September 15, 2009

My long suffering blog


Wow. Long time no blog. Twitter has killed blogging. Well my blogging anyway. HAving coffee with Andrew Anker (no Twitter feed) vp corp dev moveBle type tomorrow so I will bend his ear. I am sure they have an answer. Fuck I hate the iPhone non existent keyboard.

But the app platform is why I have this paperweight. Downloaded blogpress to do this. Hooked right up to My private install of mt. Very slick. I was worried they would only support the hosted stuff.

Blog please forgive Twitter. She is a demanding mistress

I'm Martin Tobias and I endorse this message.

Posted by Martin at 11:29 PM | Comments (0) | TrackBack

October 14, 2008

Problems with the MT 4.21 upgrade

It was a painful process as always. Here are the problems I had that were not adequately addressed in the documentation:

1. I am using a windows client to FTP up the files. In Windows there are just check boxes for the permissions on the .cgi files so when MT says "set permissions to "755", that doesn't help me as a Windows guy. I had to get my ISP to go into Unix shell and set the permissions correctly. I know some windows FTP third party clients support the UNIX number designations on file permissions but Vista does not. Come on MT, help a poor Vista user out.

2. My MT install is on a box that hosts multiple domains. MT still doesn't work well out of the box if you have multiple domains, especially if you are publishing blogs to different domains than your cgi-bin and \mt-static directories. The style sheets didnt apply the styles correctly. I just got one screen, with everything wrapped and no formatting. The support forums (very helpful) pointed me to the styles.css file and a path entry for MT-Static. But no-one there said exactly how to reformat this path entry to point to a different domain.

3. Getting the Styles.css to get formatted with the correct domain is a nightmare. You can manually edit it in the template edit screen (to a fully qualified path), but the "smarter than you style setter" overrides that in the background and resets it to the default path (not fully qualified). This took me a couple times testing and changing to figure out that the system was doing and why the published file didn't match the source screen. So now I have to figure out how to move MT-Static to whereever the system assumes it should be for the other domain. Finally had to change the styles.css global template to have th mt:StaticWebPath variable in it and then figure out what that is (there is no clear way to set it in MT 4.2, you used to be able to edit the .cgi file). then I moved a copy of MT-static directory to that place, and Voila it works!. Ugly. I hope I remember to replace that on the next upgrade...


this is too hard.

Posted by Martin at 8:28 PM | Comments (1) | TrackBack

March 10, 2008

trying to fix commenting

So commenting hasn't been working on my blog. sorry. I finally (with the help of a product manager at Six Apart and reading some code) figured out what variable wasn't being set right. so I had to edit my mt-config.cgi file to add a fully qualified URL to the CGIPath variable.

CGIPath http://www.nwventurevoice.com/cgi-bin/

Prior it had been relative:

CGIPath /cgi-bin/

which sent MT looking for the comment.cgi in each blog subdirectory instead of where the code actually resides. I think that will solve it, but now I have to regenerate the entire site.

Posted by Martin at 6:11 PM | Comments (1) | TrackBack

January 17, 2008

Finished MT upgrade and site move

Just today I finally got the last of the www.martinandalex.com directories off the server and out of my code base. The blog is now fully hosted only at www.deepgreencrystals.com. Just changed the Google search in the upper right. The last problem I had was with getting MT-xmlrpc.cgi to work correct from w.bloggar and scribefire. After alot of searching on the support sites, i find that you have to delete or rename mt.cfg. it is a configuration file used by MT 3.15 and not 4.0. Why didn't the MT-Upgrade.cgi script do that? And why did the failure of the system expose itself as an authority problem with mt-xmlrpc.cgi when the config file was there? Yet another sign I am afraid that Movable Type has lost it's way on server software. Unfortunate.

Posted by Martin at 6:57 PM | Comments (0) | TrackBack

August 28, 2007

deleting Scribfire

Rich turned me onto it, but I am not a geek enough to continue using it. Rich likes early software with lots of bugs. I don't have the time for it anymore. Scribfire could not retreive my categories from Movable Type and there was no answer to my question on their support forum and no obvious config file to edit, so asta la vista Scribfire. Welcome back old friend w.Bloggar.

Posted by Martin at 10:42 AM | Comments (0) | TrackBack

March 20, 2006

Trying Ecto again

It has been awhile since I played with blog tools. I use w.bloggar on my laptop. I tried ecto once, but didn't like it any better than the free w.bloggar. Ecto is now a 21 day trial and costs $17.95 a copy, but has continued to receive good reviews, so I am trying it again, this time on my Mac Mini.

The thing looks like iTunes. Still getting used to the "mac thing". Bonus points on start-up for figuring out my mt.cgi location automagically and connecting to the server. I like the easy access to Amazon links. First cool feature: I went to a Safari window and found my old Ecto post link, copied it in the browser then came back to Ecto. In ecto, the edit menu has option to "create URL link with clipboard". Kabam there is the link in the post. Very cool. Would be nice if I could search for prior posts within ecto automagically to relink to, but that is a nit.

Still playing, haven't decided to buy, but so far so good. And it is on the MAC where it was designed for. That makes a huge difference.

Posted by Martin at 7:08 PM | Comments (1) | TrackBack

October 6, 2005

AOL buys Weblogs Inc.

Congrats Jason and Mark. Not alot of total payout ($25M) but alot relative to the cash going in. AOL Buys Weblogs to Boost Blog Presence - Yahoo! News

Posted by Martin at 11:09 AM | Comments (0) | TrackBack

July 13, 2005

Google adsense adding blog feed ads

It was just a matter of time: Google AdSense for Feeds. I am going to try it. Look out FeedBurner.

Posted by Martin at 7:39 AM | Comments (1) | TrackBack

June 30, 2005

check out 43 places

The smart boyz and girlz over at 43 things just keep coming up with new ways to slice and dice their community of interest. They have a beta version sliced by places here: 43 Places. The part I like the best is the indexing of Flickr photos by place. It seems to do a good job with the exception of a bunch of Buenos Aries photos in the Seattle section. It was really cool to flip through images of my own city taken by other people. It was even MORE fun to flip through images of cities where I may be traveling. These are not your typical sanitized press shots. These are just real people with digital cameras and camera phones. Way better.

Not much text there, but I expect that to change as people start annotating their entries more. A cool concept. Thanks Josh!

Posted by Martin at 4:03 PM | Comments (0) | TrackBack

June 12, 2005

Trying MyBlogLog

The Blogsphere is GREAT! Anyone can write any little thing that could be useful and market it very easilly. Just e-mail people and ask them to use it. That is how I got to know about: MyBlogLog. Eric sent me this little note:

Martin --

I was hoping I could interest you in trying out MyBlogLog, a service which tracks the links that people click on to *leave* your blog. With MyBlogLog, you would know how many people you sent this week to the oilfacts pdf.

Registration and implementation is extremely quick -- five textboxes (Blog name, URL, email, password with
confirmation) and then a single line of javascript to embed in your blog template. Republish and you're immediately tracking. Under 30 seconds from start to finish.

I look forward to hearing your thoughts on MyBlogLog.
We're constantly looking for ways to improve the service for our users. Please contact me if you have any questions.

Cheers!
Eric

So of course I had 30 seconds. Here is what I found:

1. I love the simple interface. Five fields and you get the javascript to plug in. I think I did the whole thing in 20 seconds.
2. The Top Links feature that puts your top links out of your site into a side bar looks cool, but there is probably some delay in the server. I have 5 links out when I go to the mybloglog.com site, but TopLinks Javascript doesn't display any. I will give it a while.
3. I already have SiteMeter and have been happy with it, but it is WAY too much information. They could do this, but don't.
4. I like clean and simple little utilities that are 100% web services. Have no idea how Eric plans to make money (well he does try to sell me a "pro" version), but I like the free little utility for now.

Thanks Eric!

Posted by Martin at 5:33 PM | Comments (4) | TrackBack

Farting around with Text Pattern

I saw that 43things uses this for their blog. They are smart guys: Textpattern I am going to give it a try. So far, installed the code on the server, a snap. It was a bit tricky to find the main admin page. It is index.php in the textpattern subdirectory. There is also an index.php in the main site directory which is confusing. Not excited to replicate all the formatting work I have done. But I think this one will let you skip around in sections much easier. I need to replace the hard coded site for my foundation.

Posted by Martin at 12:17 AM | Comments (0) | TrackBack

May 16, 2005

Social networking software...

I am always interested to see how an hour conversation with a reporter gets parsed in a story. Recently, I spoke to Josh Jaffe at The Deal about social networking software. My comments got wrapped up into a wide ranging story from Blogs to social networking to collaborative software to wikis here: Blogs and social networks and wikis, oh my! | Josh Jaffe The Deal. If I knew that I would have had more to say. The thrust of my conversations with the reporter were around how collaborative architectures are uniquely able to solve many of the emergind distributed problems on the internet. Like SPAM, viruses and fraud. Having over a million people "sensors" voting on what SPAM is gives Cloudmark a level of intelligence and responsiveness that simply is not possible with any other single CPU based algorithm. I believe this is a fundamental, powerful and underappreciated aspect of distributed network collaboration architectures.

Josh got my comments right, but hid them in a long article. Nice to see my friends Sifry and Ross in there though!

