Korea IT Times Blog

Adobe Acrobat has been causing me an untold amount of headaches recently.  I use Adobe Acrobat to extract the text and images from our completed magazine PDF files so that I can post them on the web site.  But for some reason I do not understand, the latest version of Adobe Acrobat, version 9, creates very bad jpg images from its embedded images.  If I extract an image in the bmp format, then it’s exactly the same color as the original.  However, if I extract it in jpg format, it is red-tinted and ugly.  You may have noticed that some of the photos on the site don’t look so great this month.  It will be fixed soon, however.

It will be fixed because I found out that Adobe Acrobat 7, two versions older than the one I’ve been using, extracts jpgs better.  And that’s not all it does better.  First of all, it’s only about 200 megabytes on install, and takes about 2 minutes.  Adobe Acrobat 9, with its Creative Suite associations, takes upwards of one hour to install on this old computer and has a disk footprint of over two gigabytes.  Acrobat 7 is faster to start up, faster to read PDF files, and the jpgs it produces from image extraction look just like the originals.  Acrobat 7 is what I wish Acrobat 9 was, and perfectly fits in with my needs.  Maybe they should reference 7 for their inevitable upgrade to 10.

Update: At the request of Mr. Rosenthol of Adobe Systems I’m including four jpg photos to illustrate the problem I’m having.  I probably should have done this in the first place:

Now I can get the same results with Acrobat 9 as I did with 7 if I extract each image as a .bmp file first and convert it to jpg with GIMP or Microsoft Paint, but that’s one extra step I’d rather not do.

This week is the week where I try to set up online advertising for the web page.  All of the other steps in the future of the web site involve attracting users, but if we do that without a way to pay for those users then we’re just digging ourselves into a hole.  So before we add anything else, we need to put up advertising.

I’ve chosen to go with OpenX for an advertising solution, mostly because it is free and open source.  I was originally going to just use Google AdSense, but OpenX supports both Google AdSense and its own advertising network.  And, of course, there are already OpenX Drupal modules, OpenX and OpenX Manager, which will presumably integrate seamlessly with everything else I’ve done so far.  

There is one choice that I have to make, which is to download an OpenX installation and put it up on ads.koreaittimes.com myself, or to just use a hosted OpenX solution.  I signed up for the hosted solution yesterday evening and got an email acceptance this morning just in case, but I have yet to decide which is best for the site.  The only difference seems to be that I’ll be receiving ads from an OpenX server if I use the hosted solution, which makes me wonder how fast the hosted OpenX servers are.  Also, we’ll have to start paying if we serve more than 150 million ad impressions per month, which is a nice goal to reach for but something that isn’t going to happen next week.

One of the problems in migrating data from the old web site to the new web site was the large number of categories into which the old articles were divided.  There were actually 240 different categories, or Kickers, for the 2,934 articles that were hosted at the previous web site.  That’s only about 12 articles per category.  At that slim rate, categories really lose their meaning.  So, the first thing I did when looking over the old web site’s database is to do some Excel spreadsheet work to consolidate these categories into something more manageable.

I got it down to 30 different categories.  There were some synonyms, a few misspellings, and a lot of different ideas that could be consolidated together.  For instance, most of the old articles about different expos and events had three types of Kickers – the first was some variation on Event Preview, the second was the event’s name, and the third was some kind of variation on Event Sketch.  It was pretty straightforward to group those all under Events.  

On the new ProsePoint web site, categories are called Channels.  After setting up the web site, and creating different channels for all 30 Kickers, the menu area started to look a little cluttered.  Today I experimented with creating different section pages on the test site, similar to the New York Times web site or CNN.  I grouped all the channels into three categories: Opinion, Business, and Technology, and then set it up the channels to only be accessible from those pages.  It seemed to be even more confusing, however, and I realized that I even think that the New York Times and CNN’s pages are confusing.  Too many sections, too much stuff not viewable from the front page.  So I wiped the test site and went with the second solution for a cluttered menu – more channel/Kicker consolidation.

After some consideration I managed to consolidate some underused channels and have the number down to 21 now.  It’s one row less in the web site layout, and looks better.  Also, each section will be more populated with archive stories (as soon as I work out some kinks with archive importation) and everyone will be happier.  

I’d like to get some feedback on the header area of the web site.  Is it too long?  On my standard screen the title, search bar, menus, and breadcrumb bar (that black and white picture panel underneath the menu) take up about 50% of the screen real estate, and readers are going to have to scroll down a lot to see the rest of the site.  Is this good, bad or does it not matter?

Well the magazine edition is pretty much done for this month, so its back to working on the web site.  In order of priority, here are the objectives I’ve got for www.koreaittimes.com:

1.  Reorganize the front page links – There are three rows of channel links on the front page, which really pushes down the content too far.  On my 1024×768 screen it takes up way too much real estate, I hate to think what it looks like at 800×600.  I think there are about 30 channels up there now.  I have to do something about that.

2.  Online advertising – Since our web site is up and running, its time to set up online advertising.  My boss and the Mr. Kim, the advertising guy, are both very interested in seeing what we can do with this online space in the money-making department.  This is actually related to the front page links because with such a large header on the web site, where can we put ads?  Some reorganization might be in order.  I’d like to put Google Ads near the bottom of the articles, actually, next to the comment section. I think they would catch the most attention there.  I’ve looked into displaying ads on a Drupal site, and there seems to be a lot of OpenX integration. I’ll have to look into that after I get the links sorted out.

