Google Keyword Tool

Google’s Keyword Tool is a keyword research tool that is part of Google AdWords. Until now, it has been the most accurate, most accessible, and most useful keyword research tool available.

But there are major problems with the new version which result in much lower quality and accuracy keyword suggestions.

Sadly, Google announced this week that the new version of the Keyword Tool is coming out of beta at the end of August. At that time, the old tool will no longer be accessible. In recent months while Google has been testing the beta in general release, Google has provided an prominent link from the new interface to the old interface. That has allowed me to put off the day of reckoning.

But with the imminent demise of the old version upon us, I’ve taken some time to carefully test and compare the two versions side-by-side. The result is not pretty. My conclusion: Without a doubt that the new version generates just incredibly, dramatically worse keyword lists than the old version.

New vs. Old: Suggested Keywords Comparison

The main problem with the new version of the tool is that it generates much less useful and relevant alternative keywords than the old interface does. Like by a HUGE margin. It is shocking how different the results are.

I just compared the suggested keywords generated by the two versions using a set of ~35 keywords/phrases related to ‘web tools’ (list pasted below). While the new tool always generates worse results, this particular set of terms really highlight how extremely bad the new tool is.

New Version Results

The new tool generated a list of 800 keywords (as it does for every set of keywords no matter how long it is). The vast majority of these 800 suggested terms had absolutely nothing to do with ‘web tools’.

When sorted by ‘global search volume’, only 7 of the top 50 results even contained or was even vaguely related to ‘web tools’. Instead, as you can see in the screenshot below, the results are full of irrelevant terms like ‘joomla’, ‘drupal’, ‘ecommerce’, etc. Even wackadoodle terms such as ‘art design’ and ‘fashion design’ make it onto the list.

In other words, the new version generates a list that is essentially useless.

New Version: Suggested Keywords for 'Web Tools'

Old Version Results

On the other hand, the old version generates a highly useful list of 150 terms all of which are tightly related to ‘web tools’. Every single one of the terms in the list contains the word ‘tool’ combined with variations on ‘web’ and related ideas: ‘webmaster’ ‘web design’ ‘free website’ ‘seo’, ‘online’, etc. There is not one term that’s off target – and certainly nothing remotely approaching the uselessness of ‘art design’.

In other words, the old version generates an extremely useful and accurate list of suggested keywords.

As you can see plainly by comparing the two lists, the difference in keyword relevance between the old version and the new version is night and day. It is quite remarkable.

Old Version: Suggested Keywords for 'Web Tools'

Other Problems

There are at least two other major problems with the new version:

1. In the new version, when you click the advanced option to include only ‘ideas containing my search terms’, the new tool, which moments before eagerly pumped out 800 terms, suddenly becomes incredibly parsimonious and oftentimes offers up just a handful of terms. When I tried this for the list of 35 or so ‘web tools’ terms, I get LESS terms than I started with – only 26! And even the new ones in this list are not terribly relevant or useful. This boggles the mind.
2. The old version had a fantastically helpful option to see data limited to just ‘the keywords I entered’. This generated search counts and CPC stats, etc. limited specifically to your target list. The new version simply does not have this feature at all. That means that even if I can piece together a decent master list on my own (with or without the help of the new Keyword Tool), I can’t even validate this list using Google search volume and estimated CPC data.

I have to assume these major changes have been made by design and that Google has purposefully chosen to restrict/reduce the accuracy and relevance of the keyword data that they are sharing with us customer types.

Needless to say, I am not looking forward to losing the old version at the end of August.

List of ‘web tools’ related terms that I used to test both versions.

Buzzr: A Drupal Website Builder

Buzzr aims to make it much easier to build websites in Drupal by creating a user-friendly front end for the Drupal CMS. In a sense, Buzzr is a sort of WordPress-for-Drupal.

However, while WordPress is primarily designed to build blog-based websites, Drupal is fundamentally a more powerful and versatile website development platform that can be used to make virtually any kind of website. However, with this richness comes complexity that puts Drupal out of reach for many people.

Buzzr is trying to capture some of the superior functionality of Drupal while improving the interface to make it accessible to a much wider range of people. I think they have been largely successful in doing so.

Buzzr has been developed by Lullabot, a renowned Drupal development shop based in Rhode Island. Buzzr is in the early stages of roll out.

Buzzr is similar in concept to Drupal Gardens.com which recently launched in open beta. DrupalGardens is another simplified version of Drupal that has been developed by Aquia, another leading company in the Drupal community.

In the past week, I have tried using both Buzzr and Drupal Gardens to see how they work. I use Drupal to build most all my sites and I was interested to see how these new hosted Drupal platforms compared with the ‘real’ Drupal.

I will be posting separately about Drupal Gardens at some point, but this post is focused on the features of Buzzr.

Goal: Build a Simple Content Site

To make my testing as real as possible, I tried to put up the same relatively simple content website on both Buzzr and Drupal Gardens. This is a site is actually want to build, so I was looking to see if either platform would make it easier for me to create than doing so on a normal Drupal setup.

The site I tried to build had the following characteristics:

• How-to oriented website with ~20 web pages
• Webpages are largely text-based with a few images per page
• Design is a fairly basic 3-column layout
• Needs to support running advertisements, including header ad, ads in sidebars, and ads in body text
• Need to be able to fine-tune page layout, text styling, etc., ideally using custom CSS to augment default options
• Needs to support ‘taxonomy’ – the Drupal term for ‘tags’
• Needs to support ‘views’ – a powerful Drupal method for organizing content into lists

The above features are slightly more complex than a basic content site you might see for a small business or an artist or a personal website, but not a whole lot more. As well, it specifically includes some features (esp. ‘views’) that are somewhat unique to Drupal.