Posted by Martin at 10:14 PM | Comments (1) | TrackBack

March 31, 2005

thinking about making $$ on blogs

I have been thinking about how bloggers make money lately. Thanks Eric for the pointer to Matt Haughey's A Whole Features: Blogging for Dollars. The story of PVRBlog is pretty impressive. Matt has a good guide for how to make your site more relevant and more targetable by Google Adsense. I have been surprised by my own results with Adsense. I am not going to Vegas on it, but I am paying the hosting fees many times over.

The real issue I have with the current generation of mass market ad networks though is the cost to run them. The publisher gets the VERY short end of the ad dollar. Google pays about 35% of each ad dollar they take in on Adwords to their publishers in Adsense. Last year they pocketed over $650M on the backs of publishers. Is that fair? Doe sit cost $650M to run the network? Today they are the only game in town (that works and has scale). But how long will the masses (you and me) live on the crumbs? sounds like an idea for a company to me.

Posted by Martin at 10:15 PM | Comments (0) | TrackBack

March 15, 2005

Sifry on blog posting volume

More from Sifry: Sifry's Alerts: State of The Blogosphere, March 2005, Part 2: Posting Volume. This is no surprise to me. Spikes when big events happen. Then drops. But overall up and to the right increase in post volume. The interesting thing to track would be the interpost percentage. For example, this post is really a reference to another post with only a small amout of value added. With more blogs and everyone reading the few important ones, how many of the post are self referential and how many are truly accretive? (ok I can't spell).

Posted by Martin at 11:09 AM | Comments (0) | TrackBack

March 14, 2005

Sifry on the state of the Blogsphere


When I first met David Sifry, he was running Technorati on a box under his desk. Today the service is tracking 40,000 new blogs a day totaling over 7.8M. The blogsphere is doubling roughly every 5 months. He recently blogged some of his analysis of this growth. I was struck by the following excerpt, specifically the comments on Spam. Not only is comment spam and trackback spam a problem, but now bots are actually creating WHOLE BLOGS solely for the purpose of creating links among themselves and a host site. Everyone tries to game the gamer. So Technorati is spending an increasing amount of their time trying to purge their indexes of links to/from spam blogs. I wonder how much time Google spends on that? Or how spam links are accounted for in the Google Pagerank?

We are currently seeing about 30,000 - 40,000 new weblogs being created each day, depending on the day. Compared to the past, this is well over double the rate of change in October, when there were about 15,000 new weblogs created each day. The remarkable growth over the past 3 months can be attributed to the increase in new, mainstream services such as MSN Spaces, and in increases of use of services like Blogger, AOL Journals, and LiveJournal. In addition, services outside the United States have been taking off, including a number of media sites promoting blogging, such as Le Monde in France.

There is a dark underbelly to these numbers, however: Part of the growth of new weblogs created each day is due to an increase in spam blogs - fake blogs that are created by robots in order to foster link farms, attempted search engine optimization, or drive traffic through to advertising or affiliate sites. We have been battling the spam situation in a significant way for about 2 months - prior to January, spam wasn't much of an issue. All of these charts reflect Technorati's databases after spam blogs have been removed, and we feel that we've been able to capture and identify most of the spam out there, but one should note that there is definitely blog spam that we don't catch (tell us if you see spam in the index!). I'd estimate that we currently catch about 90% of spam and remove it from the index, and notify the blog hosting operators. Most of this fake blog spam comes from hosted services or from specific IP addresses. One of the results of the extremely productive Spam Squashing Summit of a few weeks ago is the increased collaboration between services in order to report and combat this spam. Right now, about 20% of the aggregate pings Technorati receives are from spam blogs, so you won't see that in these numbers - these statistics show only "cleaned" data.


Posted by Martin at 7:32 PM | Comments (0) | TrackBack

March 3, 2005

hey mom look I have an Alexa rating

I wonder if adding this is useful?

I don't really see the need since it is hard to gauge the absolute meaning of the number. Maybe if people click through you make few more $$ though. I know Alexa does a good job tracking the top sites but I don't know if they are doing a good job providing services to the masses. Wouldn't you think they would be interested in that?

Posted by Martin at 10:41 PM | Comments (0) | TrackBack

My mother the blogger...

Well my mother retired at the end of February. She has been a reporter for a New York Times owned paper in Florida called the Ocala Star Banner. When she retired many people e-mailed her and said they would miss her saturday morning weekly column. So my brother set up a blog for her to keep writing! Saturday Mornings with Lucy. Now her fans can continue to read and she will own the copywrite. All goodness. The ease of micropublishing like this, especially for established writers like my mother, will open up whole new markets for these people.

Posted by Martin at 7:27 PM | Comments (1) | TrackBack

February 25, 2005

shameless self promotion

Well new approaches to being a VC (including blogging) are becoming a hot topic around the water cooler. Today's WSJ business section had an article on the subject. The subscriber only link is here. For those of you less who save your daily $$ for Starbucks, I have excerpted the relevant shameless self promotion of DGC:

"In another sign of their openness to new ways of doing things, all but one of Ignition's 10 general partners have launched Internet "blogs" for regularly posting thoughts. Among the more popular are Martin Tobias's "Deep Green Crystals" and Rich Tong's "geekfishing.net," which offers an inside look at everything from new technologies to what entrepreneurs should look out for in striking deals."

Of course lots of other nice things were said about being a start-up VC and how Ignition's approach is somewhat different. But regular readers of DGC already knew that didn't you?


Posted by Martin at 7:24 PM | Comments (1) | TrackBack

February 7, 2005

Another newbie starts making money

thanks Troy for the pointer to tomEvslin's blog: Fractals of Change: Empowering the Wannabes. It seems that Andrew has gotten the Amazon Associates integration with Typepad done. At one point I thought that there might be a middle ware play to be made linking up the non-programming masses with Amazon and Google. But the non-programming masses are on the hosted code bases of Typepad, Blogger, etc. and those guys will do their own integration. Not enough self-code hosters left like myself. The post made me check my own Adsense account and I am happy to report that I am now regularlly over $40 per month in adsense revenue and about the same in Amazon associates. Even though I don't write that much about books. I am going to start a favorite books page though.

Oh, well, I am back from a week in Florida so I will post some more. But out next two weeks. May be a bit of a lag.

Posted by Martin at 1:44 PM | Comments (0) | TrackBack

January 18, 2005

the difference between Technorati Tags and Del.icio.us and PubSub and google.

So alot of people are trying to put a meta layer on top of the web and blogs and other types of data. Technorati with Tags is trying to aggregate blog entries, Flickr photos and del.icio.us entries. Technorati watchlists only search for the keyword you want in blog postings (just like typing in biodiesel on their search box). Del.icio.us serves up web and blog matches in reverse chronological order (and will automatically generate an RSS feed that matches). Del.icio.us only gives you bookmarks that its members have tabbed, they don't do any crawling. Then PubSub gets all the pings and does keyword searching much like Technorati (although the results are different yet again). Google takes web sites and blog entries (no flickr) and applies their page rank to the results.

Technorati: Tag: biodiesel

Technorati Biodiesel Keyword search

del.isio.us tag: Biodiesel

Pubsub matches for BioDiesel

Google search for Biodiesel

So, which do I like better? Well it depends on what I am looking for. If I wanted a javascript sidebar that shows me the latest news/posts about biodiesel, I would stick with Technorati or PubSub since they do a better time with the real time posts. If I were looking for a good list of general biodiesel resources, especially the most authoritative SITES, I would use Google or maybe del.isio.us. Would be nice if I could take any or all of these lists and have them auto javascript sidebared for me. That would be a cool service.

Posted by Martin at 1:27 PM | Comments (0) | TrackBack

Trying out RSS Digest

Trying out RSS Digest: RSS Digest: Show and syndicate RSS and news feed on your Web page with no programming. Basically the site can create a javascript include which will reformat an RSS feed. In a two step process I did this for B100Fuel.com.

Step one. Go to Del.icio.us and create an custom RSS feed on the keyword you want. In my case I created one for "biodiesel". The URL for the RSS feed is: http://del.icio.us/rss/tag/biodiesel

Step 2. paste this URL into the window on BigBold. Click "generate code". You get Javascript, PHP, or iFrame code. Just copy and paste this code to your template in your blog. Check it out on B100fuel.com.

I love web services!

Posted by Martin at 10:04 AM | Comments (1) | TrackBack

January 17, 2005

Technorati launches "tags" but makes them more complicated than they have to

Dave sent a mail announcing Technorati Tags today. Here is their help page. Technorati: Using Technorati Tags. It sounds like a good idea to have users set their own metadata tags and let the server worry about normalization of the database. I have long been against heavilly structured metadata systems that require everyone to use a pre-defined list of tags. Tagging is a pain enough without having to live in someone else's taxonomy. So Self Taging is good. Then apply some server logic to figure out which tags are probably related or "close enough" to allow people to find things even if they aren't tagged explicitly with the right words you are using to search.

The only gripe I have is that the presenation of the idea was overthought and shrouded in a bunch of techno babble. I got the impression from all the various threads that I had to start doing something new, adding a new tag to the HTML of my posts. But after further research, what they are really talking about is sucking up the category tag and allowing users to search on categories. MT already has tags if you are categorizing your posts! So I don't have to do anything new.

Now on the search part that would be cool. What I would like is a javascript that will pull down posts, pictures, etc. not on keywords but on "tags" or "categories". for example. BioDiesel. What if I wanted B100Fuel to have an auto generated list of posts from around the web on biodiesel? But not every relevant post has the word biodiesl in it. It may be just a pointer to something diesel related. Or crop related. Or tax realted? So working at the category level would be good. I will poke around for that. I bet it is out there somewheres.


