dalecosp;11059255 wrote:2. I am not known as the "brightest star in the PHPBuilder sky" either, so keep that in mind as you read this.
We can't all be laserlight or weedpacket (or nogdog or derokorian or...oh nevermind). You do just fine. 😃
Dalecosp is correct that HTTPS is usually sufficient to protect the privacy of communications between a client (i.e., a web app) and a server. More specifically, this kind of communication protocol means that all the packets exchanged between a web server application (or mail server or ssh server or whatever) and the client browser/mail client/ssh client should be extremely difficult, if not impossible to spy on. That doesn't necessarily mean that your web app is "secure." One of your users might email a link they requested via HTTPS to a friend (and the email gets snooped by a bad guy) or one of your users might tweet "hey lol if you wanna use my app go to https://example.com rofl." What I mean to say is that just hosting your web app via HTTPS/TSL/SSL is just a starting point. Think of it as a secure, private line that anybody can call.
Above and beyond using HTTPS/TSL/SSL, most applications require you to authenticate each user who may show up separately -- i.e., you identify each of your users separately and grant them access to the stuff they need. This usually involves prompting them for a login ID (either a username or password) and a password. It can be made more secure if you use key pairs or something like that. This would be analogous to answering the private line I mentioned above and forcing each person to identify themselves by name and provide some secret answer (or use voice recognition?) to identify them uniquely. It's important to note that you would never want anyone to mention their password or sensitive credentials unless they were speaking to you on a secure, private line. Think of it this way. If you are a bank and Alice calls and ask for Bob's information, should you give Alice his information? You really need to know who you are talking to before you start doing stuff for visitors.
Once someone has contacted you securely via HTTPS/TSL/SSL and authenticated with an ID or password, you can have your server generate a session ID and send this to the client so the client can identify themselves. Think of it like a ticket stub that can be used to locate someone's account. Like a coat check clerk at a fancy restaurant, your server needs to take care of tracking which coat (or which session) belongs to which customer (or authenticated user). PHP has features to support sessions which can be very helpful.
jrahma;11059239 wrote:Hi,
If I am going to make the service and copy it on my host then connect to it then I will achieve what I want but how about the security of the webservice? and how can I make sure it won't be indexed in Google and also how can I make sure no one will be able to access it with a kind of security?
If you want to politely ask web servers from crawling your web service, you can use a robots.txt file like this:
User-agent: *
Disallow: /
Keep in mind that every web crawler can rudely ignore that request and to try crawl any url they get their hands on. These urls tend to leak out as folks email them to each other and as users post stuff on forums and make bookmarks and stuff.
If you want to know 'about the security of the webservice' then you would need to be more specific. What, exactly, do you want to know? There's security that'll keep your little sister from reading your diary, and there's security that'll prevent nation-states with multi-billion-dollar budgets from decrypting your nuclear codes -- the cost to implement each is extremely different. Chances are that there are some common practices that will do what you want, but you'll have to decide for yourself what is "secure enough."