This type of site would be relatively easy to build in regular Drupal (at least it would be if you already knew Drupal and had a development environment in place to do it which I do). This type of site cannot be built in WordPress – the problem being that WordPress has no well functioning equivalent to Drupal’s Views module.

Buzzr Capabilities

Buzzr (and Drupal Gardens) have a wide range of built-in capabilities and I wasn’t able to look at all of them. Since I was trying to build a real website as quickly as I could, my focus was mostly on the capabilities required for getting this site built. As a result, there are a goodly number of Buzzr features – e.g., image galleries – that I didn’t look at in any depth.

Here’s a overview of Buzzr’s capabilities grouped into 4 major buckets. Again, I had a specific focus, so your experience may differ from mine, particularly in areas like ‘capabilities’ where specific website requirements are likely to vary quite substantially.

Admin User Interface

Buzzr has a very nice looking user interface (UI) for the ‘admin’ side of using the platform – basically the menus and other stuff you use to build and manage your website.

The buzzr admin UI consists of up to 3 menu bars across the top of the page, plus various editing and configuration layouts on the main page. You can find many pics and videos of the Buzzr admin setup in the Buzzr 101 section of the site.

I would say this interface is pretty darn nice. It is very graphical and much more intuitive than Drupal’s native admin controls which can be awfully cryptic.

The main admin options of ‘add content’, ‘manage content’, ‘design and layout’, ‘manage users’, and ‘site settings’ seem pretty sensible.

I won’t go into each piece part, but I wasn’t crazy about some of the screens. For instance, I did not like that there is no ‘menu’ option directly in the webpage editor – and there is no way to modify a menu title separately from the page title. As a result, when you initially create and save a page, the entire page title pops up in your main navigation menu. You can visit the ‘navigation menu’ page to change this and have it be subordinate to another menu item, but this isn’t too satisfying – and you still can’t change the menu item title there either.

Buzzr Admin Interface

A general criticism I’d offer of the layout is the size of the admin menu bars: they take up too much screen real estate. With all 3 menu bars open, they take up the top 180px of your screen. If you are on a 1080px or 1024px tall screen, these menus are taking up nearly 20% of total height – and that doesn’t include your browser frame and whatever toolbars you’ve got going. As well, many of the admin page layouts are too generously proportioned – e.g., the ‘Manage Static Pages’ page. This all needs to be more economically done to avoid too much scrolling and loss of focus.

Functionality

One way that Buzzr (and Drupal Gardens) keep the Admin UI clean and simple is that they support only part of the entire range of functionality available in Drupal proper. Since Drupal proper supports an enormously wide range of capabilities through ‘contributed modules’, this is a necessary thing for Buzzr to do.

The question is: has Buzzr chosen the right functionality to include to meet the needs of the most people? I think the answer to that is probably yes – or at least it is within striking distance. Since I know they will be adding additional features over time – e.g., I was told that they will be adding groups, calendars, contact page, and color picker functionality later this month.

The main things I personally needed to build my site were these:

- Basic content pages supporting text and images
- Upload and manage images for webpages
- Taxonomy (aka, ‘tags’) to allow grouping pages together
- ‘Views’, a special Drupal module that allows you to create lists of pages based on certain parameters (e.g., a tag)

Before I run through these quickly, I wanted to note a very important feature of Buzzr that was not required for my website, but that looks like it would help lots of other folks: their innovative Form Builder. This is a really outstanding Buzzr feature that combines two very complicated Drupal capabilities – CCK and content type manager – into one simple to use interface. This module allows you to very easily build a ‘custom form’ that contains any number or type of content fields. Using this tool you can create, for instance, a complex custom page for product spec sheets or product reviews or any other similar type of page that requires structured content. As an added bonus, Buzzr / Lullabot have contributed this module code back to the Drupal community in the shape of the Form Builder module.

Basic content pages Buzzr handles basic content pages very nicely using the ‘static page’ content type. A ‘static page’ contains several fields: page title, body text, tags, and date and time. It does not contain a ‘menu’ field as mentioned above. You are able to upload images to a static page.

In addition, Buzzr offers a number of additional pre-defined content types, including image galleries, blog posts, ‘press post’, and ‘custom pages’. The WYSIWYG editor is very nicely implemented. There is an option to switch to an HTML editor, too.

Upload images In the WYSIWYG page editor, Buzzr allows up to easily upload an image from your hard drive. The uploader allows you to adds image a title, tags, and other descriptors. You can also change the image size to fit properly on your page. You can also adjust aligned (e.g., left, right, center).

The part I did not understand was how the image library worked. While it worked fine to upload into the page directly, I would have actually preferred to upload into an image/media library so I could reuse the image on multiple pages with just the one upload. The images I uploaded directly to a page did not show up in the media library and I could see no other way to upload an image. Perhaps this is in the tutorials somewhere.

Taxonomy (aka ‘tags’) Tagging is a pretty basic capability that most all website builders offer and Buzzr is no exception. Buzzr allows you to tag every piece of content (pages, images, etc.) with one or multiple tags. These are nicely displayed on each page in the default templates.

