I have a widget that has share-on-digg and email-to-your-friends type functionality. I've implemented the emailing in what seems to be the standard fashion - a minimal dialog allowing entry of the to/from details, with the email then being sent using sendmail from my server. However there seem to be lots of issues with scalable email sending - host throttling, getting caught in spam filters etc. I'm wondering why I shouldn't change to using a browser mailto: link. I can still fill in the needed email content but the sending is then done from the users local machine and email mechanism.<p>Having each user do their own sending seems like a more scalable and robust solution to me. But it doesn't seem to be standard practice with similar apps. Any thoughts on why? Are there a significant percentage of folks who don't have an email client set up correctly for mailto: to work?
<i>Having each user do their own sending seems like a more scalable and robust solution to me. But it doesn't seem to be standard practice with similar apps. Any thoughts on why? Are there a significant percentage of folks who don't have an email client set up correctly for mailto: to work?</i><p>It might be more scalable, but it's (in my mind) not more robust. Most people I know use webmail with no mailto setup.<p>The impact will really depend on your content and your audience. Is your audience savvy enough to <i>Copy Email Address</i> if they don't have a mail client setup? Is your content worth their trouble to do so?
Most people who use webmail won't be able to use mailto: links. (Yes, I have set up my Firefox install to use Gmail for mailto: urls, but I'd been using Firefox a while before I discovered this was possible.)<p>Also, people using it from work might not want it using their work e-mail address and going through their work's mail server.