Avoiding Spam with Email Aliasing(macdrifter.com)
macdrifter.com
Avoiding Spam with Email Aliasing
http://www.macdrifter.com/2013/12/avoiding-spam-with-email-aliasing.html
12 comments
For many years I've used my own domain for emails with a wild card forward to whatever my current email address is. So whenever I'm asked for an email address I use [email protected]. So I don't have to worry about support for the + and no part of my actual email address is ever revealed. It also makes it easy for me to switch email providers.
I use the same approach. The only unfortunately side effect of doing so is that I have to deal with spams targeting at randomly generated addresses at my domain. Luckily Gmail's spam filter finishes most of them after some training.
Tip: use a subdomain as your catchall (e.g. [email protected]). I've been doing this for five years and haven't gotten a single spam to a random address. While there are various ways to learn about a domain's existence, subdomains fly under the radar.
That is an awesome tip. I stopped using a catch all when folks were doing reflected spam (trying to get my mail server to 'bounce' the email back to the spam target).
I'm curious: did you lose any emails when you turned off catch-all? I would like to turn off catch-all, but I receive a lot of (mostly automated) email to various addresses, and I don't want to set up a long list of email aliases.
My catch all dumps the from line and the subject line and date into a log. I've got a perl script that scans it once a week for anyone in my contacts list. I did create a set of ecommerce/web site aliases which are all structured similarly ([email protected]) which are processed separately.
When I moved away from catch-all, I analyzed my emails from the past 5 years to find all the email addresses I used (grep on my Maildir to find To: *, and then uniq) and then simply added ~15 forwarders. Haven't missed anything I cared about since.
Interesting! I'm gonna try it out and see if it works. But I'm wondering why?
I built a system years ago for automatically expiring email addresses: https://grepular.com/Automatically_Expiring_Email_Addresses
You can email any address of the format "YY-MM-DD"@tmp.grepular.com and it will work, but only if the YY-MM-DD date is in the future. Ie, [email protected] will work from now until Christmas and will bounce after that date.
It's really useful for signing up to services that I don't intend to use after that day, or for entering competitions etc. Competition draw is in a month? I'll pick an email address which expires in about 2 months and enter with that one.
You can email any address of the format "YY-MM-DD"@tmp.grepular.com and it will work, but only if the YY-MM-DD date is in the future. Ie, [email protected] will work from now until Christmas and will bounce after that date.
It's really useful for signing up to services that I don't intend to use after that day, or for entering competitions etc. Competition draw is in a month? I'll pick an email address which expires in about 2 months and enter with that one.
If you run your own mail server, this whole process is much easier. Just set up an alias name specific to whoever you're giving the address to; then forward it to your real email. If spam starts, you just delete the alias.
It's also fun to use aliases specific to which source it is being provided. You eventually get some surprises about the integrity of some sites when you start receiving a spam on an alias only used for that site, revealing that your information was sold or stolen.
This is especially true of recruitment agents who appear, at least in the UK, to sell your address to anyone.
Personally, I have used both the Gmail alias and the [email protected] strategies. I have since given up both as I found it too hard to remember what address I used where.
The main reason I quit doing this is because Gmail's spam filtering kept spam out of my inbox and I never cared if one of those aliases were "compromised" and it wasn't worth the effort. I guess it was a solution to a problem I never had.
I still use a [email protected] for those sites that I may need to provide an email address and I don't want to give them my real address, such as forums that make you sign in to see an answer.
The main reason I quit doing this is because Gmail's spam filtering kept spam out of my inbox and I never cared if one of those aliases were "compromised" and it wasn't worth the effort. I guess it was a solution to a problem I never had.
I still use a [email protected] for those sites that I may need to provide an email address and I don't want to give them my real address, such as forums that make you sign in to see an answer.
I use the [email protected] alias when I can but MANY websites don't support that as a valid email address. Even tech companies that should know better.
I've been receiving jcpenney spam for years because while their account system allows you to sign up with a "+" their unsubscribe system doesn't.
They _do_ know better (i.e. they know you use it for filtering) and that's why they don't accept it.
I don't like it either, but it's a reasonable choice for them to make.
I don't like it either, but it's a reasonable choice for them to make.