That said, I would have liked to see a bit more full featured implementation of tagging. For instance, there is no master screen where you can look at and refine your tags and you can have only one tagging ‘vocabulary’ whereas Drupal proper and Drupal Gardens allow you to have more than one. This is useful if you want to tag one type of thing (e.g., pages) with one set of tags while tagging another set of things (e.g., images) with a different set of tags.

‘Views’ capability One of the best features of Drupal is the ‘Views‘ module which allows you to create lists of content based on user-specified parameters. For example, you can create a list of all content pages with the tag ‘websites’ or all pages of a certain content type or published by a certain author. In Drupal proper, Views is extremely flexible and supports very complex parameters.

My needs with my sample website were more modest: In Buzzr, I simply needed to create pages based on individual tags. So, for example, I needed a page that lists all pages with the tag ‘websites’; and another showing all pages tagged ‘SEO’, etc.

Unfortunately, I almost was not quite able to do this in Buzzr. Buzzr does have a simplified Views module, but it turns out to be a bit too simplified. Instead of generating a page with just posts tagged ‘websites’, it created a page with links to all tags used on the websites. While I maybe could have worked with that, it turns out each of these subpages contained the same complete list of all posts on the site. I’m thinking the feature isn’t working quite right – or maybe I didn’t configure the page correctly.

[Update: Via email, Buzzr CEO Ed Sussman has offered several clarifications about the Buzzr custom pages (aka 'simple views') feature. First, he said that there is a bug in Custom Pages that is causing the tag links not to work properly. He says this will be fixed shortly. Second, he noted that they have not enabled the ability to generate 'single tag pages' using Custom Pages, but that this is built into the code and may be enabled in the future. Lastly, he pointed out that Lullabot / Buzzr actually built the underlying code for this feature and contributed it back to drupal.org as a contributed module called Simple Views. This looks like a really fantastic module that hugely simplifies the well known 'Views' module in Drupal. They are to be commended for contributing it back to the community.]

Design

Buzzr has some very good design tools, but overall I would say it needs some further work. It is probably a step or two behind the theme editing features that I saw on Drupal Gardens.

The design interface has several parts: ‘select style’, ‘navigation menu’, ‘design tool’, ‘design (advanced)’, and ‘widgets/layout’.

‘Select style’ allows you to select from a dozen or so ‘themes’ (the Drupal term for ‘template’). These all have basically the same page layout and vary mostly in terms of the ‘skin’ – meaning background colors, font styles, and colors, etc.

‘Navigation menu’ allows you reorder items in your main navigation menu via a drag-and-drop interface. As far as it goes, this works very nicely actually, but it looks like you can have only 1 menu and that’s not ideal for many situations.

‘Design tool’ allows you to change various theme relates settings. For instance, the global font and font size are set here. Backgrounds for the header and the site are also set here. The site logo and favicon are also here. This is good stuff, but pretty basic stuff.

‘Design (advanced)’ is a simple interface that allows you to see and edit the entire CSS file associated with your theme. You cannot change basic page layout here, but you can change lots of other styling elements of your website. For people who know CSS, this is great, but others will struggle to know what to do here.

‘Widgets / layout’ is one of the most interesting parts of Buzzr. It is basically a drag-and-drop tool that allows you to manipulate the layout of your website. Here is where you determine how many columns your site has. It also has a bunch of pre-defined widgets (called ‘blocks’ in Drupal-ese) which you can add or move around on your page. These widgets are basically small content blocks which you can put into your columns. The pre-defined widgets include ones for text or images, for tags, for a mini-gallery slideshow, and so forth. There are also widgets showing users and social widgets for Facebook, Twitter, etc.

In terms of my specific needs, it appears that most of what I needed design-wise could be done reasonably easily in Buzzr. The drag and drop widget tool is very effective.

One sticking point was reformatting the header to get a 728x90px ad to fit alongside the site name and site slogan. With some encouragement from CEO Ed, I found that I could manipulate the CSS for the header to achieve the desired positioning. This involved firstly adding a new block to the page via the ‘widgets’ editor and then using the ‘design (advanced)’ editor and Firebug (a Firefox plug-in) to modify the CSS for the default header block and my new ad block. With a bit of fiddlign, I was able to put the two blocks side by side in the header.

Help / Support

Buzzr provides pretty strong and easy to understand documentation. The website provides a very helpful how-to section plus hands-on demos in a series of nice video tutorials. Since the whole thing is designed to be straightforward and user friendly, this documentation is plenty to get you going in building out your website.

While the documentation is quite good, a concern I would have is that there is no obvious venue for support beyond this info – like if you still don’t understand something, how do you get help? On Drupal Gardens, they’ve got very active user forums to help you out. I found these invaluable as I worked through building out a site there. It may be that Buzzr is intended more as a platform for Lullabot clients than as a general use website builder. If that’s the case, they don’t make that obvious.

But I can’t complain too much about support because I received an immediate response to the one query I did send them via email – and that came from their CEO, no less. Obviously that’s not scalable, but it does suggest their commitment to the product.

[Update: Ed Sussman, the Buzzr CEO, has now sent along numerous emails clarifying various points about Buzzr. He's been very helpful.]

Pricing

Buzzr has a tiered pricing scheme that seems fairly reasonable if you are planning to put up one or two sites. Prices start at $19.95 for the ‘Bronze’ plan. This seems designed for a smaller site typical of a small business or independent professional. If you have a larger site and require more hosting resources, then you can choose one of the 3 more expensive plans ranging from $39.95 to $119.95 per month.

