So I really wanted WLW to work. It's got a nice interface, some clever features (tables!), a plug-in community, etc. Just a few problems:

1. It regularly times out while posting, throwing an error and never showing a post as finished and posted (even if it is). It may not fully be WLW's fault -- Flickr does the same thing when posting to my blog -- but some applications don't, which means it takes some blame (and it's annoying).
2. The image uploader doesn't allow for any hand-crafted customization, only what MS has canned with it. In ecto, I can have the images wrapped with the drop shadow DIV stuff (and I can have drop shadows without actually modifying the image).
3. There's no native custom snippet/tag support. There are some plug-ins that will do custom snippets -- but not tags (i.e., select some text, click on the tag, have it wrapped with something in front and behind). That's a step backwards compared to what I have in ecto.
4. Category support seems wonky -- some things posted without (perhaps part of the time-out), others it wasn't clear how I could designate a primary category. (Yes, I know that "categories" are so 1990s, while tags -- esp. Technorati tags -- are what all the cool kids do. Bah.)

Windows Live Writer would be a very nice desktop client if I were just starting off with one. The complaints I have above are more a matter of advanced features I've come to count on, rather than something that everyone is likely to use. I'll be curious to see if MS does any significant improvements or push of WLW in the coming years.

Meantime, I'll be heading back to ecto, until I find something better.

Filed under :: Blogging - Technical
Link · Print · Edit · Comments (3) · Pings (0) · TR/G

Been poking around (as previously discussed) with desktop blog clients.  My old standby, SharpMT, has essentially shut down.  W.bloggar is more powerful than ever, but I've grown fond of WYSIWYG writing, and it only supports 12 snippets/tags/blocks.

So one tool I've often heard highly touted is Windows Live Writer.  Yes, working with the Evil Empire itself -- or with its tools.

I will say that:

a.  The interface is pretty.

b.  There is a healthy number of plugins to play with.

I'm not sold yet, but I like what I see so far.

More later.

Filed under :: Blogging - Technical
Link · Print · Edit · Comments (2) · Pings (0) · TR/G

*sigh* 

I've been using Ecto as a blogging tool for a couple of years now. It's a nice offline blogging client with some features I can't imagine living without. In particular it has some great formatting tools, a versatile picture uploader, and a template/tag/snippet tool that makes my life huuuuugely easier.

But ...

Ecto was originally created for the Mac, with a separate developer working on the Windows version. Upgrades of the Windows version have gotten slower of late, and now the two versions have reached a parting of the ways. While the developer says he'll continue to support the Windows version, the signs aren't promising:

1. The announcement was the end of March.
2. A new name, "by the end of next week," hasn't come up. Indeed, nothing has come up on the site about the product.
3. The support forums have been moved to a new site -- but empty. A promise to migrate the previous outstanding issues (including a few of mine) "some time next week" hasn't happened.
4. The product is no longer for sale (but still available for download). 
5. The future of the product, moving into v.3, "remains to be decided."

On the other hand, a new point version was posted at the beginning of April (first since January). And that fixed the one most annoying bug in the program. And I've always found the developer to be responsive until the last several months.

So I'm worried. And it's not about whether the product is good (it is), or the developer isn't well-meaning (he is), just whether he's going to have enough time (so far, not), and whether the product will actually grow in the future (prognosis cloudy).

*sigh* 

I'll stick with it for now, but I'll be keeping my eyes open for a replacement, as well as considering whether it would make any sense to shift over to using the MT4 native input screen.

Irksome.

Filed under :: Blogging - Technical
Link · Print · Edit · Comments (0) · Pings (2) · TR/G

Having recently re-added the books/movies I've seen recently (in the "***Recently" sidebar), I'm now also adding the most recent wines consumed via an RSS feed I get from CellarTracker and a clever little sidebar RSS widget from  SpringWidgets, by way of FeedBurner (though it doesn't require FeedBurner).