Posted by Martin at 2:01 PM | Comments (4) | TrackBack

January 6, 2005

I blog therefore I am

Nice article on Venture capitalists that blog today. Venture Capital Journal And yea I am in it.

Posted by Martin at 3:20 PM | Comments (0) | TrackBack

January 5, 2005

Six Apart buys Live Journal (rumor)

Here is a juicy one: Om Malik on Broadband � Six Apart to buy Live Journal. Live Journal really doesn't have any options other than getting funding and growing management themselves. With Six Apart already funded by August Capital (a GREAT VC), the shorter route to success is merger. I will back-channel this one.

Posted by Martin at 1:12 AM | Comments (0) | TrackBack

Technorati launches significantly improved watch lists

David and his merry pranksters just keep getting better. Sifry's Alerts: Get your keyword watchlists here!. Now I want that live feed to be a java script include (like a blogroll and the code PubSub lets me put in).

Posted by Martin at 1:09 AM | Comments (0) | TrackBack

January 4, 2005

Fixing my trackback spam

Going through the clean-up process on track-back spam today on Deep Green Crystals. I had over 1000. The job took me over 2 hours. It should have taken 10 minutes. I run 12 blogs. This management task alone could cause me to publish less..

I like the tool in MT, but it is too hard to use. I noticed some interesting trends. It seems like many of the track-backs were done by the same guy since they came in three waves roughly a month apart. Different IP addresses, IDs, and comments, but definitley in batches. A couple of feature requests:

1. Allow me to "ban the IP" and delete the trackback with one click. Now you have to ban the ip, go back to the screen, select the comment, click delete, confirm delete, then go back again. Too many clicks, should be one.
2. SHould be able to ban an ID as well ans an IP. The spammers change their IP addresses more often than their ID or source blogs. I want to do both.
3. Should have a power editing mode to do the job more quickly.
4. Should integrate some of the blacklist or key-word searching screens on track-back comments. There is NEVER a case when I want someone to put the word "rape" or "piss drinker" or "incest" in a trackback.
5. Since these track-backs are done in batches by a bot, the ability to select by date/time would be good. I could delete the whole batch at once.

In the end, I just requested all trackbacks and selected them all and then unchecked the good ones. That was WAY faster. But I would still like the delete function in that case to ban the ID and the IP address.


Posted by Martin at 10:59 AM | Comments (0) | TrackBack

January 2, 2005

upgrading to MT 3.14

Movable Type Publishing Platform : Support

no real features. Just want the latest. I gotta find a way to stop track-back spam though. Would be nice if they had a track-back review feature like they do with comments so you could manage them there and block the IP addresses, etc. With Jay Allen there, I am sure this is in the works sometime.

Posted by Martin at 1:38 PM | Comments (1) | TrackBack

December 28, 2004

Installing a discussion board

Well I started another blog B100Fuel.com , just what I need, more work. But this one came with some fun! I decided to add a discussion board and installed InvisionBoard. Here it is: B100Fuel.com (Powered by Invision Power Board) I even figured out the template structure they were using and inserted google adsense in there. It was totally easy using CPanel at my ISP. It took some time to figure out that a Category was the Parent of a Forum, but that was cake compared to the bugs I was expecting. So far, a completely smooth install with no configuration bugs! Can't say that about too many software installs. Especially on a linux server in Huston by a Windows guy!

Posted by Martin at 3:58 PM | Comments (0) | TrackBack

December 21, 2004

Pluck Sucks (again)

I have posted about Pluck before (back in March). At that time I was casting about for an RSS Reader. The author of Pluck was quite Plucky at the time and sent me a comment.

So this time when I was looking for a simple to use RSS reader for my team to standardize on, I gave Pluck another try (this time version 1.095.005). Unfortunately it fell short in two key areas.

1. It only works in IE. Half our team uses Firefox (which has it's own very good RSS reader integrated with Bookmarks thankyou).
2. It can't read secure pages. For my team blog there are two levels of passwords. Pluck doesn't prompt you for the password the server is asking for, it only tells you it can't find the .RDF file (404 error).

So I recommended NewsGator which works in Outlook of course, but I spend more time there than the browser anyway and yes it handles secure pages just fine, asking for the log-on credentials.

Keep working at it Pluck.... It took Microsoft till 3.1 to get Windows usable!

Posted by Martin at 8:08 PM | Comments (1) | TrackBack

my foaf

Martin's Foaf

Posted by Martin at 10:22 AM | Comments (0) | TrackBack

The best 2004 lists

This guy has been doing a good job collecting all the "best of" lists he can find. 340 this year. I wonder if you can make a business re-distributing that stuff? I wonder what the rights are. It seems to me that there is a need for agregation of content like this and re-distribution on a micro basis. While I don't like the micro-publishing model of WeblogsInc and that ilk, I do believe there may be a business in micro-web services.

Posted by Martin at 10:08 AM | Comments (2) | TrackBack

December 20, 2004

Blog beats Wiki

So after alot of gnashing of teeth, I have choosen Movable Type Publishing Platform 3.1 over a Wiki for our Ignition Team group sharing tool. Basically wiki's require you to learn a new tag format and come with varying degree of WISYWIG tools for publishing. Wiki's are also an order of magnitude harder to use than Blogs by the simple fact that they have more features. I could see using a Wiki to manage a workflow in a highly technical team instead of Outlook/exchange.

In the case of our little VC team, all we needed was a central place where we could post thoughts about companies, upload documents, have threaded comments, and simple search. These requirements do not a wiki make. The sharing tool had to be totally light weight and simple to use. The only thing I that would make this blog easier would be if I could copy it on an e-mail so that many of the internal e-mails on companies could just automatically go into the blog and be searchable. I will work on that this week.

Bottom line. For simple collab, go Blog not Wiki. I installed and ran through the features with Kwiki, TikiWiki, SocialText hosted, PMWiki, Moin Moin, and tWiki. For ease of set-up and rich features go with SocialText hosted if you must have a collab solution.

Posted by Martin at 2:02 PM | Comments (1) | TrackBack

December 6, 2004

HTML Help!

OK, I give up. I can't get the large space between the text and the table in this entry gone. Deep Green Crystals: Amazon beats E-Bay on price!. I have spent too much time on it. Any ideas? I don't see any code that says make three pages of spaces before the table. I exported the table from Excel as HTML and already took all the junk out that Excel put in. I have edited it in Frontpage and it previews ok there, but shows up wrong here. When I take the code from "< table >" to "< /table >" out, the spaces go away and the text shows up in correct order.

I am stymied...

Posted by Martin at 10:13 AM | Comments (4) | TrackBack

November 24, 2004

Funny bug in W.Bloggar or MT with < / BR>

I was posting something over at NW Venture Voice and having to edit the HTML from the e-mail that I received it in to a format that would look good in MT. After much messing around deleting unnecessary images and reducing the table width to fit in the window I have, I was stuck on the problem of excessive spaces at the top and bottom of the table. The code gave no clues. In w.bloggar I couldn't find any extra < / br > statements, or weird references to style sheet commands. When I previewed the entry in W.Bloggar the spaces were not there. When I viewed it in Firefox (or IE) they were. But I had put many commands on a single line for readability in the front and back. Like:
< html >
< body >
< table >

When I went into firefox and viewed the source of the page, somehow a carriage return < br > had ben placed after each of these commands. Even though that tag was not in the source editor I saw in W.Bloggar OR in the Edit entry window of MT. Maybe when the browser renders the page, it puts in the extra carriage returns. In any case, when you put the tags all together on the same line, the spaces disappear. Like this:
< html > < body> < table >

Now isn't that interesting? Sure hinders readability.

Posted by Martin at 10:39 AM | Comments (0) | TrackBack

November 8, 2004

Watch your < /div >s

After the election last week, I went into my MT template and removed the javascript that linked to Election-Vote.com (since it is all over). My happy cursor (unbenonst to me) grabbed an extra < /div > (had to put in the spaces so the browser doesn't interpret this one as an actual "div") when selecting the code snippet. Today, I noticed that the google ads and the All Consuming "reading" pages were not showing up. No error messages, simply white space. The only change I made was removing the election-vote code, so I go over to count the div's. Yup one missing. I put it back in and the right column shows up again. Having consumers editing HTML code is not a good idea...

Posted by Martin at 9:37 AM | Comments (0) | TrackBack

October 27, 2004

How do I comment on DGC?

I have found a work-around if you want to keep using IE and are having problems signing on to comments. The problem which I posted about before is how IE handles third party cookies. I call it a bug. Msft calls it a "Feature". Six Apart says "use Firefox".

Well if you want to stay with IE, go to Tools/InternetOptions/Privacy and click on the Sites tab. Add these sites:
martinandalex.com
nwventurevoice.com
typekey.com
typepad.com

Your IE should handle third party cookies just fine for my site (and maybe others).


Posted by Martin at 7:21 PM | Comments (1) | TrackBack

Mary Meeker on Blogs and RSS

Mary Meeker update from Digital World Covers RSS, Yahoo's RSS embrace and Blogs. Pretty meaty and tasty.

Posted by Martin at 9:23 AM | Comments (0) | TrackBack

Google site search beats MT

Ok, I am testing Google site search against the standard search that comes with Movable Type. There is a reason Google is worth billions.