The monthly price includes on-going upgrades of the Buzzr platfrom. This is a major benefit over a self-hosted version of Drupal since, unlike WordPress, the upgrading process is not automated and really quite a pain in the a**.

Buzzr has a 7-day trial period where you can demo the platform. This a great way to try out the platform and see if it is a good fit for your needs. I used only this option because it took just a few hours to conclude that I wouldn’t be able to do what I wanted to build out my site.

Even if I’d been able to get the website up and running, pricing might have been an issue for me because I would have wanted to host multiple small websites on Buzzr. Via email I learned that you may be able to get a price break if you put up a number of websites.

Conclusions

Buzzr appears to be very promising Drupal website builder. It is a relatively new product, but has many outstanding features and excellent potential.

Via email, CEO Ed Sussman offered this comment about Buzzr and it’s future:

“Buzzr is designed to be a real time saver for the average website owner who wants the power of Drupal in a more user-friendly package. We’re very pleased with the capabilities of Buzzr today, but we’ve got many important enhancements in the pipeline for the coming months. When the fully packaged version of Buzzr is released in September, you’ll see some really great new time saver tools beyond anything now in the public beta. I think people will be excited and they will clearly see things that set us apart from Drupal Gardens.”

Unfortunately, despite the great features, I was not able to get my website built in Buzzr. It currently doesn’t quite support my needs. Some of the problems I had were:

• Couldn’t generate ‘views’ type pages for individual tags
• Couldn’t [easily] manipulate layout of header to include an ad [as described above, this was eventually accomplished via custom CSS]
• Couldn’t change menu item titles [turns out this is wrong - titles can be edited, but the edit screens are very confusing and should be clarified]
• Cost to host multiple small websites on Buzzr is too high [but it should be noted that Buzzr costs are comparable to other similar tools - they appear high only in the context of my specific set of needs which calls for many small sites]

Note: Text in brackets above are edits / clarifications added after the original post went live.

Despite these issues, I would definitely recommend that others try Buzzr and see what they think. There are a number of really strong positives and surely the platform is going to continue to expand and improve.

Here’s my take on Buzzr’s notable features:

• Modern, very easy to use admin user interface that makes Drupal very accessible to non-programmers
• Strong set of Drupal functional capabilities that will meet the needs of many websites
• Excellent Drupal feature innovations in both the ‘simple views’ and ‘form builder’ modules that simplify and enhance Drupal usability
• Commendable generosity on the part of Buzzr / Lullabot to contribute their module innovations back to drupal.org
• Nice mix of content types and other specific features out of the box
• Built-in support for image galleries with lightbox-type overlay/slideshow (caveat: I didn’t test this too thoroughly)
• Simple, easy to understand theme structure (in contrast to Drupal Gardens which has mind bogglingly complex themes)
• Beautiful drop down menus based on Drupal ‘Nice Menus‘ module (note these may require a bit of CSS styling to get the right look)
• Excellent drag-and-drop widgets tool that makes it very easy to organize your website page layout
• Nice social media, newsletter, and other community features

If your website needs match up well against Buzzr’s current capabilities, you could be up and running with a high quality Drupal-based site in a matter of hours. Since using Buzzr is so straightforward, it’s well worth giving it an hour or two to see if it’ll work for you.

Here’s the details:

Netvibes Wasabi – A Very Readable Layout

Here’s a screenshot of my old Netvibes Wasabi reader – one of the basic ‘official’ themes of the zillions they offer. This layout was just a dream to read. I loved the spacing, the use of favicons, the nice date headers in the item stream, and the colors, among other things.

In general, Netvibes comes with a lot, lot more configuration and, especially, theming options than does Google Reader. One feature that I especially liked with the “Give feed items more space” option – this incrementally opened up the spacing giving feed items just the right amount of breathing room.

Google Reader – Too Dense and Cluttered

Here’s a screenshot of the same feeds in the default Google Reader skin.

The main differences are too much white, item list and feeds sidebar are much too dense, and there’s a lot of unnecessary visual clutter (e.g., stars on every item). It’s hard to believe that the same company that designed a clutter-free search page and it’s highly efficient and readable search results format came up with this.

Apparently a lot of people like this density and visual style, but for me it makes my eyes glaze over.

Google Reader – With a Simplified New Skin

Once I found that I could do some fine tuning of the Reader skin using the Reader Plus extension, I focused on opening up the spacing in the sidebar and item feeds, removing unneeded stuff (stars, folder icons), and refining the fonts types, font sizes, and font colors. The changes to the sidebar are more dramatic than the ones to the item lists.

Here’s Google Reader after I’ve applied the custom CSS to make it look more Netvibes-like.

Obviously, Google Reader still looks mostly like Google Reader, even after applying 40-50 new bits of CSS. But still, the new bits change the layout just enough to make Reader a lot more readable – at least for me.

I’d like to (and probably will) do more with it as time goes by, but at least this gets me to a place where I can scan Reader every day without too much eye strain.

What’s Still Missing from Google Reader?

Even if I tweak the CSS for a while longer, there are some things i can never get to with Google Reader. They require real programming and/or maybe not even that will do it.

One key thing that can’t be changed via CSS is the lack of date headers in the item stream. I really like having date headers to separate out one day’s items from the next. Netvibes has beautiful fat date headers. My favorite iPod Touch RSS Reader, an app called ‘Reeder‘, also has very nice date headers that float along as you scroll thru the feed items.