I completely disagree! This is not a reasonable choice. Email forms must confirm to the RFC standards or else there is no standard. The biggest offender I've run into is Yahoo. Yahoo's email address filtering won't allow me to use my email address because they reject my TLD. Even lousier, their technical support form also rejects my email address, so I can't even send them email to have them correct the problem.
I presume you mean the SMTP RFC standards. If so, then it would be unreasonable to respond with 501 (Syntax error in parameters or arguments), just because they don't like your email address format or TLD.
However, if I'm offering a service and collecting email addresses as part of that service, my mail form could do additional validation (e.g. I might not want to collect email addresses from the UK (.co.uk) or from .edu or from people with 'john' in their email address.
I'm not violating an RFC with my email form, which may provide some benefit (e.g. stop fake email addresses, reject customers I don't care about) at the expense of losing/angering some other customers. That's a business question, not a standards issue.
(BTW - I've never restricted email addresses on a form in this way, but can empathise with those who do)
However, if I'm offering a service and collecting email addresses as part of that service, my mail form could do additional validation (e.g. I might not want to collect email addresses from the UK (.co.uk) or from .edu or from people with 'john' in their email address.
I'm not violating an RFC with my email form, which may provide some benefit (e.g. stop fake email addresses, reject customers I don't care about) at the expense of losing/angering some other customers. That's a business question, not a standards issue.
(BTW - I've never restricted email addresses on a form in this way, but can empathise with those who do)
I honestly don't think they do know. It's more likely that some frontend dev just copy/pasted some email validation regex they found somewhere and called it a day.
Same here, but its better than what I hit the other day - the ptrchase form accepted my +'d email, but the sign in form wouldn't. I had to get their staff to change my email in their database just to get the thing I bought...
You can rearrange the dots in your username to create variations. err.or54 e.rror54 er.ro.r.54 then you can filter the variation to a less frequently checked folder.
I had previously dismissed this method for spam-prevention because how many dot permutations can you possibly create, and how would you remember which permutation corresponds to each company (so you know which one to block when you start receiving spam). With just 11 letters in your username though, you could generate upwards of a 1000 unique email addresses! No need to remember them in a database, because you can just block the address that's getting spammed.
I use this approach in conjunction with server-side sieve, that sorts messages into individual folders.
if envelope :detail :matches "to" "*" { set :lower :upperfirst "name" "${1}"; fileinto :create "INBOX.${name}"; stop; }
if envelope :detail :matches "to" "*" { set :lower :upperfirst "name" "${1}"; fileinto :create "INBOX.${name}"; stop; }
I have my domains hooked up with google apps (there was a time when it was free, still is if you got in early), and they allow you to have a catch-all for email addresses that you can redirect to one of your user accounts. So each time I need an email for a site, I just put <site_name>@some_domain.com.
This way there's no need for the +alias in the user part, and some site tend to block this because they know it's used for aliases.
Doing this, I know exactly where this email was used. Been doing this for years, and works great.
This way there's no need for the +alias in the user part, and some site tend to block this because they know it's used for aliases.
Doing this, I know exactly where this email was used. Been doing this for years, and works great.
Have you actually encountered a site blocking + in email addresses? Because that's just categorically wrong behaviour.
My mail server (running qmail) uses - for this. Have been doing this trick for at least a decade now with it. I've never had a site complain about it.
My mail server (running qmail) uses - for this. Have been doing this trick for at least a decade now with it. I've never had a site complain about it.
What's stopping spammers from ignoring what comes after the plus?
Too much effort. They mostly go for fire-and-forget strategy. Any additional processing time means less sent emails = less profit.
What? It's minimal effort. 'Found' a gmail address? Drop everything after the + sign.
Spammers spend a lot of processing time on their emails (e.g. testing them against spam filters to tweak them to get through), so this extra step is so small as to be insignificant.
Spammers spend a lot of processing time on their emails (e.g. testing them against spam filters to tweak them to get through), so this extra step is so small as to be insignificant.
its so well known and common, I would be surprised if spammers already weren't ignoring dots and the plus for gmail addresses
Obligatory link to xkcd: http://xkcd.com/1279/
I use a free service called Gishpuppy, which operates kind of similarly to the fastmail aliases, except it just forwards the email to your Gmail, so you don't have to stop using Gmail. When you get spam, you simply cancel the alias and it becomes a blackhole instead of forwarding email to your real address.
With gmail spam filtering and the new tabs I find all this junk a waste of time. And who wants to use all these different addresses for logging in?