I tried the query "you perform" (without quotes), looking for the software product of that name. The MT Search Results picked up 23 entries. 21 of them had nothing to do with "you perform" the software program. They didn't even have "you perform" as a string together. The entries did have the word "you" AND the word "perform" somewhere in them. But in many entries, "perform" was actually a part of "performance".

Now if you are a search junkie and think about it, you can put quotes around "you perform" and get two entries back - the right two!

But Google is more accurate right out of the box. They show four entries, two of which are for the software I am looking for. They also include a helpful hint about how they tried to make the results the most relevant:
"In order to show you the most relevant results, we have omitted some entries very similar to the 4 already displayed.
If you like, you can repeat the search with the omitted results included."

Of course with Google, if you fully qualify your search with quotes "you perform" you also get only the two you are looking for.

Looks like I will probably be ditching the MT search and going with Google only because the relevance out of the box is so much better (and I can get paid for click-throughs).

Posted by Martin at 8:51 AM | Comments (0) | TrackBack

Funny behavior on Google Adsense Search


Ok, got it working now. Initially I put in my site address as www.deepgreencrystals.com as the local site to search. But that is a re-direct to www.martinandalex.com/blog, so the Google indexes weren't finding any pages with deepgreencrystals in them. The search found no results. So at first I just hacked the javascript on my index page to have www.martinandalex.com/blog. Still no results. That is because google indexes only have the root domain. Updated the JS to be www.martinandalex.com and bob's your uncle. Search results just fineo.... now to get a logo working on the results page....

Posted by Martin at 8:04 AM | Comments (0) | TrackBack

Adding Google Adsense for Search


Been playing around with more ways for small publishers to monitize their sites. The latest method is Google Adsense for Search. I have implemented it here and look forward to your feedback. Initially I wanted to replace the Movable Type site search with Google (since they do local site search too), but Google doesn't have the ability to resize the search box very well yet. It is default of 255 px long. My left link column is 200px, so the box hung over and was covered by the entries. So for now I am putting it up in the upper right of the banner. You need to use the
< div style="float:right;"> tag to get it over there.

For now I am leaving the MT search there as well so I can compare the experiences and results. Will letcha know.

Posted by Martin at 7:23 AM | Comments (0) | TrackBack

September 9, 2004

Upgraded to MT 3.11

So I just upgraded. The new features claim to be:

1. Dynamic page building
2. Scheduled posting
3. Subcategories
4. Application level callbacks

and a not so well documented one:

- adding a ping to technorati.com as a check-box under pings.

What I am going to use:

The Technorati trackback and some comment subcategories (later this week).

I tried to turn on dynamic archive generation and am getting tons of errors. When you turn it on, it renames your index.html files under each category and date subdirectory to index.html.static. When I click on my archive links I get a "file not found" error. Must have something not configured correctly for static generation. Will poke around the support forums, but for now going back to static generation. It doesn't take that long and I don't rebuild that much so it isn't a big benefit to go to dynamic (although it was ALOT of code).

Posted by Martin at 1:47 PM | Comments (0) | TrackBack

September 8, 2004

MT comment problems...

So I am still trying to work out why comments don't work. It apparently is an IE problem with handling cookies from a domain different from where MT is installed. Or is it a MT bug? It works fine if you are using Firefox or Netscape (I tried). I hope MT comes out with a fix and doesn't just sluff it off as an IE problem. That is 90% of the browsers out there guys...

Here is the mail that MT support sent me:

Hi, I'm able to log in for commenting with Firefox - are
you using Internet Explorer? I see you have MT installed on
a different domain name than your blog's domain name, and
IE has a problem with the cookies when the domain names are
different, so it may not be recognizing the sign-in because
of this.

Regards,

Shelley

Posted by Martin at 12:49 PM | Comments (1) | TrackBack

September 2, 2004

trying to fix comments

Ok, thanks to readers who wrote saying my new comment feature isn't working! The message people are getting is: The site you're trying to comment on has not signed up for this feature. Please inform the site owner. I am informed and working with TypeKey to see what the problem is. Off for the weekend so it probably won't get fixed till Monday. Sorry...

Posted by Martin at 9:26 AM | Comments (1) | TrackBack

September 1, 2004

Well Mt 3.0 didn't solve my comment SPAM problem

Supposedly TypeKey registration was supposed to solve the comment spam problem. But I have as much or more of it than before! They just get approved accounts I guess. I am tightening up on my allowed comments. I had "enable comments from unregistered users" but that only required a name and a confirmed e-mail address. Easy to fake. I set it that way because I didn't want to have to require people to have a TypeKey account. To have a TypeKey account, you need a blog. The skeptic could view this as an attempted lock-in from Movable Type and a lead generation vehicle. Well, I am changing settings to require registration now and we will see what that does to comments.

Sorry to my regular readers for the hassle.

Posted by Martin at 12:52 PM | Comments (0) | TrackBack

August 25, 2004

Blogger integrates AdSense

Blogger has integrated AdSense. Duh. Both owned by Google. And Typepad will probably do the same thing. Keep your eye on NWVENTUREVOICE.com for a longer post about my thoughts on e-commercizing blogs.

Posted by Martin at 10:40 AM | Comments (0) | TrackBack

August 23, 2004

Tossing Ecto

I installed Ecto as a blog client but have been unimpressed so I am deleting it from my system. Going back to w.Bloggar which is solid and has an upgrade coming. Major issues:
1. confusing interface.
2. not intuitive how to post. No dafault button, you had to use a Ctrl key sequence.
3. Did not read formatting defaults from MT configuration correctly and I had to keep resetting formatting to convert line breaks.
4. No multiple file upload for images. Neither does W.Bloggar, but that is a feature I would have switched for.
5. Preview not as good as W.Bloggar.
6. Slower than piss. Slowed my whole system down. You could watch metal rust in the time it took to post.
7. No compelling new features.


ouch.

Posted by Martin at 9:43 PM | Comments (0) | TrackBack

I scooped Wired

Got my September Wired today. On the last page they have a photo of a lost file sign posted on a tree in South Park by their offices. I picked one of those things back on Aug 1 and posted the picture Aug. 18. MGTMoblog: Lost thesis. Lost in the air from a WiFi connection to the FTP server.

Interesting deal. Next time I will post sooner.

Posted by Martin at 1:08 PM | Comments (0) | TrackBack

August 16, 2004

A web publisher finds a profitable model

Wired News: Net Publishing Made Profitable shares Adam Engst's adventure in micropublishing PDF books. Buried in the article is a tidbit which is an "Ah Ha" moment that many may miss. For the price of his PDF books ($5-$10), you get free updates. These are mainly MAC technical books. The content changes as software gets updated and patched over time. Woa, so all of a sudden you get subscription like access to current technical content for one up-front cost? Physical books can't provide that. Higher priced web content portals can't match the economics. wow, that just might work.

Posted by Martin at 1:48 PM | Comments (1) | TrackBack

August 12, 2004

How NOT to upgrade from MT 2.6 to 3.0

OK, I am over it. Here is what NOT to do:

1. Do NOT create a new database
2. Do NOT run two parallel installs
3. Do NOT export from 2.6 and import to 3.0.

This causes all sorts of unnatural acts. Since I don't have a back-up of my old 2.6 database, I can't recover from this mess, so I am just going to take my lumps and move on. I will leave the 2.6 individual archive files out there to keep google page rank, but they won't work well. Hopefully google will find the same entries with the new archive file names and start linking to those.

The RIGHT way to upgrade:

1. Upload 3.0, overwrite the 2.6 files.
2. Run upgrade script.
3. Make CSS and Index template changes that you want to keep.

done.

Learn from my mistakes...

Posted by Martin at 6:20 PM | Comments (0) | TrackBack

Well I agree with Jason on one thing

Vibrant Media, for desperate and stupid publishers. - The Jason Calacanis Weblog - calacanis.weblogsinc.com

Posted by Martin at 2:24 PM | Comments (0) | TrackBack

August 10, 2004

Trying new Blog authoring client

Zempt didn't do it for me. Don't feel like reconfiguring Bloggar for my new paths for mt 3.0. I am going to try ecto blog which is recommended by Rafat Ali, the man, the myth behind Paid Content.

Posted by Martin at 10:40 PM | Comments (1) | TrackBack

Oh, and Trackback is broken too

When you delete a weblog from an old installation of MT, you remove all the entry numbers from the database so the trackback script pukes. Check this out: Deep Green Crystals: this is a cool site

Lesson: You can upgrade from MT 2.64 but you can NEVER get rid of it! That doesn't sound like much of an upgrade to me...

Posted by Martin at 6:20 PM | Comments (0) | TrackBack

More MT archive upgrade problems

Thanks John for talking about my MT archive comment problem today. The suggestion to just turn off comments on those old archive pages would be grand if I hadn't already deleted the old weblog in MT 2.64 (stupid!). Don't do this! Now I can't change the configuration of those individual archive files without running some massive script over them to remove the comment code. I did just that on one entry here: Deep Green Crystals: this is a cool site and it worked fine, but is not scalable to 732 entries. I thought about importing the entries again into MT 2.64, but the app would give them all new entry numbers and generate different file names anyway which would break the Google links also.