3.  Put up previous issues - This is turning into a more difficult assignment than I first thought.  While I do have the database of the old site, and I have managed to install and successfully use Node_Import to import a CSV file containing all the past web site data, two things continue to elude me.  First, the ProsePoint customizations put the subhead and the published date in a different kind of field, so I am unable to import them.  That gives me 2,900 articles all published on the day I did the import with no subheadlines, which is not an easy spot to work from.  I’m going to have to work very closely with the ProsePoint developer to see about getting this done.

 4.  Register with google news, reuters, etc etc. - After some consideration I think I’d better hold off on trying to register our web site as a legitimate news source until after we have our archives up.  That makes us look a little more respectable.  And, of course, before registering with Google I’m going to have to set up a News Sitemap.  I think it’s possible with the XML Sitemap module for Drupal, however that is still in development so I would have to do some tweaking on my own I think.

I’ve been poking around Google News, trying to figure out how to register as a legitimate news source with the new web site.  There is a long list of requirements to qualify to be listed in Google News, and the new web site qualifies for all of them, except one.

Here’s a list of the technical requirements, for those who are interested.  The only qualification that our new web site doesn’t pass is under Article URLs:

Technical Requirements: Article URLs

In order to be included in Google News, your articles URLs should meet the following guidelines:

Be unique. Each of your pages that display an article’s full text needs to have a unique URL. We can’t include sites in Google News that display multiple articles under one URL, or that do not have links to pages dedicated solely to each article.

Be permanent. Our system is unable to crawl sites that use a single URL for multiple articles. For example, we wouldn’t be able to crawl the page www.yoursite.com/news1.html if it displayed a different story every day. In order to ensure that our links to articles function properly, each article on a news site needs to be associated with one unique URL, and that URL must be permanent (i.e., it can’t be recycled).

Display a three-digit number. The URL for each article must contain a unique number consisting of at least three digits. For example, we can’t crawl an article with this URL: http://www.google.com/news/article23.html. We can, however, crawl an article with this URL: http://www.google.com/news/article234.html

Keep in mind that if the only number in the article consists of an isolated four-digit number that resembles a year, such as http://www.google.com/news/article2006.html, we won’t be able to crawl it.

The URLs on www.koreaittimes.com are automatically-generated aliases that consist of most of the words of a story title.  For instance, in the December story A Wave Too High? by Chun Go-eun, you may notice that the URL follows the format of web site name, then the word story, then the article URL which is wave-too-high.  This is really convenient for our users I think because it’s easy to remember, and it looks nice too.  However, it doesn’t contain a three digit number, so it doesn’t qualify as an article by Google’s guidelines.  I am going to have to think of a solution to this next week.

One thing I have been pondering is to put the node ID of each node into the story URL.  It could look something like www.koreaittimes.com/story/5534/this-story.  However, that would turn the human-friendly URLs into something that is impossible for casual readers to remember, with that 3 or 4 digit number in the middle.  I’m not sure how to resolve this yet, but I do think we need the traffic that Google News can generate.  I’ll have to think of something.

Our magazine regularly features a How To section which details projects that one person can do in the IT field.  Much of the rest of our magazine is related to big picture topics such as display industry growth for the third quarter of 2008 and other high-end topics, so these How To articles bring the focus back down to the individual.  And, actually, one of the reasons that I like the IT field so much is that it brings so much power to the fingertips of the individual.

And, if you are an individual looking to set up a magazine site similar to www.koreaittimes.com, just look in this month’s issue for instructions on how to do it!  Or, if you’d rather do it online, you can read this article for a week early preview.

How to Set Up Your Own Open Source Magazine Web Site

As you may have read or seen already, the Korea IT Times has a new web site at www.koreaittimes.com. It was all created using free, open source software that is readily available for anyone to use. The software is called Drupal. If you or someone you know would also be interested in starting up a magazine or newspaper-based web site, just follow this guide to get yourself started.

Read the rest of this entry »

Our new web site is not anything revolutionary in the era of online news, but rather just bringing our outdated previous web site to a respectable standard of Web 2.0 techniques.  We are still proud to be a latecomer into the ranks of acccessible, socialiable online news.  

My major aims in making the web site were to make the work of the dedicated reporters here at the magazine accessible to a larger audience (that includes my own I must admit) and to develop a community of people who are interested in online news about the Korean IT industry.  Quite frankly I find the whole subject fascinating and I really want to discuss it with other interested people rather than boring my friends at parties.

So, here is a draft of the article we will run in this month’s magazine detailing the major new features of the web site at www.koreaittimes.com.

Comments and Social Bookmarks and Archives, Oh My!

The Korea IT Times is proud to announce that it has a new web site, which is located at a new web site address:  www.koreaittimes.com.  We encourage all our readers to stop by and visit the new site, and to experience the new features that we have on the site.

Read the rest of this entry »

Today marks the official launch of the new web site for the Korea IT Times.  With the redirect meta tags put into the existing web site and the official sitemap uploaded to google we can hope to start getting regular visitors now.  I have google analytics accounts set up for both domains, so I can see if it is effective to direct people to the new site.  I have some doubts because the previous site had some peculiarities with its domain names.  The domain ittimes.co.kr held only a small .html file which redirected a full-page frame to another domain containing the web site information.  So in order to direct both people who visited the main web site and people who directly visited stories, i had to insert several meta redirect tags throughout both HTML and PHP pages.  We’ll see how effective it is at driving traffic to the new web site.

In the next few days I’ll have to write up a comprehensive post detailing all the features of the new site and how I did it, which will also be articles in this month’s magazine.  But for now, I’ll just take a bit of a break.