aparadja

Thanks for the feedback, pandadoodle and Glowbridge! The drop shadow is a great idea. Now that I look at the page, my eyes beg for the shadow. Never even thought about it before. Extending the "black glass" to the top is an interesting idea. I'll try it out first thing tomorrow. Maybe having a whole new navigation concept to work with will help me break through my lack of inspiration for it. Same thing with the footer, although I hadn't considered it a problem before. Mo' exposure, mo' problems, I guess .

Hi, I'm primarily a coder, and making web sites is sort of a necessary hobby to get my software out. This is my first landing page for a commercial app, and so I would like to make it as good as I possibly can. Here is the current landing page. In general, I'm quite happy how the site turned out, but there are a couple of things that I cringe to look at: The top navigation. It only has the product name and the FAQ link at the moment, but I need to add more navigation links soon. I have no idea what I'm doing here. Using the same font for the navigation and the product name feels wrong. Neither looks like a link. I'm not sure at all how they should relate to the underlying 960gs grid. Can I break the grid in navigation? How about the twitter link? I love the rhythm in the bottom of the page (with the benefits and the little icons). None of that smooth rhythm is in the navigation. The calls to action ("Buy" and "download"). My color choices are purely random, and the buttons currently feel somehow obnoxious to me. I've done some a/b testing with them, and the conversion rate is OK. But still, for some reason, they don't feel right. Are they too wide? Too big? Too close together? Or am I just overly critical? Those are the two main things I'm concerned about, but I would love to hear your feedback on anything on the site. Even small improvements can make a big difference. Thank you.

Thanks, everyone! And of course I meant "the best". The "good" was a typo.

IMHO, the "speed of the language" itself doesn't matter at all. For a typical web app, we're talking about milliseconds or microseconds. Compared to the delay caused by actually sending all the html, the images, js scripts, and css files of your website, those tiny fractions of seconds are totally meaningless. What matters is your development speed. If you can finish your app in half the time by picking PHP, then that's the right choice. If, one day, your app is so popular that it has to serve thousands of requests per second and those millisecond delays are starting to cause trouble, you can probably simply buy another server. Just don't worry about it now.

Why hash it? I see two downsides to it. You're probably trying to avoid filename collisions, but md5'ing the file actually leads to collisions. If two people upload the same file (some popular meme, for example), you have two files with the same name, guaranteed. When one of them deletes their account, will the other also lose their profile pic? If you're concerned about your code's speed, the hash calculation takes way more time than the if clauses. It's not usually a huge issue, but if you are already worried about performance, an md5 isn't going to help. I'd suggest using a random filename. Put 10 random characters together and call it a day. A collision is so unlikely that you don't even have to think about it, especially when dealing with non-critical things like profile pictures.

Hi, I'm a web developer from Finland. Lately, I've spent quite a bit of my time on my own web projects and struggled mightily with the design part. The creative juices tended to stop flowing when it was time to point the editor to the CSS file. I basically saw three options. Admit defeat and despair, live with crappy engineer art, or learn the basics of design. Well, here's me trying to learn. I've read some of the essential books, consumed web design sites, and most importantly, practiced and worked on my projects. I'm starting to actually like web design now. This site seemed like a good design forum, so I wanted to participate.