There is a bit in MT 3.01 which says "keep backward compatible file names" which sounds promissing, but does not deliver. They simply create 001234.html file names instead of the new "friendly" file names of Cool_post_here.html. But when you import to 3.01 you get all new entry numbers as well so the files are not name compatible. What would solve all this would be an option in MT 3.0 import which says "preserve entry numbers". Then warn you when you are trying to overwrite and offer you suggestions for magaging conflicts. Then you could have mt 3.0 regenerate the same file names that google has but with the new formats.

Another annoyance is that the 2.64 and 3.0 style-sheet.css files are different with different names for tags. This means that the 2.64 archive files will render in the 3.0 index template, but when they use discontinued css tags, those show up in the post as random codes. Very distracting.

Basically, MT fix the upgrade process with archives !

Posted by Martin at 6:14 PM | Comments (0) | TrackBack

August 9, 2004

Upgrade to MT 3.01 has NOT fixed my comment spam problem

Since MT changed archive file names in 3.0, I had to leave the old ones there. And the comment spam bots hapilly find the files and use the comment script to make comments. Unless I blow those files away (and my Google pagerank with it), I can't get rid of spam. I have a couple of ideas I may try. Maybe using the Mt3.0 comment script on the MT 2.64 archive files? Will letcha know.

Posted by Martin at 10:43 PM | Comments (0) | TrackBack

Fixed CSS problems with new three column layout

Thanks to Steve for pointing out formatting problems with my new CSS for DGC under MT 3.01. God how I wish MT would fix these problems FOR me! Well, after two hours of futzing around with CSS, here is the solution. Basically with the Absolute tag in CSS, it will position your content (in this case the right and left link boxes) relative to the browser window in absolute terms from the upper left corner being 0,0. That means that the absolute is in relation to the browser window itself. So with my prior settings, when the browser was full size, it worked fine. When you resized, it made the link bars overwrite the rest of the content.

The fix is to get rid of the left and right absolute tags and use a big left margin to move your right box over. Leave the Top tags to get the boxes to appear at the top of your page. If you take that out, the linkboxes go to the bottom of your page.
Whew...

#leftlinks {
position: absolute;

top: 85px;
border:1px solid #FFF;
width:200px;
background-color: #FFFFFF;
overflow: hidden;

}

#rightlinks {
margin-left:729px;
position: absolute;

top: 85px;
border:1px solid #FFF;
width:170px;
background-color: #FFFFFF;
overflow: hidden;
}

Posted by Martin at 10:35 PM | Comments (3) | TrackBack

August 5, 2004

Moved to DGC to MT 3.01

So with the family out of town, I finally get a couple hours for delayed SysAdmin stuff. Two hours and fifteen minutes later and I have my Deep Green Crystals and all 624 entries and 204 comments ported over to MT 3.01. I hope this addresses my comment spam issues. I waited for 3.01 because the import function of 3.0D did not import entries from 2.64 correctly (lost comments and other things). The process was mostly copy and paste of code in templates then an export/import, but there were a few gotchas to watch out for...

- Had to use Netscape to export entries. IE didn't work even with the tips suggested. And make sure you use Save As a .txt file. I tried to copy and paste the entries into WordPad and created a .RTF file. MT 3.01 didn't understand it (MS probably put some random bits in it).
- The new Style Sheet CSS in Mt 3.0 changed some of the formatting codes and names of sections (for the better). It requied a little hunting and poking to get the right tags around my new text. Would have been nice if the divisions were named the same though.
- Same problem adding three column format as I had with 2.64. Come on guys, provide a standard 3 column layout! I still have some resizing issues because I am using Absolute positioning on my left and right columns, but I will fix that tomorrow. Here are the three key bits of code to change/add to your .CSS file if you want a three column look:
1. Make the container larger:
#container {
line-height: 140%;
margin-right: auto;
margin-left: auto;
text-align: left;
padding: 0px;
width: 900px;

background-color: #FFFFFF;
border: 1px solid #FFFFFF;
}
2. Give the Center div a left and right margin
#center {
margin-right:170px;
margin-left:205px;
margin-bottom:20px;
border:1px solid #FFF;

overflow: hidden;
}

3. Create a left links division
#leftlinks {
position: absolute;
left: 62px;
top: 85px;
border:1px solid #FFF;
width:200px;
background-color: #FFFFFF;
overflow: hidden;
}
4. Create a right links division.
#rightlinks {
position: absolute;
right: 62px;
top: 85px;
border:1px solid #FFF;
width:170px;
background-color: #FFFFFF;
overflow: hidden;
}
Then put what you want in your Main Index template for the leftlinks and rightlinks section.

One trick: Leave your 2.64 files on the server, especially the archive files. They will render with the 3.0 templates. Won't look great, but you won't break so many links in Google. With the new file names and formatting I am waiting to find the broken links, but I trolled Google a little and most are working because they go to the old Archive files that are still there.

Off to bed!

Posted by Martin at 12:16 AM | Comments (2) | TrackBack

August 3, 2004

Joining the MoBlog craze

Well my Nokia 6230 made me do it. Don't know exactly what I am going to post, but I found one very scary sign at Kuleto's in San Francisco. Check it out.

Posted by Martin at 3:10 PM | Comments (0) | TrackBack

July 28, 2004

See me on a BlogOn panel

on Friday I was on a panel about investing in social media at BlogOn conference. Here is the video.

Posted by Martin at 10:30 AM | Comments (1) | TrackBack

July 26, 2004

BlogOn comments

I was thinking of posting my own impressions of the BlogOn conference I was at last week, but Rafat Ali did such a good job in PaidContent.org: July 25, 2004 Archives
that I won't opine much more. Basically good networking, poorly organized conference. My panel was at 5:00 pm on a Friday. There were supposed to be 300 people registered, but I counted less than 100 in the room. We had fun as a panel, but don't know how much it really helpled.

Posted by Martin at 4:00 PM | Comments (0) | TrackBack

July 23, 2004

Movable Type announces 3.1

Watching Ben and Meena announce Movabletype 3.1 right now at BlogOn. They are having a problem with the server speed, so the major features are:
1. movabletype Pro features like; Post scheduling feature. Auto posting when you are gone. Wow, will it write it for me too?
2. Subcategories are implemented. Very cool. You can move them around to different major categoris. Very cool UI and flexible.
3. Dynamic pages. Original design was to create static pages due to their better performance. Now they will have dynamic pages per template. You can pick static or dynamic per template. You can integrate PHP directly into the templates. All template tags re-written in PHP. This is a check box feature add/subtract. This is very cool for low traffic sites.

This release is out in 3 months instead of 1 year for the last one. This is enabled because they are charging for the product now and they can hire developers. Meena did a good job of showing how you get more features sooner when the thing is paid for. They are on the right track.

Posted by Martin at 3:22 PM | Comments (1) | TrackBack

Video on your blog

First Stream Business Solutions announced today at BlogOn that they have a blog client to add video to your blog easilly. This reminds me of the early days of Loudeye Technolgies. This consumer video idea is not very profitable. It is certain the FirstStream is failing at their business video solutions and see all the people on blogs and they think there is growth there. Can't see paying ANOTHER $10 per month just for video.

Posted by Martin at 3:10 PM | Comments (0) | TrackBack

July 22, 2004

Newspapers enter bloging and sell ads

From the old media jumping on a trend file comes this: Poynter Online - E-Media Tidbits. A Spokane Wash newspaper started publishing blogs (lost of reporters sitting around) and is now selling ads on them. Apparently profitably. Now I wonder if this is a testament to the ad sales force that the newspaper already has, or a testament to the relevance of those blogs. I actually don't read the local paper. But they sure sell alot of ads! Do I really want to go read local newspaper written blogs with car ads in them?

Posted by Martin at 7:20 PM | Comments (0) | TrackBack

June 7, 2004

RSS linked to radio...

The Doc Searls Weblog : Friday, June 4, 2004 points out a Nokia/HP program to syncronize RSS feeds with radio broadcasts. There have been a number of these attempts over time to get radio listeners more information in some sort of a datacast. Sony sells a little dongle thing that you can click when listening to the radio and when you are back at your PC, plug it into the USB port and get info on the song, buy it, etc. Apparently what Nokia is doing is trying to get radio stations to program RSS feeds syncronized with radio broadcasts. These would be received in mobile phones as either WAP or better, SMS. The user could answer questions, get more information, interact in a number of ways. This one actually has promise. I like the idea of driving to work listening to NPR. They have a story about something and ask for an audience poll. My phone vibrates and I vote SMS. Then hear the results. Pretty cool.

RSS is not just for weblogs anymore.

Posted by Martin at 8:08 AM | Comments (0) | TrackBack

June 1, 2004

Interesting paper on collagorative weblogs

This recent paper talks about four top sites that are big collaborative Weblogs. The challenges of the model and what works/doesn't work. Good micropublishing insight. Kinda academic though.

Posted by Martin at 12:56 PM | Comments (0) | TrackBack

New poll service a bust...

A helpful reader suggested that I try PollMonkey. So of course I did and compared it to Blogpolling which I am now using. The Poll Monkey guys obviously spent alot more money on design of their site and it is very pretty. The reason becomes immediately obvious when you keep getting pop-ups that try to upsell you t a $19.95 per month "pro" version. The basic version comes with 100 responses per month. I don't know if there is a limit to the number of polls you can have as I only set up one. Also some features are only available to pro users like security, HTML results, and e-mail results. Funny, when did security become a "premium" feature? Well you have to draw your feature line somewhere. So I set up a poll and copied the javascript into my blog template and hit rebuild. Nothing showed up. Big white space. I clicked on a couple other links, then came back and the pool was there. So there is some problem with their server getting the picture there quick enough, or being in sync with my browser reload or somthing. So I make my selection and hit "vote". The poll goes away and the white space returns. No results. After a couple other links and multiple re-loads the results miraculously appear. But then they won't go away again.

