Welcome to Orcmid's Lair, the playground for family connections, pastimes, and scholarly vocation -- the collected professional and recreational work of Dennis E. Hamilton

 

Computing Milieux

Cooperative, Collaborative, and Social Software

Standing in the Users World ... And Not

There are titles here again because posts go to their own files, and headings and such are sometimes worthwhile.  This particular running monolog continues unfettered ...

Stopdesign | The New Blogger.  I found out about "The New Blogger" by selecting Blog This! yesterday and having an unexpected experience.  I am having that experience now, too.  I can't post any longer, I must save as Draft or Publish.  And if I go to the Edit Posts via the button that is provided when I do save this draft, I will see something completely unrecognizable and a bit unpleasant.  So it is interesting to read here how much effort was invested and what is thought to be cool.  I want to know how this team arrived at what constituted an improvement, and for whom?

[later 2004-05-10-15:31 -0700] This is starting to piss me off.  I discover that the prohibition of <br> and apparent failure to break paragraphs in comment previews is a preview bug: the double-space does cause a paragraph break, but a <br> and an indented paragraph would give more cohesion to the comment.  There's a bigger surprise.  Blogger now saves independent pages for each blog entry as well as providing the lair.asp recent-blogs page and the archive pages that have always been produced.  The new individual pages are inside of a subdirectory structure that I was given no choice about.  In addition, that breaks the relative links that I have in my template to (1) the atom feed icon, (2) other intra-site icons and targets (such as the Hard Hat Area), and (3) to the archive index.  The archive pages, built off of the archive index, have their permalinks to the independent pages, and comments have their own permalinks to fragments of those independent pages.  Now I have to figure out how to get some coherence in this structure.  I wonder if there is no way to get there from here!  Grumble, grumble, grumble.

[later still 2004-05-10-22:44]  I am just wrapping up the manual refreshing of every post in the current blog page (back to 2004-05-03) so that the permalinks are correct in the Atom feed and on the page itself.  I don't know about all recent archive pages yet.  I will probably have to refresh all other posts for the archive having 2004-05-03 in it also.

[still later 2004-05-10-22:49]  I even refreshed the posts that were properly permalinked in the new folder structure, because the page template had to be corrected and reapplied to properly make links to images, buttons, and other parts of the site regardless of the level of folder nesting at which the template is used.  I am not going to make any more repairs tonight, because I am fatigued and may make too many mistakes. Instead, I will backup my work to my development image of Orcmid's lair and see what else to do in the morning.

[day later 2004-05-11-10:16 -0700]  Jeffrey Veen posted an article on how Working with Blogger is Fun, and I left some rude words last night and this morning while waiting for Blogger to quiet down and let me in to edit more material.  There are interesting problems when interaction design does not account for the entire user experience, not just the part directly-attributable to the facade.  (For one thing, it reveals that the facade is just that.)  There are also matters of conceptual models, failure modes, and what happens when the user stumbles into an incoherency crevice.  The breakdown here is that interaction design was applied to achieve an user experience, and that seemed to have dealt with the psychodynamics of rounded corners on logos, but not the experience of actually driving the car in traffic.  Jeff is probably not the guy to talk to about this, though I think his perspective (and Alan Cooper's) would be valuable for discussion.  Nevertheless, people are linking to Jeff's blog as if it is that of the designer and the architect.  He should not be too surprised that his possibly-premature self-congratulation and exuberance over work on the Blogger redesign may have his team be tarred by the same brush as those anonymous technical guys who seem to have missed key aspects about risk management, disruptive wholesale change, and annoying oversights in the plumbing.  Jeff has a case study on the Blogger project and it will be interesting for a little forensic analysis here.  One thing that sticks in my craw is the clear imperative of being standards based and then offering an apologia around how the current template pages do not produce valid HTML.  This makes no sense at all, especially the intimation that we are going to endure on-going development and change in vivo.  I can't wait to see what was done about accessibility standards.

[mid-day later: 2004-05-10-12:47 -0700]  I see that I somehow missed reposts on some of the items back through 2004-05-03 and that I will need to touch them up to repair bogus permalinks.  That is not too difficult.  I also see that I can completely restore site integrity if I go back to 2004-05-02 and repost those entries in the new post format, with comments enabled.  Then I can lock down everything prior to 2004-05-02, make sure it is backed up on my development site, and never ever do anything that leads to republishing of any of those entries.  Those are history, baby.

One nice thing about the blogger redesign did show up.  There was a bootstrap effort, even if it seems to be fragile or have failed, depending on how badly the mess is if you are burned by it.  The original archive structure and index approach is preserved, including the fragment-identifiers of indexed articles.  That is, if I have handed out links and made trackback pings using old permalinks of previously-archived entries, those links still work.  The old permalink reaches an entry on the weekly-archive page, and the permalink on its time-stamp then goes to the (new) single-page post that is now the ultimate authoritative entry.  Although this transitional model is a little weird, it does keep previously-referenced archive material reachable, even if it has been updated to the new post format for the purpose of enabling comments or simply cleaning up a permalink-klobber mess left by Blogger. (I guess I shouldn't call them archive pages any longer, since the weekly "archives" I have Blogger maintain are no longer authoritative, but it escapes me what the appropriate concept might be in the new model.)

It just dawned on my how to have the lockdown work though. I can change the name given to the weekly archive pages, so that new ones won't ... uh, never mind, I have to figure out how to preserve the index before I try anything like that, though I have an idea.  Enough for now.  I take a break and then complete cleaning up the broken permalinks.  Then I'll worry about locking down everything prior to 2004-05-02, and I'll come at it top-downward rather than the bottom-upward way I almost tripped myself up with.

Note to myself: The individual post format needs its own template, in contrast with the page of multiple latest-posts and the weekly archives (or whatever granularity those pages are created with).

Second note to myself: If the daily-weekly-monthly or whatever chronological archive pages are still producable, then it is not true that a calendar of where there are entries can't be produced to facilitate navigation when editing posts, especially when having to reach back a ways and clean up posts, as I will when going back to 2004-05-02.  What may be more important is that producing a calendar on the page may not be all that difficult.  I am congratulating myself that I had the foresight to use *.asp filenames for all of my Blogger pages and I can introduce server-side solutions for this at some point.
posted by Dennis at 5/10/2004 01:17:15 PM