Background info: I am making an open source backup manager that does not require shell access. This is especially important to those users who are on subdomain hosting and can not use shell because, well, any smart host wouldn't. Because this script does not require shell access, the cron job backup ability is run through http://setcronjob.com to a page with GET variables to process the backup. Currently there are options to backup all public_html files, backup MySQL tables, backup a specific folder, and backup a specific file. Most importantly, an access password is required for authentication. My questions are:
1. Should I remove the specific file backup option? I mean... anyone can backup a file easily by copying its contents. Also, it seems pointless to create a zip file for one file inside, but if it's something that you would like, I can go with it.
2. Should I create an option to backup public_html files and MySQL tables in one process? This can get very lengthy depending on connection and size of the required backup. A sub question: Should I create a JQuery loading message telling the user to wait or something? In my opinion, it's a lot better than seeing a blank screen for 30+ seconds.
3. Should I create a backup specific MySQL table option? I'm debating on this one honestly. Seems smarter to backup all while at it, but you never know the circumstance the user wants.