BlogPolling gave me none of these problems. Sometimes simplicity is the best answer. I am staying with Blogpolling for now.

Posted by Martin at 10:43 AM | Comments (0) | TrackBack

May 25, 2004

Bloglet is cool

Bloglet is a service that will put an e-mail box on your blog and automatically send subscribers your new posts each day. More blog specific than notifiylist.com. I may try it out.

Posted by Martin at 9:10 AM | Comments (4) | TrackBack

May 21, 2004

Here is a nifty web service

free mailing lists from Notify List . com. Just sign up on their site for an account and copy some javascript into your site. When people put their e-mail into the box, it goes onto your list and you can mail away through their browser. Pretty nifty and simple, database-less, programming-less way to implement one-way e-mail notify lists. Have NO idea what the business model is, but I will us it!

Posted by Martin at 2:00 PM | Comments (0) | TrackBack

May 20, 2004

MT 3.0 first impressions

Installed MT 3.0 today. I installed the free version since it supports the number of blogs and authors I need so I won't be jumping in on the licensing comments. Some first impressions:

I installed the system in a separate directory with a separate URL so I could keep running 2.664. This reqired a couple of things that were not clear in the installation scripts. To make my updates to mt.cfg, I just opened two WordPad windows and scrolled thorugh my 2.664 file, copying the changes to my new 3.0 file. This worked fine, except I was hoping to reuse the same MySql database. No such luck. The MT.CGI wouldn't let me sign-on and the mt-load.cgi barfed saying "MT-Author" table already existed. So I created a new database, changed the mt.cfg and ran MT-load.cgi. Bob's your uncle, mt.cgi log-on worked fine.

There is a little inconsistency in the installation script for the full 3.0 version and the upgrade. In the upgrade part they talk about moving mt.js to the mt-static directory, but don't mention this in the fresh install. Being safe, I moved it anyway.

The next glitch was the import of entries. I had of course backed up all entries before I started any of this just in case something walked over the prior install. The import function does not correctly interpret HTML. So all the links come through as text instead of hyperlinks. Ouch! I wonder if I twiddle the configuration around auto interpretition of HTML if that would fix it. Need to work on that.

The next set of weirdness is the new features. I don't see any meaningful ones. There is alot being said about the comment upgrades, but it is just adding the ability to accept TypeKey authors and to moderate comments. TypeKey is something that can be spoofed as the barrier to an account is very low. And it is only designed for other blog authors. What about people commenting who are just readers? Anyway, you still need to run MT-Blacklist and to rename your comment script to be safe. Which I did. We will see how it goes when I go live with the convert. In the last week, even with MT-Blacklist and renamed comment.cgi I have been hit with over 100 spam comments.

The much vaulted new plug-in architecture is hard to figure out. I am working with a programmer to write a plug-in for their competition, but his initial review of the code is that they didn't go far enough to open up the UI. Wait for a full review of those functions.

Posted by Martin at 4:10 PM | Comments (0) | TrackBack

Blog platform review

Thanks John, this is just what the doctor ordered.
a little ludwig goes a long way: CMS comparison chart

Posted by Martin at 3:48 PM | Comments (0) | TrackBack

Write your first MT plug-in

this article: O'Reilly Network: Developing Movable Type Plug-ins [Mar. 19, 2003]
Gives a very good step by step process through writing your first MT plug-in. Since I am doing that, it was a great intro. It was written for 2.6, so the new 3.0 stuff is not in there, but you can find that in the MT documentation.

Posted by Martin at 1:53 PM | Comments (0) | TrackBack

May 19, 2004

MT licensing uproar

this thread Movable Type RIP | Metafilter and others point to a rift in the base of MT early adopters. Now I am not of the "I only use free software" camp. Like John, I don't mind paying for professional services. One thing I do mind is that initially there was no upgrade pricing. I made a $40 donation to Six Apart. And I paid for an installation (of an early version) which I believe Mena did herself. Getting SOMETHING for that would be great. I hear that you can get credit for your previous donations by sending in your Paypal transaction ID. I am testing 3.0 now and will probably stick with the free version for now as I don't host that many blogs and only have one author. If you are hosting blogs for other people and have lots of authors, you should probably pay.

Posted by Martin at 12:56 PM | Comments (0) | TrackBack

Micropublishing gets money

The New York Post Online Edition: business is confirming that Mark Cuban invested in Weblogs Inc. And that Gawker Media, their competitor, is in the black with their two publications. Micropublishing is moving on...

Posted by Martin at 10:21 AM | Comments (0) | TrackBack

May 6, 2004

Google Adsense in Archive templates

Thanks to E. Timothy Uy for the code to add Google Adsense words to my Archive pages. It wass much easier than adding to the main index page. Since the archives all have a standard header basically over on the left, if you just float a couple ads on the right, you don't have to mess with the body of the template. So this code puts two side by side google ads to the right of your header.

Go into the Category, Date and Month archive templates. Scroll down to the body. Just under the "Banner" div, add the float right style div with your google script pasted in. Sample section below. Replace the script code with what you copy from google. I put some spaces in the code just after the "<" to make MT interpret it as text, so you have to take those out if you want the code to work.


< div id="banner">
< div style="float:right;">
< script type=" text/javascript">
< !-- google_ad_client = "insert your client id here"; google_ad_width = 468; google_ad_height = 60; google_ad_format = "468x60_as"; google_ad_channel ="insert your channel id here"; google_color_border = "CCCCCC"; google_color_bg = "FFFFFF"; google_color_link = "4A7184"; google_color_url = "4A7184"; google_color_text = "444444"; //-->< /script> < script type="text/javascript"
src="http://pagead2.googlesyndication.com/pagead/show_ads.js">
< /script>


< h1>< a href="<$MTBlogURL$>" accesskey="1">< $MTBlogName$>
< span class="description">< $MTBlogDescription$>
< /div>

Posted by Martin at 1:35 PM | Comments (2) | TrackBack

May 4, 2004

Adding Google Adsense to my blog

So after seeing my server logs and realizing that people actually read my blog in some significant quantities, I decided to see if the adding Google Adsense ads to the site could generate enough to pay for my daily latte habit. Here is the blow by blow implementation....

Step 1: Get a Google AdSense account. You would think that this is trivial. It is not. I first applied about six months ago when I had very little traffic and infrequent posts. I was denied. They sent a nice mail saying "your site doesn't meet the criteria". Maybe one wrinkle is that my blog is at a subdomain off my main domain. The blog is at www.martinandalex.com/blog and Google won't let you enter that subdomain, only the top level which is www.martinandalex.com. When their bots went there and found a personal web page, instead of the insightful, engaging blog content, they probably ran for cover. When you sign up for an AdSense account, you use your e-mail address to log-on. So when I went back this time and tried to apply for an account they said I (my e-mail address) was not elligible (based on six months ago scan). Being wiley and persistent, I signed on with a different e-mail address and gave the domain again. This time their spiders came out and found the blog, all the insightful postings, and approved my application. Now I don't know if the AdSense criteria have changed in six months, or if simply over time and more traffic, my site now qualifies. It is a bit of a design error with Google that they blackball your e-mail address and only allow one domain per e-mail address. Good thing e-mail addresses are easy to come by. So I got the account, figured out what ad format I wanted for the site and copied the javascript code. Ready to paste into my Movable Type index template.

But not so fast....

Step 2. Add the AdSense code to your Blog template. Nothing could be as easy as it sounds. My Movable Type Index template was two columns with a small left sidebar and a variable content main body. I wanted the ads down the right side of the blog. This means a three column blog. So I had to break it down into substeps.

Step 2a. Make the blog a three column format. Now not being a programmer (anymore), I asked my pal John who is. He didn't know exactly how to add a right column, but pointed out the Cascading Style Sheets part of MT configuration. So I had to become a CSS programmer. MovableType does not provide any three column templates standard (bummer). Luckilly Google provided the answer at this very helpful site. There was the three column CSS code. And different variations. I chose one with a fixed left and fixed right margin with floating text in the middle. That means that as you resize the window, the text in the middle will resize and wrap as the left/right stay there. The main work was creating a "Rightlinks" and "Leftlinks" division in the style sheet. And modifying slightly the #Content (center) portion code. Here is the relevant new code which I added to styles-site.css in MT.

                  #content {
background:#FFF;
margin-right:163px;
margin-left:225px;
margin-bottom:20px;
border:1px solid #FFF;
voice-family: "\"}\"";
voice-family: inherit;
margin-left: 225px;
margin-right:163px;
}
html>body #centercontent {
margin-left: 225px;
margin-right:163px;
};
}

#leftlinks {
position: absolute;
left: 10px;
top: 85px;
border:1px solid #FFF;
width:200px;
}

#rightlinks {
position: absolute;
right: 10px;
top: 80px;
border:1px solid #FFF;
width:164px;
}

Step 2b. Change the index template to pick up the new style sheets. Basically, you have to rename your "links" division to "leftlinks" and then add a "rightlinks" division. Sounds simple, but after a couple of placements, I kept getting the adsense stuff either appended after the bottom of the "powered by" part of the left links or written directly over top of the calendar part. There was a very nice white space over on the right that was totally unoccupied. After a couple hours of futzing around, I decided to go back to an old programmer trick of counting begins and ends. It looked to me like the "rightlink" division was being treated as part of the "leftlinks" division. If that was so, there was probably a missing "< / div >" somewhere. That was in fact the case. When I fixed that, the following code slapped the google Adsense ads right over on the right like it is supposed to. (had to insert some spaces to get MT to see this as text and not try to run the code).

