). The newest. Similar or better efficiency, better privacy properties for complex spending conditions, and improved multisig. Support is broad now but not universal — a few older services still can't send to it.

If you've ever wondered why your wallet suddenly generates addresses that look different from the one you wrote down three years ago, this is why. Nothing is wrong; the fitting was upgraded.

2. Why the Fee Differs

Recall that fees are charged by transaction size, not by amount sent. The address type changes the size.

SegWit — activated in 2017, and the subject of the conflict I wrote about in the nodes article — restructured how signature data is stored, so signature-heavy parts count less toward the size calculation. The practical effect is that spending from a native SegWit address is meaningfully cheaper than spending from a legacy one. The exact saving depends on the transaction's shape, but the direction never changes.

Taproot takes it further for anything complex. A multisig spend can be made to look like an ordinary single-signature spend on-chain — smaller, and considerably more private, because observers can't tell a complicated arrangement from a simple one.

For most people the summary is: if you're still using 1... addresses, you are paying a permanent tax on every transaction for no benefit.

3. What to Actually Do

• Use bc1q (native SegWit) as your default. Universally supported, cheap, boring. If your wallet offers it, use it.
• Use bc1p (Taproot) if you're comfortable checking compatibility, and especially if you use multisig. Almost everything supports it now; a small number of older services still don't.
• Keep 3... only where a counterparty demands it. Some legacy business systems still do.
• Stop generating 1... addresses. You can still receive to them, and old ones still work forever — nothing is lost — but there's no reason to make new ones.

Migrating is simply a matter of creating a new wallet (or a new account within your wallet) of the modern type and moving your coins to it, ideally during a quiet fee period. You'll pay one transaction to save on every future one.

4. Two Things That Trip People Up

Sending to an unsupported type. If an exchange's withdrawal form rejects a bc1p address, that's an old system, not a broken address. Use a bc1q address instead, or a different service. Your funds aren't at risk — the transaction simply won't be created.

Case and typos. Modern formats (bc1...) include an error-detecting checksum, so a mistyped address is almost always rejected rather than accepted into the void. That's a genuine safety improvement over the old format. It doesn't protect you from a valid address belonging to someone else, which is the address-poisoning attack — so still check the middle of the string, not just the ends.

5. Why Bitcoin Ended Up With Four

This is worth a paragraph because it explains something about how the whole system changes.

Each new format was introduced as a soft fork — a tightening of the rules that older nodes still accept. That's why every format keeps working forever: nobody was forced to upgrade, no coins were stranded, and the old doors were never bricked up.

The cost of that approach is exactly what you're looking at: four address types coexisting, a bit of confusion, and a decade-long migration that's still finishing. The benefit is that nobody woke up one morning to find their wallet invalid.

Bitcoin chose slow, backwards-compatible change over clean breaks. The clutter in your address book is what that choice looks like from the inside.

The Point

The front door still works. It has worked since the house was built, and nobody is going to brick it up — but it's heavy, it sticks, and there's a perfectly good flap two metres to the left that costs nothing to use.

Tuga worked this out in an afternoon. Most bitcoin users are still shouldering the front door open several times a year and paying for the privilege.

Check what your addresses start with. If it's a 1, you're overpaying. 🐾⚡