The other thing I’d really like to have are favicons in the item streams in folder views. Netvibes does this nicely and I think it helps with readability. I’ve put in a request to the ‘Reader Plus’ developer to possibly add this. He’s got something close already, but it requires a bit of tweaking to show just the favicons only in the folder views.

Google Reader – CSS for Custom Skin

I’m not sure the image above does it justice, but if you are interested to give this layout a try yourself, please install the Reader Plus in your Chrome browser.

Then go to the options screen and choose ‘Themes/skins’ > ‘Relook’. In the textfield that opens up, please paste the following CSS (delete what is there):

Switching to Google Reader

Back in March, I did a detailed comparison of Windows RSS readers and concluded that Netvibes Wasabi and Google Reader were the two best feed readers.

For desktop viewing, the main tradeoff at that time was between design / readability and feed handling. Google Reader was hands-down better at handling a wide range of different types of feeds. Netvibes was, to my eye at least, the hands-down winner in terms of design and readability.

Despite the problems with the feed handling, I ended up sticking with Netvibes.

But as of yesterday, that’s changed: I have now permanently switched over to Google Reader and dropped Netvibes.

The reason for the change was three-fold:

• I have some new responsibilities and simply can’t work around Netvibes feed handling issues any longer
• I found a Chrome extension that allows you to easily create your own skin for Google Reader, so I added some custom CSS and now have a Netvibes-type layout (see separate post)
• I’m need better mobile access to my feeds and that favors Google Reader by a lot

The feed handling piece deserves some further discussion. Turns out that I am putting some abnormally tough demands on the readers because I want to use messy feeds generated by RSS feed remixing services like Yahoo Pipes, FeedRinse, and Feedweaver.

As I noted in March, Netvibes handles normal feeds that most people are going to be using perfectly well – a CNN feed or a Techcrunch feed, etc. But it is a lot less effective than Google Reader at handling the funky feeds these remixing sites pump out.

To be clear, these sites pump out feeds that are often not truly ‘valid’ as defined by the W3C. Many of them fail the W3C Feed Validator test. In some cases, e.g., FeedRinse, the Validator isn’t even really able to test them, but nonetheless, Google Reader cranks out these feeds out reliably where Netvibes will not. In the case of FeedRinse, Netvibes often can’t even determine there’s a feed there to be read.

Having used Netvibes for many years, I have something of an emotional attachment to them, so I’m sorry for the parting, but so it goes.

Netvibes Wasabi RSS Reader:
The Best Feed Reader?

It all started when I became aware of the new Netvibes Wasabi edition. This version was released last fall (here and here).

Wasabi provides an alternate format for viewing feeds within Netvibes. The new Wasabi reader is more in the style of Google Reader as compared to the ‘start page’ (or ‘widget’) style Netvibes reader I’ve been using for some years. You can switch back and forth between the ‘widget view’ and the ‘reader view’ with a click of a button.

While my initial reaction to Wasabi was very positive, I had some problems with my feeds. Thus began a month long re-look at all the major Windows RSS readers to see if there was something better.

At the same time I starting looking into the top feed readers, I also cleaned up and expanded my custom feeds. These include feeds generated by Yahoo Pipes, Google Alerts, Google News, and Twitter Search. Turns out these custom feeds are fairly problematic for many feed readers and were the stumbling block that got me looking elsewhere beyond Wasabi.

Here’s a recap of what I found about the best RSS readers for Windows.

The Most Popular Windows Feed Readers

There are quite a few Windows feed readers available, but the main options I looked into were:

• Netvibes – Wasabi edition
• Google Reader
• Bloglines – beta edition
• FeedDemon (by Newsgator)

There are some other readers out there, but none of these seemed compelling enough to try. A few others I glanced at included: FeedReader3, RSSOwl, and BlogBridge. All of these are Windows desktop readers, but browser-based readers seem the way to go.

Windows RSS Readers Comparison

So with the top 4 feed readers, I did a deep dive review of their various features. I installed and configured each of them with my 100 or so feeds. I used an OPML import to get things going and then spent a fair amount of time with each reader customizing the layout and setup. Among other things, I worked on getting the individual feeds and feed groups organized in a particular (non-alphabetic) order that I like. I also spent a fair amount of time using each reader, though Netvibes and Google Reader got the most playing time.

The main things to look for in a feed reader are: feed handling and readability. In addition there are some secondary features worth looking at, such as sharing, saving, tagging, and feed importing/exporting.

Comparison of Windows RSS Readers

Features	Netvibes Wasabi	Google Reader	Bloglines Beta	FeedDemon
Cost	Free	Free	Free	Free
Type	Browser	Browser	Browser	Desktop
Feed Handling
Normal feeds
Yahoo Pipes feeds
Google Alerts / News feeds
Twitter feeds
Readability
Layout / readibility	most readable, good use of color, lines	simple, functional design, least readable for me	nice design touches, 2nd most readable	could be good – but the chiclet count is off the charts
Viewing options	3 – list, expanded, mosaic	2 – list, expanded	3 – list, expanded, hybrid	3 – list, summary, expanded
Organizing feeds	any order (Ajax)	any order (Ajax)	fixed order, alpha only	fixed order, alpha only
Themes / skins	many skins, but can’t change feed/item colors	none	none	1 skin, but nice options to change feed/item colors
Other Features
Sharing / Saving / Tagging	email, social sites (FB, Twitter), save (no tagging)	email, save, tag (no social sites)	email, save (no social sites, no tagging)	email, save, tag (no social sites)
OPML Import / Export	/	/	/	/
Pros & Cons
Pros	most readable / best design by far handles most feeds cleanly / fast ‘widgets view’ option with huge library of widgets	handles any type of feed fast feed updating rock solid, reliable can be used to feed a lot of smartphone mobile RSS readers	2nd most readable layout 3-pane viewing option is nice	nice transitions between feeds some nice readibility touches
Cons	slow updating of ‘non-standard’ feeds (e.g. Pipes) sometimes requires page refresh to update feeds hot links to original posts too small, bad position still a work in progress	overly functional hard to read	worst feed handling – even a lot of ‘normal’ feeds didn’t work or updated very, very slowly no ‘refresh’ button to update feeds can’t reorder feeds no OPML export	can’t reorder feeds way too many chiclets & icons not browser based – items open inside FeedDemon ‘browser’ (not good) or in your regular browser

What’s the Best RSS Reader for Windows?

