Robert Jacobsen

Project work: So far, so good.

Monday, 6. October 2008, 18:12:04

As I've written earlier, we're having a web project at school, and now we're finally getting somewhere. Before I go on: Yes, I've re-used the background from this blog in the headers on the site.

The website parses RSS information from Norwegian technology sites, and try its best to comprehend what happens. And it works... somewhat.

The thing I dislike is that I due to the differences of the sites I am checking have to have several conditions for the parser, thus the parsing function becomes an abomination. The code is very much "as is", and not particularily optimized. Goal one is to have a functional code. Goal two is to get it optimized, if possible.

If you want to browse the site, you can check out the working copy at http://s.voltage.nu/~robert/leep/. Feel free to give comments regarding the design and layout, as I am not certain what I should add to give it some more "spice" and yet keep it slick and clean.

Weekends, weekends, wonderful weekendsOpera 9.6 is out!

Comments

Telperion # 6. October 2008, 18:45

It looks awesome!

Espen André Øverdahl # 6. October 2008, 18:55

Hot! Works great in Opera. Good job!

Site:

* Is it possible to have more text to each news story? It wouldn't hurt to have at least one more paragraph of text. You can maybe have a word count (max 250)?

Design:

* How does it look with white background?
* Have you tried with some more space between the header image and top?
* Maybe seperate the menu from the header image for something new?

I think this site could very useful for students on mobile phones using Opera Mini... Easy access to content when you're on the bus, watching your GF shop, etc.

Robert Jacobsen # 6. October 2008, 19:40

Thanks :-)

@Espen - I am aiming for it to work in Opera as my main priority (since Opera is my main browser)

Regarding the text, it's quite hard to get more text since it's parsing RSS and there is no more text to fetch from the feed. There's an option to get more text by parsing the website itself. I'll look into it, but I cannot promise anything since there are no such things as standards of these sites. I cursed many times trying to parse ITavisen, and even moreso when I tried to parse dingz.no, but I gave up since it was futile as the last site worked against every standard I have been taught to follow.

I will try out the design suggestions. If I render the site correctly with those improvements in my head, it might prove to be a great improvement. :-D

Thank you for your feedback. :-)

Donny # 7. October 2008, 10:20

I would recomend not bothering with parsing the pages. Get what you can from the rss feeds. Pages are probmelatic as they are, plus they can change at any time.

Otherwise, nice

Though I don't understand a thing

Robert Jacobsen # 7. October 2008, 11:32

The problem is that according to our specifications we're bound to have images on the sites, and all but one of the feeds give no images. And even more annoying, you can't create a general rule just to get image art_img_123123 where 123123 is an unique identifier, no they are totally random in an own system spanning over several sites.

I'm closing in on getting the rules to work for the current sites, so I think it should go alright in the end, unless the sites suddenly changes their systems.

Aleksander # 8. October 2008, 08:49

looking good!

Donny # 8. October 2008, 14:46

well either take random images from a folder, or maybe put images that represent the feeds.

Robert Jacobsen # 8. October 2008, 15:29

The site is not supposed to parse the sites everytime someone updates it (it's just because we haven't agreed on the database structure yet), but rather be on a cronjob that runs, say, every 2 minutes. Unless the sites are changing radically, there should not be any problems parsing them in that way, I think. :-)