< div id="rightlinks">


< div class="side">
< script type="text/javascript">< /script>
< script type="text/javascript"
src="http://pagead2.googlesyndication.com/pagead/show_ads.js">
< /script>
< /div>

< /div>

Step 3. Rebuild the site and start cashing the checks. Well, no checks yet, but I will keep you updated.

In all it took about five yours to implement. Too long for the average guy. And I had to become a CSS and HTML programmer. This has to be an order of magnitude easier to do.

Posted by Martin at 1:31 PM | Comments (7) | TrackBack

April 19, 2004

Google Adsense hack for Blosxom

Mehack
has posted an interface for Blosxom to google Adsense. The monitization of blogs has begun in earnest!

Posted by Martin at 9:39 PM | Comments (0) | TrackBack

Ads everywhere

Lots of media recently about what is next in search. The real question is what is next in terms of monitizing context. And what are the new contexts? e-mail, RSS, mobile are the new frontiers. A quick round-up of efforts here: The Next Context. The conclusion, no-one has a solution for RSS contextual ads yet.

come on people, send me a plan to fund!

Posted by Martin at 8:40 PM | Comments (0) | TrackBack

April 14, 2004

What is the blog business model?

BuzzMachine... by Jeff Jarvis is leading a panel this weekend in Boston at BloggerCon.
He has solicited ideas on how to monitize blogs and collected the initial ideas in this post. A very thorough list. I don't believe most of them. In the end with every new data type, the guy who does search makes the most money. The publishing tools get driven to zero. Hosting gets commoditized. It costs too much to build a brand or to deliver premium content. Consumers want free and easy access.

Posted by Martin at 8:35 AM | Comments (1) | TrackBack

April 13, 2004

Newsgator launches multi-media edition

Well it had to happen.
Product extension. Basically it looks like an interface to some of the p2p distribution networks. You need NewsGator online services to get multimedia content. They aggregate it at their server and you pull it down when you need it according to your subscriptiions. Basically NewsGator is a subscription manager for multiple datatypes for you. Don't know if I want them to be that. Happy now with them as a reader though...



"Living Room" Interface Allows Users to Read Selected Content, or Watch On-Demand Video Content

DENVER, CO -- April 13, 2004 -- NewsGator Technologies launched NewsGator Media Center Edition today, which allows users to read syndicated content feeds on their TV with Windows XP Media Center Edition. Both text and multimedia content is supported, with an interface designed to be used with a remote control from across the room. NewsGator Media Center Edition shows information that has not already been viewed on another device by synchronizing user subscriptions with NewsGator Online Services.

Audio and video content is only one click away. Users see visual cues when a feed contains multimedia content; they can then instantly view this content using the remote control. There are a number of featured feeds for the launch, including a video feed from Microsoft.

This is the latest step in the "any time, any place, any device" strategy that distinguishes the NewsGator product line. Users typically use more than one device throughout the day, such as Microsoft Outlook at work, a mobile phone on the road, a web browser in their home office, and a TV in their living room. NewsGator Online Services allows users to read one set of content from any of these devices, without any duplication. The combined power of NewsGator Media Center Edition and Online Services offers customers a productive and fun way to access their personalized subscriptions and information from any device, whenever they need it.

"NewsGator Media Center Edition is a great example of cutting edge ISV innovation on the Microsoft platform," said Sanjay Parthasarathy, corporate vice president of Developer and Platform Evangelism at Microsoft Corporation. "The deep integration of technologies like XML and RSS, combined with rich multimedia, and presented on a device as pervasive as the television, gives consumers access to the information they care about, where and when they want to view it, truly delivering on the vision of a connected home. We're proud to work with NewsGator to help bring this innovative application to market for our mutual customers."

NewsGator Online Services also includes other three content reader editions, which allow users to read content they have subscribed to from any web browser (with Web Edition), any email client (with POP Edition), and from a mobile wireless device (with Mobile Edition), which is a powerful feature for road warriors who use mobile devices to access information while on the road.

The service also provides exclusive, subscriber-only content to its subscribers, as well as the ability to search for content that matches a specific keyword or URL, and return that content in a feed.

NewsGator Media Center Edition is included with the NewsGator Online Services subscription service, and is available immediately at http://services.newsgator.com. Pricing starts at $5.95/month per user.


[NewsGator News and Updates] [channelTitle]

Posted by Martin at 9:09 PM | Comments (2) | TrackBack

What good is the Technorati API?

Been thinking about Technorati's business lately. And FeedBurner's. And the whole idea of value added services around Blogs. What is the value of collecting lots of information about blog posts from different platforms in a central database? What value can be added?

Boing-Boing recently replaced comments and track-back with a new feature integrating the Technorati API. There are others as well:
chaotic intransient prose bursts: An evolved Technorati plugin. What do you get by using a Technorati service instead of something like TrackBack? Well Technorati indexes non MT blogs for one. Now there are other blog platforms that are supporting Trackback, but not as many as ping Technorati. You will get more results with the Technorati API. In addition to the links themselves, you can also pull other useful structured data about that link that you can use in your publishing system like the number of posts referencing the link, the time of the latest link, minutes since last link, summary text around the link, author of the link master weblog, and more information to come. I believe you can get more information from the Technorati server than from the Trackback server. But this will probably be an arms race. It is interesting to note that Technorati is starting to attract developers to their platform in much the same maner as Movable Type.

Posted by Martin at 8:15 PM | Comments (0) | TrackBack

April 10, 2004

Screenshots from BlogBot

Kudos to Joshp who in 15MB of fame: Blogbot? managed to turn up screen shots of Microsoft's upcoming BlogBot. The troubling thing about BlogBot is that it will only search "relevant" blogs. Relevancy apparently determined by Microsoft. That is not search in my book. That is editorializing. That is news aggregation. That is convience. That is not search.

Posted by Martin at 8:42 AM | Comments (1) | TrackBack

April 5, 2004

Well I bought an RSS reader and already regret it

So I had winnowed the RSS reader search down to FeedDemon and NewsGator. I have been traveling for two weeks, so I haven't really had time to do a head to head comparison, and had only played with each for about two hours. My trial on both ran out, so I had to bite the bullet. My decision criteria really wasn't that complicated. I didn't want another icon on my desktop, so I went with the Outlook plugin. Plunked down $29 for NewsGator. They should change their name. Gator is a bad word.

The first sign of trouble was in configuring postings out of the thing. With FeedDemon, they had a handy feature where they just auto discovered your posting app (I have been using w.Bloggar) and they automatically set it up. Well NewsGator doesn't work with w.Bloggar and you have to install another app from another guy: public MattBerther : ISerializable: MovablePoster. This plug-in is free now, but no gaurantee of how long. And it is definitely beta stuff. It took three trips to the support boards to get it working and configured correctly. And when you do post, NewsGator conveniently slaps it's "related" link on my posts which encourage people on my blog to click through to NewsGator's on-line services. Now I paid for my software and I don't want to be a shill for their viral marketing please. I haven't found where to turn this off yet. Also, this plug-in formats posts VERY differently than Bloggar and the difference is obvious. Now I am sure that with some programming in the template I could fix that, but I am a busy guy. This is alot of braindamage for a simple thing. Oh, and the "post to my blog" tab is well hidden in a couple of NewsGator menus. It is not a right-click option or an easy to find button on a menu bar. Why not?

I also have noticed that my outlook has slowed down considerably. Right now I am working over a VPN from home to the office server. I bet NewsGator is pulling everything from the net and putting it on my server then sending it along to my client. In that respect, a separate application would be better because they would just pull directly to the client and bypass the VPN/server. When I was running it, FeedDemon was very fast. Then there is the problem of new postings to my RSS feeds. With FeedDemon I got a cute pop-up window from the taskbar for new posts like I do when I get new e-mail. Of course this can be turned off. In News Gator it can't be turned on or off, it is not available. Another feature I kinda liked in FeedDemon was the ability to define a category of feeds and have the application assemble a "newspaper" for you with all the recent posts. Very cool.

FeedDemon also did a very good job of auto discovering the different RSS and Atom formats and finding the right place for a feed. NewsGator is very touchy that you get the stuff exactly right.

And then there is the problem of business model. With NewsGator you have to pay extra for things like reading your feeds in a browser, and custom searches. Many other readers offer that free. I hope some of this gets included, there will definitely be price pressure to do so.
Wow, sounds like I should have bought FeedDemon eh? Well I am an American at heart and am easilly swayed by convenience. Lured in with a promise of easy integration. I will stick with News Gator for awile, but may end up changing. Not sold yet. But they do have my money...

Posted by Martin at 7:55 PM | Comments (2) | TrackBack

April 4, 2004

Some guys at HP muse about information architecture in blogs

A couple researchers at HP have published a musing about information architecture in the blogspace. They are proposing a new way to rank postings in the space called iRank (similar to Google PageRank). I don't know if they have nailed it, but the idea is correct. The interconnections are different in blogs than raw HTML. There are time stamps, authors, FOAF files, personal relationships, etc. Since I am an investor, I hope Technorati solves the problem and defines the access paradygm.