After way too many hours fiddling around with this, I think the answer is ‘none’ – in the sense that not one of the 4 readers did all the key things you’d like it to do. Two of the readers – Netvibes Wasabi and Google Reader – are clearly way ahead of the pack. They win top honors, for sure. Which one of these two is best for you depends on your specific needs.

Netvibes Wasabi is the most readable, user friendly feed reader of the bunch – by far. It is uses colors and lines, font treatments, skins, etc. to make reading feeds a (comparative) pleasure. On the downside, Wasabi feed handling isn’t rock solid. Wasabi does a decent job with most feeds most of the time, but now and then it can be slow to update mainstream feeds (like Techcrunch last week). Plus, it definitely struggles with some of my custom feeds, especially feeds off Yahoo Pipes. I’ve been in touch with their tech guys and I know this is a recognized problem that they are working on. If/when Netvibes delivers bullet proof feed handling, Wasabi will be the top reader for most everyone. In the meantime, I’d recommend Wasabi as the best feed reader for anyone who mostly or exclusively uses ‘normal’ feeds.

Google Reader shares the top honors with Wasabi – and it is the most popular reader in terms of numbers of users. Google Reader delivers bullet proof feed handling – but the big issue I have with it is readibility. Some folks may prefer the stark layout, but I find it very tough to pour through a lot of content on Google Reader. The layout is either over-dense (‘list’ view) or it is not dense enough (‘expanded’ view). The overly functional design just doesn’t offer up enough in terms of readability features. Beyond readability and feed handling, Google Reader also has another huge thing going for it: many mobile RSS reader apps (e.g., Newsgator’s NetNewsWire for iPhone/iPod Touch and most Android RSS readers) use Google Reader as a launching point. They pull feeds off Google Reader and repackage them for delivery onto your smartphone.

FeedDemon is the next best RSS reader. It is outstanding when it comes to handling feeds. From my testing, it appeared to deliver feeds as well as Google Reader. But on the downside, the fact that it is a separate program is a real drag. FeedDemon has a sort of built in ‘browser’ to view the original post on their live sites, but the browser isn’t that great, so you end up having to switch back and forth between your browser and FeedDemon. That’s no good. In terms of readability, FeedDemon is in the middle of the pack – some good features, some not so good.

Bloglines Beta, while it has a nice pretty nice design – the 3-pane viewing layout seems promising – so I rate it as the second best in terms of readability. But the killer is that Bloglines just doesn’t handle feeds that well. It wouldn’t display any of my custom feeds at all – Pipes, Google Alerts, Twitter, etc. Worse, it isn’t even able to deliver a lot of normal feeds like the Google Blog or this blog in a timely fashion (e.g., when I posted this article, the other readers had the feed item within <2 minutes - Bloglines? not so much). I'm not sure how this could be, but adios Bloglines.

Bottom line

Netvibes Wasabi and Google Reader are the top two Windows RSS readers, for sure. However, since neither is the perfect complete package, you may find that you need to maintain them both — especially if you are a heavy consumer of RSS fees. Netvibes Wasabi gives you the polished, user friendly readability features you want while pouring through piles and piles feeds day after day. But the rocky Netvibes feed handling means you want Google Reader at hand in case there’s a glitch.

As well, you may have to have Google Reader around to power the mobile RSS reader you use on your smartphone. If you are on an Android or iPhone, it’s pretty likely that your mobile reader is going to pull your feeds off Google Reader. Blackberry not so much.

Drupal Related Content Modules

I came to use these two after trying out a whole bunch of Drupal related content modules. Check here for a good review of the related content choices – there are a lot of them. And here’s another useful, but less comprehensive review of Drupal related content modules. I suspect neither list is actually completely complete.

Most of these modules work off taxonomy terms in various ways. Some are more widely used than others and/or better supported. Interestingly, none seems to have really gained traction as the ‘go to’ module for related content.

I went with Similar Entries because it develops matches off the title and the body content vs. just taxonomy terms. In general, this seems like a good idea to me. As well, at the time I wasn’t using tagging on my sites, so Similar Entries was one of the few modules that would work at all. Now I do use taxonomy, but I’m not a prolific tagger, so I still like the idea of matching off titles and body content.

The module works pretty easily out of the box. It has relatively few settings, but has enough to get the job done well. Configuration involves activating a Similar Entries block somewhere on your page – then you check off several configuration options. These are related to teasers, to search engine handling, to the number of matches to display, and to the types of nodes used in generating matches (you can exclude some types).

I used this latter option – type of nodes to use in matches – to help solve a problem I was having with the module: from time to time, it would offer  up matches like my privacy policy or terms of use pages. This was absurd, but it was quick work to fix this by defining a new content type and reassigning these types of pages to that new type (using the Nodetype module). Then I excluded that node type from the matching process and, viola, worked beautifully.

