4 ms·

Sciter's author here. I am considering creation of EMail client, as a next step of my https://notes.sciter.com https://notes.sciter.com. The Sciter Notes rese

Sciter's author here. I am considering creation of EMail client, as a next step of my https://notes.sciter.com https://notes.sciter.com. The Sciter Notes resembles email client UI pretty close, so UI portion is feasible in months time frame. At the mean time I am doing html-notepad that has also close UI: https://sciter.com/html-notepad-new-kid-in-the-block/ https://sciter.com/html-notepad-new-kid-in-the-block/ If someone knows IMAP/POP/SMTP stuff and want to join or invest into the project - please let me know. I believe it is finally the time to make convenient email client.

johntash 8y ago
I only recently learned about sciter, and was pretty excited to start using it on a personal project. Thanks for that! I'm not exactly an expert on imap/pop/smtp, but I'd potentially be interested in helping depending on what your plans are for it. Honestly I'd be happy to help with any new privacy-focused e-mail client that has a convenient ui.
stevekemp 8y ago
> in months time frame. I'm going to assume you meant you could start in a month, rather than be completed in a month, because otherwise I'd like to save you some pain! I wrote a mail client, two actual, focussed upon the console. (Think "scriptable mutt".) Writing a mail-client is hard. Partly due to the sheer amount of bogus mails you'll get, and partly due to people having very firm beliefs about what they want. The user-interface is the easy part. More important is to design how you're going to work - Originally I allowed something similar to google's tagging, so you could "open a folder" which contained mail from multiple Maildir collections. It was not a success. Using a database to store your state is a cheap alternative, but risks issues getting outdated if you ever interact with the mail outside your program. You either have to force an "indexing" step - like notmuch does - or scan the filesystem tree. Neither of those approaches are great, for obvious reasons. (Also God help you if you want to work with remote IMAP servers. Standards? What are they!)

johntash 8y ago

I only recently learned about sciter, and was pretty excited to start using it on a personal project. Thanks for that! I'm not exactly an expert on imap/pop/smtp, but I'd potentially be interested in helping depending on what your plans are for it. Honestly I'd be happy to help with any new privacy-focused e-mail client that has a convenient ui.

stevekemp 8y ago

> in months time frame. I'm going to assume you meant you could start in a month, rather than be completed in a month, because otherwise I'd like to save you some pain! I wrote a mail client, two actual, focussed upon the console. (Think "scriptable mutt".) Writing a mail-client is hard. Partly due to the sheer amount of bogus mails you'll get, and partly due to people having very firm beliefs about what they want. The user-interface is the easy part. More important is to design how you're going to work - Originally I allowed something similar to google's tagging, so you could "open a folder" which contained mail from multiple Maildir collections. It was not a success. Using a database to store your state is a cheap alternative, but risks issues getting outdated if you ever interact with the mail outside your program. You either have to force an "indexing" step - like notmuch does - or scan the filesystem tree. Neither of those approaches are great, for obvious reasons. (Also God help you if you want to work with remote IMAP servers. Standards? What are they!)