Posted by Martin at 8:17 PM | Comments (0) | TrackBack

March 22, 2004

The central MT anti-spam tactics page

I knew with a little searching I would find such a place. .:: SPAM FIGHTING TECHNIQUES TO PROTECT YOUR BLOG ::.. Here are 10 or so tactics to reduce spam on your MT blog. All different approaches. I just implemented the rename your comment cgi script tactic. haven't gotten any more spam (in last 4 hours). If I start to get it again, I will probably try some of the others. I don't like the blacklist thing (it has already gotten corrupted once) so I will probably un-install that and try something new if they get through.

Posted by Martin at 2:26 PM | Comments (0) | TrackBack

New Technorati site!

Technorati: Searching the World Live Web Want to know what is going on in the blogsphere? Want to know your place there? Want to track keywords? All can be found here....

Posted by Martin at 10:04 AM | Comments (0) | TrackBack

March 18, 2004

Another survey tool

Thanks to a Kevin, a faithful reader for a pointer to: SurveyMonkey.com - Powerful tool for creating web surveys. Online survey software made easy! I am currently trying BlogPolling. I haven't had much time to play with it, but 8 people have taken the poll in the first week, so somebody is reading! Next poll I will try with Survey Monkey...

Posted by Martin at 10:49 PM | Comments (1) | TrackBack

March 15, 2004

Supersize your RSS

The Blogsphere is a wonderful place. These guys: FeedBurner - The spark for syndication success will add all sorts of stuff to your RSS feed to improve them. They basically route your rss from your blog server through their servers and add stuff. Like converting the different RSS formats and Atom. And inserting your AMazon associates ID to Amazon links. And tracking lots of stats. I think there are 100 ways to add value to RSS by pulling it through your servers. Very interesting concept.

Posted by Martin at 10:21 PM | Comments (0) | TrackBack

Another RSS reader off the list

Just un-installed SharpReader RSS Aggregator as well. I read a number of reviews that picked it as a simple, sturdy reader. It is that, but that is not what I need. Does a couple of things funny.
1. The pop-ups for new posts that come out of the task bar are too big. They include a summary of the message. I just want the headline.
2. I actually liked the auto-discover features, they auto discover most RSS feads without alot of screens (good thing).
3. auto lookup of linkcosmos in Technorati was great!
4. No integration with publishing. This was the killer for me. If I am reading a post that I like and want to link to it on my blog, I want easy integration.
5. My Firewall (Zone Alarm) also complained every time I started SharpReader asking if I really wanted to set up as a server. I thought this was client software. I don't know why they are setting up a server, but I don't want that.

Posted by Martin at 3:30 PM | Comments (1) | TrackBack

Two RSS readers off the list

I have been testing different RSS readers lately. Two have fallen off the list:

1. Pluck: RSS Reader & Power Search. At first I liked the idea of reading RSS in the browser, but after trying it I hated it. I want to read in a separate application, or maybe inside Outlook. Not IE. Also, after you install it, it is not abundantly clear how to access it. It does not always load in IE. You don't get a toolbar, you get (sometimes) a sidebar. Out of sight, out of mind. Nix.

2. RSS Reader.
This was an early favorite of mine because it was free all day all the time. But the features just aren't there. I guess that is one thing you get when you pay for software. Adding channels was hard, no autodiscover feature. Drag and drop of XML tags not available. Simply not rich enough yet.

What is left? FeedDemon and NewsGator.; both require payment (pain!), but are full featured enought that I may actually pay after the trial. I like the 30 day trial on FeedDemon, 15 days on Gator seems short sighted. NewsGator needs to change their name, my spyware software keeps trying to delete it as the "evil" gator. I am addicted to the auto generated newspaper feature within a channel group in FeedDemon. I also like the integration with publishing in w.Bloggar. They carry the URL there seamlessly. Very neat.

Posted by Martin at 3:23 PM | Comments (1) | TrackBack

Trying out Blogpolling

I have been looking for a quick and easy way to add polls to my blog and am gunna try this one: BlogPolling Voting System :: Free Polls for your Website or Weblog Go vote on the SPAM poll today!

Posted by Martin at 9:22 AM | Comments (1) | TrackBack

March 10, 2004

PicoStation

PicoStation Your Station For Wireless Mobile Blogging - Start Have a wireless phone with a camera? Download this little app and start a cell photoblog. I don't have one (use blackberry), but I may get one. The CEO started 4TH Pass and is a very creative entrepreneur.

Posted by Martin at 8:33 PM | Comments (0) | TrackBack

March 2, 2004

A micro portal

One of the themes we have been batting around at Ignition is the idea of micropublishing. Do Blogs allow you to publish authoratatively on multiple subjects for a fraction of the cost of New York Times? Yea, probably. Along this idea is single purpose sites that do one thing REALLY WELL. Think the old time wasters we love. Along this vein is a site that tells you what glue to use when you want to glue any one thing to another thing. This to That (Glue Advice) I can see this as a web service. I am going to use it tonight actually.

Posted by Martin at 2:29 PM | Comments (2) | TrackBack

February 19, 2004

My Rss Reader

Ok, so I hate to pay for an RSS Reader at this point in the technology. It is not clear to me what the value added over parsing a standard XML feed is. I gotta try a bunch to figure out if there is anything worth paying for. I did not install NewsGator because it was only a trial that required payment soon. I prefer a totally free thing now. So I moved on to RSSReader. Installed without a hitch. It was a bit tricky figuring out what URL to point it at. After a couple trips to the Movable Type support site, I figured out that the standard is point to your index.rdf file. It would be nice if the reader just figured that out for you. Tell it which blogging tool you are using and it can figure out where the standard stuff should be. I have been using it for a couple days and it seems fine. Things I would like:

- A unified view of all new posts. Not have to navigate by blog through posts.
- Easier hook-up to blogs
- A search box that searches the entier blogsphere.
- Read in OUtlook
- have on-line clientless access

I guess NewsGator does much of the above. I guess that is what you get when you pay for something. I will try that one next.

Posted by Martin at 2:52 PM | Comments (2) | TrackBack

February 18, 2004

Useful review of RSS readers.

RSS Readers (RSS Info)
I tried to install their #1 pick, SharpReader and got an error. I am moving on to another. I have used Radio UserLand before, but I will try NewsGator this time. Hope it doesn't come with too much spyware.

Posted by Martin at 10:10 AM | Comments (6) | TrackBack

How to implement RSS feeds on your site

I have been thinking about adding RSS feeds to my site and found this very useful! Learning Movable Type: Displaying an RSS Newsfeed on Your Site

Posted by Martin at 10:03 AM | Comments (2) | TrackBack

January 20, 2004

Technorati Beta site

Ever wanted to search blogs? Technorati: Web Services for bloggers. Or figure out who is linking to you? Or how "authoratitive" you are on a particular topic? Or how a blog you reads ranks in relation to others? Check out the world's largest Blog index and author tool.

Posted by Martin at 9:24 AM | Comments (0) | TrackBack

December 3, 2003

Some very big words about search

I invested in Technorati. They are at the VERY early stages of doing something to search weblogs and make sense of all the metatags in there. This guy has thought alot about XML search which is quite similar. ongoing On Search: XML.
I am hoping that searching blogs is different enough and hard enough that a separate company becomes the expert there.

Posted by Martin at 3:00 PM | Comments (0) | TrackBack

November 13, 2003

An Iraq blog without politics

blogs are everywhere now... Nabil's Blog

Posted by Martin at 3:32 PM | Comments (1) | TrackBack

October 28, 2003

here is the code to change you MT upload directory

So you can find it again. movabletype.org : Support Forum
Yes my blog is now my memory.

Posted by Martin at 9:43 PM | Comments (0) | TrackBack

How to configure thumbnail image upload in MT

I have been hacking away trying to figure out how to get images in my blog. As you can see below, they are now working thanks to: Sniptools | Tutorials | Thumbnailing with NetPBM and Movable Type. I found this VERY helpful tutorial by following a bunch of links. My ISP didn't have IMAGE:Magick, so I installed NetPBM and it was up in a jiffy.

Posted by Martin at 8:23 PM | Comments (0) | TrackBack

October 7, 2003

More fun blog indexing

Just registered at: Blizg - The Blog Resource. looks like they are trying to get more people to use metadata. That is a good thing. Not sure how useful it is, other than allowing computers to understand more about a page than simply scraping the text. Structured data actually. Interesting effort.

Posted by Martin at 11:33 PM | Comments (0)

October 3, 2003

A twist on the Blog...

One cool thing about a Blog is that you can use it sort of like a memory. I have replaced my IE Favorites with the blog. I just post links here that I want to share with other people or even just ones I want to remember. One interesting twist on this is <a title="RAM" href="http://www.randomaccessmemory.org/">RAM</a>. They are trying to get people to post memories by year. Then they aggregate them by year so you can see other people's memories next to yours. Not a very good search tool. This may be good for public memories like who won the super bowl and what the hit song was, but it doesn't scale down very well to groups or personal memories. What this site is doing could be done equally as well by just mining the existing BLOG databases like <a href="http://www.technorati.com/">Technorati</a>

Posted by Martin at 1:26 PM | Comments (0) | TrackBack