The most complicated configuration option has to be with how (if at all) you want Similar Entries to factor in taxonomy tags. If you choose one of the tag-dependent options, the module will determine best matches by scanning title / body content within a relevant set of tags.

Node Recommendations is a newer module and has very few users, but it is very powerful and flexible. At the same time, it is considerably more complicated and requires more energy to implement — it sufficiently complicated that I’m still not sure I’ve got it configured correctly.

Since I operate a lot of small sites, added configuration complexity effort is not a good thing — especially since these options are spread out over multiple screens in different admin locations. It seems like a lot of presence for a relatively simple function.

Node Recommendation module supports both automated matching and manual selection of related pages. The automated matching is done entirely off tags (which doesn’t work so great on my older sites which aren’t fully tagged — or even on my newer sites since I’m somewhat skimpy with tags).

Node Recommendation also allows highly customized page-by-page matching. For each node, it allows you to override the default matching parameters and/or to hand pick individual ‘related content’ matches. This feature seemed like an awfully good idea when I had only a few sites and was obsessed about the quality of matches for each and every page, but became wildly impractical as I launched more sites. I realized I just gotta live with some imperfection.

Neither of the modules generates match results for pages generated by the Views module. Both modules work well on pages with a lot of CCK content – though Node Recommendation probably gets the edge here because Similar Entries doesn’t have as much body content to work with.

At the end of the day, the choice is clear: Similar Entries is going to be my default related content module for my Drupal sites. It generates the best match results over the widest variety of conditions. It is also much easier to administer. Node Recommendations is just a couple notches too complicated – though I will miss the option to do manual selection

Tray icons missing or grayed out

In my case, the missing icons are usually one or all of the volume icon, the network icon, or the power icon. The clock icon always shows up for some reason.

When I try to re-enable them, they are often grayed out and I cannot check the boxes. Sometimes, they aren’t gray.

After some googling, I found a solution to this missing tray icons problem – though apparently not a permanent one.

This page provides a recap of Microsoft’s official method for fixing the problem. It involves going into your registry keys, deleting the ‘iconstreams’ keys, and then restarting explorer.exe (which runs the taskbar / start menu). The fix works for Windows Vista and Windows 7.

I tried this method a couple of weeks ago – and it worked! — for a while. Unfortunately, here we are a couple weeks later and my icons are graying out one by one. Now back to where I started.

I guess what is happening is that the registry keys are getting corrupted over time.

So that means I need a repeatable way to fix this problem – ideally one that doesn’t require a system reboot as the Diurnal / Samuel’s blog methods do. System reboots just take too much time.

My new, improved methodology is as follows: download and use Microsoft’s ‘Fix it’ program which repairs the missing tray icons problem. You can download this little program off Microsoft Windows 7 missing tray icon help page.

I have saved this mini-program to my desktop and now can run it anytime I start seeing grayed out tray icons.

This method has several advantages:

1. I don’t have to mess around with registry keys.

2. It creates a ‘restore point’ during the repair procedure. This adds a measure of safety.

3. It doesn’t require a full system reboot.

4. It involves fewer steps.

One thing I have noticed is that after running the Fix it program is that the icons are not always re-enabled. In such case, I have to take a couple of extra steps:

1. Stop and restart explorer.exe as shown in step #8 of Microsoft’s manual repair process as follows:

8. Restart the Explorer.exe process. To do these, follow these steps:

1. Press CTRL+SHIFT+ESC.
2. On the Processes tab in Task Manager, click the explorer.exe process, and then click End Process two times.
3. On the File menu, click New Tasks (Run), type explorer, and then click OK.

2. Recheck the missing tray icons: right click on taskbar / menu bar, choose ‘properties’, choose ‘notification area’ tab, recheck missing system icons, click ‘ok’.

For me, I’ve found this makes things right again – and is as close to a permanent solution as I’ve found.

Windows Vista UAC Takes a Dive!

Yesterday, I disabled User Account Control (UAC) in Vista for good. Could there be a more invasive, yet ultimately useless program?

After 13 frustrating months using UAC, I still have no idea what it is supposed to be protecting me from — whatever it is, it has never, ever happened.

As part of my recent efforts to increase productivity and streamline workflows, I realized it was time for a change yesterday when, for the nth time, I tried to figure out what UAC is supposed to do — and I couldn’t.

In my search for answers, I found only techie jibberish or vague explanations, such as this from Microsoft’s website:

User Account Control (UAC) is a feature in Windows that can help prevent unauthorized changes to your computer. UAC does this by asking you for permission or an administrator‌ password before performing actions that could potentially affect your computer’s operation or that change settings that affect other users. When you see a UAC message, read it carefully, and then make sure the name of the action or program that’s about to start is one that you intended to start.

Uh, huh. I suppose it sounds vaguely useful – until I realized that not once in the 13 months has it ever activated when I haven’t been explicitly trying to do something. Unlike my Norton Anti-virus program which is constantly (and more or less silently) picking off various threats that are invisible to me. UAC is constantly stopping me from doing things that I purposefully want to be doing. How is that helpful?

Even after trying to understand, I still don’t really quite know what the point is — or what some relevant ‘use cases’ might be. Is it to stop my cubemate from tinkering with my computer? Except I don’t have a cubemate. Is it supposed to prevent viruses that somehow get past Norton from doing more damage? Except that has never happened.