So now whenever I post a review of something I pull from my cellar, you can see it here, woo-woo. Click on the title (CellarTracker) to get summary reviews of all, or on any of the individual wines to see the details.

I'm not wild about the SpringWidgets Flash widget, but it looks a bit less unattractive than what Grazr offers, and even though I have the feed already showing up in the sidebar via Google Reader, I can't quite figure out how to get Google Reader to display it.

It's interesting -- I'm trying to turn this page more and more into sort of an aggregation portal for what I browse, read, see, and now drink, rather than relying on static links and blog posts.  I have specialized ElseBlog items, the Google Reader shared items (Unblogged Bits), my Recently "wishlist" from Amazon via Dealazon, and now this feed via SpringWidgets. All of which are meant to save me having to write added blog posts about stuff if I don't have the time or inclination -- the Lazy Man's way of blogging, I suppose.

I think SpringWidget is meant as either (a) an easy RSS feed aggregator for folks to post on their blog (or even their desktop), or (b) a cheap way to publicize one's site ("Hey, get the SpringWidget for my site!"). I think it makes a reasonable aggregator for my own personal feeds.

Though ... all this stuff has gotten a lot easier, but nobody has a complete model -- something that will take any source and spit it out in any other format without wrapping all sorts of crap around it (as even SpringWidget does). The things I can do with Google Reader come close -- but, as SpringWidget shows, even there I can't easily grab the contents of a single feed and share them in my sidebar -- and the shared "clip" like I do with Unblogged Bits has serious formatting limitations.

On the other hand ... I can do a lot more of this sort of thing than I could when the blog started (both personally, in my level of "expertise," and in what the Net provides in the way of services). Which is kinda cool ...

Filed under :: Blogging - Technical :: Food & Drink - Wine
Link · Print · Edit · Comments (0) · Pings (0) · TR/G

I finally found a replacement for Blogfuel to show my "Recently" Read/Watched sidebar block. The Wishlist Badge creates a nice little dynamic set of cover images for items that I've put on my Amazon "Recently" Wish List (which is sort of a Post-Wish List, if you follow).

(In other words, it's intended to point to an Amazon Wish List -- but I have a second Wish List, that I use to track things I've been consuming.)

Not sure what happened to Blogfuel to make it not work, but now I can continue to show off what I have been/am reading/watching, for your edification or entertainment. Yay.

Filed under :: Blogging - Technical
Link · Print · Edit · Comments (0) · Pings (0) · TR/G

Well, that little digression turned into a big time sink.

A few weeks back I converted my WIST quotations site into a static published site, rather than dynamic. That took a chunk of disk space, but the performance for calling up individual pages (and having them index) should save time.

So, today it I was in the mt-config.cgi file (to turn off the autosave function), when I realized it had been three weeks since I renamed my comment and trackback scripts. I've found that's a moderately helpful way of foiling certain spammers.

I was about to do that, when it suddenly occurred to me that, unlike before, not all of my individual entry pages in my various blogs were still dynamic. WIST's pages are static (actual HTML files generated at creation, vs. dynamic pages generated from the database on the fly). If I changed the names of the comment and trackback CGI files in the configuration, I would have to rebuild all my pages. Which, last time I did it, was a multi-hour task (that may have been in part because it was the first time I'd done it; I haven't timed it again lately).

Ugh.

So instead, I needed a way for the script names to be dynamic when a given page is called, but the rest of the page to be static the rest of the time. Here's what I did.

1. Create two new index templates, one for each script name. 

I figured I could use Server Side Includes (SSI) as the dynamic source of the CGI script names. You can create SSIs from MT without too much trouble (as the output file associated with a template), but to have them "built" with the value they need, so they need to be done as Index Templates (vs Template Modules), flagged to rebuild with each rebuild.

So I created a "dynamic comment script" index module that creates dynamic_comment_script.inc, and has as a single line:

<$MTCommentScript$>

That tag will return the of the comment script.  Then I did the same trackbacks.("<$MTTrackbackScript$>" etc.)

