Rethinking My Architecture
As always, I'm rethinking my architecture.
Now that the wife and kids are all in bed, I've finally had some real time to go over what I've accomplished in the last two weeks and map out my next steps. The very next step is to get the RSS import functions working on the home page.
Here's the thing: I intend to export RSS feeds of all my content channels, and (as already mentioned) I'm going to be importing RSS from elsewhere; so the begged question is, "Why not just import my own feeds when building the homepage?"
Yeah, that would imply all the code I wrote over the last week in my spare time (hah!) was purposeless.
Actually, most of it is entirely reusable; it can just be swapped around so that it becomes my core RSS export algorithm.
You know, I went round this circle a hundred times when planning, so it irks me that I'm going around again. Something just seems wrong about exporting RSS documents from my SQL database, only to just slurp them back up again to post on my own site.
But what's the alternative?
Well, the path I'm going now will require me to publish twice, as it were, every time new content is added. Once to the xhtml on the site, once to the rss files left behind for content subscribers to pick up. Is that better than publishing once and then using the published rss materials as sources for my own homepage?
Actually, the rss generation is not necessarily synched with the xhtml generation. The plan for publishing, right now, is to ignore messing with cron schedules, and instead just publish the xhtml on the site when an author/editor commands it. The publishing model will actually generate static pages from the dynamic output of the php and database. When a new article is added, it can be previewed via the "live" php pages connected to the db, and when it looks good, the author just hits "publish", and the php will splurt out static xhtml that gets served on demand in end-users browsers.
The rss, on the other hand, is (was?) only going to be generated as requested by outside agents. I.e., there is no "rss file" sitting on the server, waiting to be picked up, getting regenerated every hour/day/week/whatever. Instead, when an agent requests the rss feed, a php script will dynamically generate one on the fly, using whatever the then-latest content is, as needed.
Hrm. Well, none of this changes the need to bring in rss/rdf/xml data from outside syndicates. I'll just bang that out and then revisit this internal problem afterwords.
Aside: BBEdit rocks. It's been a very long time since I've used it. I've got the 6.5 Demo running right now; it's much better than the 5.0 version I used back in the day.
Opera also rocks. And Netscape is back on track.
I used to be a major Netscape fan. I'm by nature irrationally anti-Microsoft, and since I got on board with web development within days of Mosaic being launched, I've got a soft spot in my heart for it's first child.
Still, Microsoft's no-need-to-think interoperability is nice. And the ubiquitousness of their "standards" makes many things easier. And ASP was a great development. And MSWord finally replaced WordPerfect on my desktop due to a couple of specific features. And Exchange convinced me I wanted Outlook in all its bloated glory at work, and thus my work pc became a windows machine. And when I was writing Java, the Explorer kicked Navigator's bee-hind when it came to conforming with the "standards".
And suddenly I'm a MS fan. My company is a Certified Gold Partner development company. (Not that we do only MS work, but the majority of our programming rev comes from clients that are distinctly MS-based.)
And then IE on the Mac, touted as one of the first "fully compliant" CSS1-savvy browsers, failed me. Right here, just in the last week or so. IE 5.1 Mac is rated around 98% CSS1 compliant. It's that final 2% that has me pulling my hair out. Their "float" selector is buggy as all getout.
And then, for testing purposes, I loaded a bunch of "forrun" browsers on my G4 at home, and on my Win2k machine at woik. And I spec'd out the conformance matrices for Navigator 6 and various vers. of Opera wrt CSS1. And you know what? I'm about to abandon Explorer as my browser of choice.
For now, at any rate.
EXCEPT THIS ONE PAIN IN THE REAR THING THAT IS HAPPENING RIGHT NOW!!! I'm in Navigator 6 on my Mac right now, typing away into the Blogger window (which I hope soon to replace), and I reach for my mouse, wander up to the top-right corner of the screen, hit "Post"...
...No, wait a minute - I don't hit post. THE BUTTON ISN'T THERE! Weird. I guess I'll just copy this longwinded post to the clipboard, fire up Explorer (or, actually, Opera, cause I'm perverse), and paste it in there.
<*shurrrp*>
[..........]
<*THPBLAT*>
Okay, I just pasted it into the Blogger window on Opera. It looks kinda funky, but there's a "post" button, so I guess it'll work.