Looking back, I’m not sure why I soldiered onward with UAC for so long given how much time I’m on my computer — or how bogus UAC alerts I’ve clicked thru. I think I’d have ditched it much earlier if I hadn’t been using Norton Labs’ UAC tool. That program has probably cut my warnings by 90%, but even the remaining 10% is too many.

So when I paused and realized that it had never, ever stopped anything I hadn’t wanted done, I realized it was time to say…

Bye, bye!!

Click here if you want to turn off UAC, too.

Google Chrome launches extensions

I’ve been using the Chrome Developer version for many months – it was the only version that worked well with my dual monitor setup.

In December, Google launched an apps store for Chrome – except that they call apps ‘extensions’. I’ve been using these extensions for several weeks and LOVE them! but they haven’t been available in the general release versions of Chrome.

Today, Google unveiled a new general release version of Chrome – now extensions are available to everyone. You should definitely try them out. Quite fantastic stuff!

My favorites are:

• xmarks - manage bookmarks across browsers and/or computers located anywhere
• lastpass - manage passwords and login info across browsers and/or computers
• evernote - the official extension for the evernote note taking / web clipping service that is all the rage (and deservedly so)
• delicious - beautifully designed widget for adding bookmarks to your Delicious account. cleaner and better than the Firefox version.
• netvibes - auto-detects RSS feeds in Chrome and provides one-click button to add it to your Netvibes RSS reader (still only available for developer version, i think).

These are essential in my book.

I’m also trying out the aviary, picnik and webpage screenshot extensions for taking screenshots. They all seem to work fine, but not sure I am 100% sold on any of these yet and I may be sticking with jing, a  program I’ve been using screenshots for a while.

I have also been using one of the several Page Rank tools – Kuber Page Rankchecker. This seemed the simplest and best of the bunch. Some of the others seemed to be trying to do too much and I worried about slowing down my page loads while they collect all their data.

I haven’t been too happy with the RSS auto-detect extensions for Chrome. Unlike other browsers, RSS auto-detect isn’t built in, so you need an extension to handle that. The existing ones all seem geared to plugging the feed straight into your reader, especially your Google Reader (which I don’t use – I prefer Netvibes, especially since they launched their new Wasabi Reader which is fantastic). I’d like a Chrome RSS extension that simply does auto-detect and then shows you the actual feed url.

The Chrome Firebug extension isn’t where it needs to be – or at least it wasn’t a couple of weeks ago. The Firefox version wins hands down at this point and is largely why I continue to use Firefox very heavily. Hopefully, the Chrome version comes along and becomes a near clone of the FF version.

I also hugely, hugely like how all the Chrome extensions present in the exact same area in the browser frame – vs. Firefox where they can be located in any of at least 4 places in the browser frame. That’s very confusing and means you need to remember exactly what extensions are located where – and I’m not smart enough to remember stuff like that.

I do notice that with a pile of extensions running, Chrome doesn’t seem quite as light as it did before. I believe each extension opens up a new process on your computer, so now I’ve got a lot more Chrome processes running and consuming memory. That’s a potential issue that they maybe need to address – trimming Chrome memory consumption.

Links:

Find Chrome extensions

Download Chrome

Some other bits about this from ReadWriteWeb and Techcrunch.

Comment tracking & aggregation

But how? The answer: Backtype, a free comments tracking service. Backtype describes itself as a “real-time, conversational search engine” because as a result of tracking people’s comments across the web, it also includes features that allow you to search web comments in a way that you can’t with Google. I don’t have too much interest in the searching function, so the value-added function for me is having Backtype track and catalog my comments on the web.

Backtype works by scouring the web, indexing all the comments people make on blog posts. It doesn’t spider every single blog, but covers all the major ones and lots of not so major ones, too. It also seems to index blogs of any size that reside on the major hosted blogging platforms – e.g., WordPress, Blogger, Typepad, Posterous, etc. That’s why the comments on my tiny, tiny personal blog, John’s Bytes, are getting indexed.

Once you register and identify yourself, it keeps track of your comments on a personal webpage – here’s a link to my web comments. It does this on a continuously as it tirelessly scours the blogs. New comments are posted to your Backtype page on reasonably timely basis – usually within 24 hours, often much faster.

Backtype seems to identify your comments by tracking against the URL you enter when you leave a comment. It is not clear to me whether that domain name has to be unique to you or can be something that multiple people might use. In other words, using my own domains (like this one – www.WebBizGeek.com) works fine, but what if you worked at Fidelity or some other company and didn’t have your own URL? I’m not sure how that works. Perhaps by tracking against both a URL and your ID / username, it works out.

If you aren’t entering a URL when you make a comment (for instance, in many forums), then Backtype seems not to have a way to trace comments back to you. That means that a lot of comments I make – e.g., in the forums on drupal.org, openx.org – aren’t tracked. That’s sort of a disappointment, but perhaps they will eventually figure out how to track those, too.

Backtype is similar to another service called CoComment. Both are widely used. I’ve not tried CoComment, so can’t say which is better. Maybe I’ll give it a shot as some point just to see.

As an added bonus, today I was able to integrate my Backtype comment stream directly into my blog here: my Backtype comment stream. I did so using a WordPress plugin called My Comments Elsewhere. Installation was quick and straightforward. The plugin supports both Backtype and CoComment. You need to have an account with one of the other since it relies on those services for the actual comment aggregation.

If you want to keep track of your comments, too, I recommend you give Backtype a try.