So now whenever a rebuild happens, those two .inc files will have the name of the script (e.g., "xyztrackback.cgi") in them. And when I change the name of the two scripts in my mt-config.cgi file, instead of rebuilding all the individual entries in WIST, I just have to do an index rebuild (which takes just a minute). And if I forget, it will still update the next time I add a new entry.

Note that SSI is not available on all hosts. It is on mine, though.

2. Change the comment form. 

Now to change the reference to the comment script. In the comment code, there's a form call that starts something like:

<form method="post" action="<$MTCGIPath$><$MTCommentScript$>"

That's the part that needs to be fixed. And, fortunately, it's simple.

<form method="post" action="<$MTCGIPath$><!--#include virtual="/dynamic_comment_script.inc"-->"

That's an SSI call there at the end. It's calling the contents of that .inc file I created in step 1, literally sucking it in at the time the page is loaded. Thus, the page is static (on file), but that particular piece gets pulled in dynamically. And recall that .inc file contains the name of the comment script, as most recently generated (even if that's after when the entry's static file was generated).

3. Change the Trackback text. 

I still have the default trackback address text at the bottom of the individual archive page, in case someone's doing a manual ping that isn't doing an auto-discover on the file. That line usually looks like:

TrackBack URL for this entry: <$MTEntryTrackbackLink$>

Instead, I do the same trick as above:

TrackBack URL for this entry: <$MTCGIPath$><!--#include virtual="/dynamic_trackback_script.inc"-->/<$MTEntryTrackbackID$>

MT has all sorts of tags for this sort of stuff, so it was easy enough to (by looking at what was actually generated) find the surrounding pieces and substitute in the SSI of the TB script in the middle.

4. Change the Trackback autodiscovery code. 

A lot of blogging software can autodiscover trackback addresses for a file through special RDF tags embedded in it. So MT has a simple tag to generate the RDF tags:

<$MTEntryTrackbackData$>

That actually creates a 14-line set of tags and info for the trackback discovery process. Unfortunately, part of that info is the location of the trackback script (so that the autodiscovering system can generate a trackback entry).

Fortunately, though long, the format of those tags is pretty straightforward and the content is reproducable. So in my Individual Archive template, in lieu of the one line above, I now have:

<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
        xmlns:trackback="http://madskills.com/public/xml/rss/module/trackback/"
        xmlns:dc="http://purl.org/dc/elements/1.1/">
<rdf:Description
   rdf:about="<$MTEntryLink$>"
   trackback:ping="<$MTCGIPath$><!--#include virtual="/dynamic_trackback_script.inc"-->/<$MTEntryTrackbackID$>"
   dc:title="<$MTEntryTitle$>"
   dc:identifier="<$MTEntryLink$>"
   dc:subject="<$MTEntryCategory$>"
   dc:description="<$MTEntryExcerpt$>"
   dc:creator="<$MTEntryAuthor$>"
   dc:date="<$MTEntryDate format="%Y-%m-%dT%H:%M:%S-07:00" />
</rdf:RDF>

All the stuff in the first block is literal info as generated currently by MT. The next stuff is all use of MT tags -- including, note, the SSI to get the current trackback script info into place.

The only thing I didn't bother to look up or figure out how to do is on the last line. The "-07:00" is the GMT time zone offset, in this case Mountain Time. I don't know if MT has a tag to do it, and I really didn't feel the need to look it up (since I don't plan to permanently move out of my time zone any time soon).

5. Kick back and relax. 

And that's it. With those steps, I now can change my comment and trackback script names, do a simple index rebuild on WIST (or even -- given that the trackback and comment traffic isn't all that heavy -- let it rebuild itself when I add more quotes each weekday), and the changed CGI script will be present in all of my entries without having to do a full rebuild.