Until next time....
--------
Now that the wife and kids are all in bed, I've finally had some real time to go over what I've accomplished in the last two weeks and map out my next steps. The very next step is to get the RSS import functions working on the home page.
Here's the thing: I intend to export RSS feeds of all my content channels, and (as already mentioned) I'm going to be importing RSS from elsewhere; so the begged question is, "Why not just import my own feeds when building the homepage?"
Yeah, that would imply all the code I wrote over the last week in my spare time (hah!) was purposeless.
Actually, most of it is entirely reusable; it can just be swapped around so that it becomes my core RSS export algorithm.
You know, I went round this circle a hundred times when planning, so it irks me that I'm going around again. Something just seems wrong about exporting RSS documents from my SQL database, only to just slurp them back up again to post on my own site.
But what's the alternative?
Well, the path I'm going now will require me to publish twice, as it were, every time new content is added. Once to the xhtml on the site, once to the rss files left behind for content subscribers to pick up. Is that better than publishing once and then using the published rss materials as sources for my own homepage?
Actually, the rss generation is not necessarily synched with the xhtml generation. The plan for publishing, right now, is to ignore messing with cron schedules, and instead just publish the xhtml on the site when an author/editor commands it. The publishing model will actually generate static pages from the dynamic output of the php and database. When a new article is added, it can be previewed via the "live" php pages connected to the db, and when it looks good, the author just hits "publish", and the php will splurt out static xhtml that gets served on demand in end-users browsers.
The rss, on the other hand, is (was?) only going to be generated as requested by outside agents. I.e., there is no "rss file" sitting on the server, waiting to be picked up, getting regenerated every hour/day/week/whatever. Instead, when an agent requests the rss feed, a php script will dynamically generate one on the fly, using whatever the then-latest content is, as needed.
Hrm. Well, none of this changes the need to bring in rss/rdf/xml data from outside syndicates. I'll just bang that out and then revisit this internal problem afterwords.
Aside: BBEdit rocks. It's been a very long time since I've used it. I've got the 6.5 Demo running right now; it's much better than the 5.0 version I used back in the day.
Opera also rocks. And Netscape is back on track.
I used to be a major Netscape fan. I'm by nature irrationally anti-Microsoft, and since I got on board with web development within days of Mosaic being launched, I've got a soft spot in my heart for it's first child.
Still, Microsoft's no-need-to-think interoperability is nice. And the ubiquitousness of their "standards" makes many things easier. And ASP was a great development. And MSWord finally replaced WordPerfect on my desktop due to a couple of specific features. And Exchange convinced me I wanted Outlook in all its bloated glory at work, and thus my work pc became a windows machine. And when I was writing Java, the Explorer kicked Navigator's bee-hind when it came to conforming with the "standards".
And suddenly I'm a MS fan. My company is a Certified Gold Partner development company. (Not that we do only MS work, but the majority of our programming rev comes from clients that are distinctly MS-based.)
And then IE on the Mac, touted as one of the first "fully compliant" CSS1-savvy browsers, failed me. Right here, just in the last week or so. IE 5.1 Mac is rated around 98% CSS1 compliant. It's that final 2% that has me pulling my hair out. Their "float" selector is buggy as all getout.
And then, for testing purposes, I loaded a bunch of "forrun" browsers on my G4 at home, and on my Win2k machine at woik. And I spec'd out the conformance matrices for Navigator 6 and various vers. of Opera wrt CSS1. And you know what? I'm about to abandon Explorer as my browser of choice.
For now, at any rate.
EXCEPT THIS ONE PAIN IN THE REAR THING THAT IS HAPPENING RIGHT NOW!!! I'm in Navigator 6 on my Mac right now, typing away into the Blogger window (which I hope soon to replace), and I reach for my mouse, wander up to the top-right corner of the screen, hit "Post"...
...No, wait a minute - I don't hit post. THE BUTTON ISN'T THERE! Weird. I guess I'll just copy this longwinded post to the clipboard, fire up Explorer (or, actually, Opera, cause I'm perverse), and paste it in there.
<*shurrrp*>
[..........]
<*THPBLAT*>
Okay, I just pasted it into the Blogger window on Opera. It looks kinda funky, but there's a "post" button, so I guess it'll work.
Until next time....
--------