Forum: Editors/Reviewers Hangout

It sounds interesting. I can volunteer as I am a pretty fair proofreader. I can use and have over a dozen dictionaries including Webster's. I also can do Office Word, ODT, .TXT and a shift amongst several different document writers (Word Processors)including editing PDF.

send me an e-mail with a direct address.

Send me an e-mail through SOL under "authors" "Goldfisherman". Include your e-mail direct and I will include my direct e-mail. I have broadband. Attachments OK.

For the information of others here, I asked the webmaster about this kind of situation recently: how can an editor contact an author who doesn't have an author page yet, because they haven't posted their first story?

You can post emails online, but there are many online spambots (programs designed to view public sites searching for openly stated emails for them to dump with useless spam) looking for them. If you choose that option, you don't want to type out the full address, but instead use "username at service provider dot com".

For now, at least, the spambots seems unable to read the plain text version of email addresses.

Since that trick has become quite common the spam bots anticipate that and find these with ease.

Alas, my bots aren't quite that smart, as they haven't figured it out yet. So, crossing my fingers, I'll keep using it until they do, as it's also common for the same bots to now use common image-to-text conversion routines. :(

I rarely use PHP and I am not familiar with this specific problem so you should search for it and no doubt something useful will pop up.

This page has several: https://stackoverflow.com/questions/12592363/looking-for-a-php-only-email-address-obfuscator-function#12592483

My favorite is the Markup + PHP solution (scroll down to the third or fourth answer)

I have seen that solution but as stated in one of the replies it's probably countered by the email crawlers by now. I saw 2 other 'solutions' somewhere else on stackoverflow. The first was 'Don't bother, just rely on the spam filters'. The easiest method and will work for most. The other one I liked was 'Invent your own algorithm'. That will work too because it's not one of generally used obfuscating methods. It doesn't have to be complex, just different from the mainstream methods.

I've found the harvesting bots to be not too sophisticated (I've done my own testing).

That said, I rely on a combination of RBLs (Postfix simply rejects any mail from a server listed in an RBL), fail2ban (which blacklists, among other things, attempts to send to non-existent users as well as hacking attempts), and SpamAssassin.

Such that I get, perhaps 1 piece of spam in my mailbox per month, and because of whitelisting, my 'false positive' rate is similar - about one per month.

Of course, I run my own server, so I have complete control. :-)

How do you check for false positives?

My ISP introduced e-mail spam-checking by default. By chance I discovered it was occasionally bouncing legitimate e-mails. My ISP's support team tried to stop me switching spam-checking off but eventually relented and told me how to do it. Now I get the occasional flurry with half a dozen spam e-mails in a day (thanks to a research-related blog), but I'm now reasonably confident I'm not missing anything important.

AJ

How do you check for false positives?

My ISP introduced e-mail spam-checking by default. By chance I discovered it was occasionally bouncing legitimate e-mails. My ISP's support team tried to stop me switching spam-checking off but eventually relented and told me how to do it. Now I get the occasional flurry with half a dozen spam e-mails in a day (thanks to a research-related blog), but I'm now reasonably confident I'm not missing anything important.

For RBLs, I get a list of to/from addresses emailed to me each day. A quick scan tells me if something was bounced before Spamassassin got to look at it. That happens rarely because of whitelisting.

For Spamassassin it gets sent to a special user mailbox ('spam') which I check on a daily basis. For those rare 'false positives' I bounce it to the appropriate user and add the from address to the whitelist.

I probably spend about five minutes a day, total, checking for false positives. I think the key is whitelisting (e.g. I provide email services for a pastor and his church mailing list, and by adding all the members to the whitelist, we never run into 'false positives'.

Another thing I do is use 'throwaway' email addresses where the mail only goes to a 'catchall' box. If the source of email to that address becomes trusted (over time), I'll redirect it to my main mailbox. If not, I block it. I do this for sites that require you to sign up with an email address (because I refuse to use Facebook and have weaned myself from all Google services).

Such that I get, perhaps 1 piece of spam in my mailbox per month, and because of whitelisting, my 'false positive' rate is similar - about one per month.

Unfortunately, you've got to be cautious, as many honest contacts from your non-regular contacts will never show up, forcing you to dumpster dive through the hundreds of fire-walled Viagra and 'Girls in your area' spam.

My Apple devices no longer even lists the items in my Junk Folder, so I'm not even sure how to check for what's buried in them (hopefully nothing, but who's to know?).

Every now and then, I'll be forced to use a different email while on the road, and my editors simply never see it, because they didn't whitelist the alternate email (which is often the only one Apple allows me to use away from my network due to their new 'secure website' restrictions). :(

Unfortunately, you've got to be cautious, as many honest contacts from your non-regular contacts will never show up, forcing you to dumpster dive through the hundreds of fire-walled Viagra and 'Girls in your area' spam.

That doesn't happen. The RBLs get rid of most of that even before it hits the spam folder. I have to review no more than perhaps two dozen emails a day.

An image is simple, just create an image with the email address and display that instead of the actual email address. It's a dirty trick but it works... so far.

Alas, it won't work with most bot-check software, as they 'read' images of characters (as does Google Search Images). However, since so few people write script anymore, and banks have unilaterally given up on requiring them for checks, creating an image of an actual signature (rather than a common 'signature' font) is virtually foolproof, assuming you can still write your name legibly!.