Of course, you might say, I could simplify things by eliminating trackbacks -- but I'm stubborn about this, as I think the TB concept is delightful, and use it a lot for internal cross-references if nothing else. I hate to let the spammers "win" on that one. Ditto for comments -- the WIST site would not lose a lot by losing comments -- but I'd be irked and saddened. So, to me, it's worth the effort to have done this.

And to have shared the wealth with anyone else who's looking for something similar.

Filed under :: Blogging - Technical :: Spam
Link · Print · Edit · Comments (0) · Pings (0) · TR/G

Margie commented a few days ago that, in doing an iTunes upgrade, she'd been offered a chance to install Safari.

"Safari? That's a Mac browser."

"Well, it offered to install it for me."

"Huh. Weird."

Turns out that Safari is now available for Windows, too (not sure for how long) -- but Apple's drawing heat by slipstreaming it in as ... well, as an "update" to iTunes (which it's not), installed by default (bad dog!).

I agree with the critics -- new products, or products that are not the core of what you say you are trying to install, should be clearly labeled and should not be chosen by default. I dislike it when other companies pull that sort of trick ("No, for the twelfth time, I do *not* want to install Yahoo's browser bar while downloading your upgrade to something that has nothing to do with the browser!"), and I don't dislike it any less that it's Apple doing it this time.

Regardless, that's why I always do a "Custom" (rather than "Default") install when I load things onto my machine -- so I know what's actually being put in.

Filed under :: Blogging - Technical :: Hi-Tech
Link · Print · Edit · Comments (8) · Pings (0) · TR/G

Time once again to clean out those browser tabs ...

1. When is dead dead? Not a trivial question when you're talking about organ donations.
2. The Toilet Sign Project. There are some pretty clever ones here.
3. The worst foods in America. I've eaten ... well, one of these (the Cinnabon).
4. One of those stealth patent infringement suits may have Sony's Blu-ray in its sights.
5. Not exactly how you'd expect a stingray to kill someone.
6. For all I've bitched about Microsoft not adhering to web standards ... its moves toward doing so will likely cause more problems. (More)

Filed under :: Blogging - Technical :: Food & Drink :: Health :: Hi-Tech :: Media :: Potpourri
Link · Print · Edit · Comments (2) · Pings (0) · TR/G

I've been using a system by BlogFuel to list "recently read" books in my sidebar. Alas, the service seems to be not working (at least it's not showing up on my page any more, nor on Kate's), though there's nothing on the site to indicate a problem.

I'm poking around on Google for how to turn an Amazon Wish List into an RSS feed, which you'd think is a problem that should have been skinned twenty different ways. If I can get an RSS feed, I can use Google Reader tools to create a sidebar from it. No immediate, easy, trivial luck so far, but I'll try to continue digging. 

Filed under :: Blogging - Technical :: Media - Books
Link · Print · Edit · Comments (3) · Pings (0) · TR/G

The neat, sleak, redesigned interface for MT4 is driving me nuts. While it looks pretty, and I've managed to figure out where things go, and it even has some nice features (e.g., the post preview is excellent), I find it slow, slow, slow. Everything is a javascript construct, even the input fields. If you do anything to push the systems -- as I do with my 6,000-category WIST "blog" -- then it goes from manageably glacial (in MT3) to nearly unusable. That's as true the post creation side (where it takes about 30 seconds each time to generate the drop down to select categories) as it does on the category management side (where it takes forever to load, and then keeps trying to update itself periodically, all taking forever).

Ugh.

Oh, and the java-controlled drop-down menus are a pain in the ass, as they tend drop down on a casual mouse-over and then obscure the screen, waiting to be clicked on.

Ugh.

I know a big chunk of this stems from my using the system a way it's not really designed (thousands of categories is possible, but I doubt anyone else is doing it), but it's still a pain in the patootie, even in those cases where I do something on other than just WIST.

Filed under :: Blogging - Technical
Link · Print · Edit · Comments (1) · Pings (0) · TR/G