Month: 2002/08
-
2002 August 30
-
LiveJournal largely ignored, but why?
Caught this snippet on tweeney.com via JOHO: "...LiveJournal.com (which most weblog news stories overlook for some reason) boasts more than 650,000 [users]..." Why does everyone seem to ignore LiveJournal? It's very, very, very rare that I see a LiveJournaller's posts linked into the Blogosphere at large. Granted, I know that the median user was a 15 year old female who complains about Mom and her boyfriend, when last I checked. But, as the adage goes: 90% of anything is crap. There are, nevertheless, a good number of worthwhile streams of narrative in that space. On the other hand, I don't see many of the people behind LJ stepping out and making noise in the Outer Blogosphere either. I think many of them are just plain busy keeping the site afloat, or having lives, and LJ is world enough for them. But 650,000 users... that's a lot. More than Radio and rivalling Blogger.com. Is there a real qualitative difference in writing between the groups? I would still imagine there's a lot of crap to be found via Blogger.com. I'm not sure about Radio, though, since I get the impression that the 15 year olds have yet to flood into the userbase and its following seems more tilted toward professionals. But as for the software & service itself... As far as I can see, LJ is one of the easiest paths out there to starting a weblog/journal online. And it was one of the first sites I ever encountered that had a desktop-based client app for posting to it. And, though not prominently placed, they have RSS feeds for every single journal on the site. They're even working RSS aggregation features into the place by gatewaying external RSS feeds in as special LJ users to be added like any other LJ "friend". So, to me, LJ sounds like a top competitor to every other blog/news aggregation product or service out there - yet I rarely hear about it. Hrm. Anyone have a theory why? [ ... 1126 words ... ]
-
Interop: Movable Type IM-based pub/sub?
Even more gas is flowing - from andersja's blog: Movable Type notifications to Instant Messenger? I really want to see this, and I want to see news aggregators exploit this - why poll a feed once an hour when you can just have the feed tell you to come 'n get it? Basic publish/subscribe model. But the advantage Radio has over MT in this regard is this: Radio is a persistently running app/server/daemon thing. MT is a collection of scripts that does nothing until asked to run. Radio can connect to IM and stay connected. Something like MT would need to login to an IM service each and every time it wanted to ping. Maybe that's not such a big deal, really. I also have a hunch that there would be some difficulties with web hosting sites who don't really want customers emitting IMs from their CGIs. Maybe not a big deal either. Just seems like an impedance mismatch, though. What I'd like to see is something like this: An XML-RPC/SOAP <-> IM gateway. And then, eventually, I'd like to see a decentrallized P2P network with XML-RPC/SOAP entry points that can smoothly replace centralized resources that have XML-RPC/SOAP entry points, maybe using IM networks as one possible rendevous point. (Just remembered this project: JXTA Bridge. Mental note - play with JXTA again and poke at SOAP some more.) [ ... 483 words ... ]
-
Outlining + IM = On Fire
Dave writes, "I have my instant outliner going again." Kick. Ass. Now we're cooking with gas. I think I need to get a fresh install of Radio going on this iBook again, especially since it seems renice works now and can tame the CPU hungry beast to managable levels. I was kinda waiting to see if/when the people hacking with IM in Radio would close this loop. Now I want to see what this can do to news aggregation, pub/sub, change notification, etc and more. [ ... 86 words ... ]
-
2002 August 29
-
Would the real RSS please stand up?
Sam Ruby responds to Dave's question:I am not a stakeholder in this naming issue, nor can I claim based on personal experience that any of the above references are authoritative, but based on these references alone, it seems to me that one could ask the converse to both of these questions: i.e, why is RSS 0.91 called RSS, and why did the RSS 0.91 branch use instead of [rdf:rdf]? This is basically what I'd started writing as a response the other night, but I lost it in a browser crash. I browsed around a bit, revisited some old links I'd followed when first I started hacking with RSS and frustratedly discovered the format fork. As Dave tells it, his adoption of RSS started in cooperation with Netscape, and the later resulted in his continued support and development after Netscape had wandered away from it and shed the developers responsible. So from his seat I would think that it looks like RSS was abandoned in his lap and thus he felt free to continue on shaping it toward his own ends, namely those of simplicity and real dirt-under-the-nails use versus further design and planning. Then, later RSS 1.0 comes into the picture. About this point in history I'm fuzzy. Most accounts and articles I see do not mention how v1.0 came about and who birthed it. Most of what I've ready just basically says, "And then there was v1.0" But who said so and when and why? That 1.0 has more in common with the v0.9 roots than any of Dave's v0.9x series is clear, so maybe in this respect one could say that this naturally lends v1.0 a "natural" claim on the name by virtue of bloodline and original intent. On the other hand, one could also say that Dave's v0.9x series has a claim on the name due to virtue of having actually been in active development directly after the original v0.9. Call this a claim by virtue of squatting rights? By virtue of principles of Do-ocracy, as Sam wrote about Apache? But then, on another set of hands belonging to another body altogether, why did Dave keep the RSS name if he was so radically changing the nature of the format by ditching RDF? (But the history confuses me - I seem to remember that change started when it was still in the hands of Netscape.) So where does that leave things? Ugh. Seems like Dave's RSS should be so because he kept the torch burning in the Dark Times. But it seems like the Other RSS should be so because it's the heir to the original throne. Then again, I could have the whole history mixed up. I say, everybody get up! I'll start the music, you all run around, and pick new format names before I hit stop on the CD player. [ ... 547 words ... ]
-
2002 August 26
-
"...ideas are fucking worthless."
...that's what's got me so bothered about people musing in their weblogs about projects they'd like to do. Stop talking about it an just build it. Don't make it too complicated. Don't spend so much time planning on events that will never happen. Programmers, good programmers, are known for over-engineering to save time later down the road. The problem is that you can over-engineer yourself out of wanting to do the site... [Andre Torrez, Even You Can Do It] Just found this post today, via Danny O'Brien. This is why I threw this site together, and why I have a link to ReleaseEarlyReleaseOften on the front page. Before this, I would spend years working on something in silence, only to have it fall over on top of me and end with me never wanting to touch it again. For almost 2 years, I was working on a Zope replacement in Perl called Iaido before I finally created a ?SourceForge project for it and invited some people in to play. By then, I was already disgusted with my ugly code, wanted to scrap it and restart, but wasn't nearly enthusiastic enough to do that. And by then, there was just too much code - and yet too little documentation - for anyone I invited in to really dive in and muck about. And this project was to be the core of a community site for coders, web designers, and general all around mad scientists. It would be named ?NinjaCode, at http://www.ninjacode.com. Well, you can see, the community never got off the ground, and I don't even have a hold on the domain anymore. It coulda been a contenda. But you can see now, on my Projects page, that I've been gradually working up and spinning out little hacks and widgets. Eventually, they combine into bigger widgets, like MTXmlRpcFilterPlugin. I'm thinking that this is the way to go. And, even if/when I do get a grandiose idea, I need to start off releasing the widgets early and show the build up process here in this weblog. Then, there's some documentation from the start and maybe even some enthusiastic co-conspirators from the start. [ ... 425 words ... ]
-
2002 August 25
-
AmphetaOutlines fix download page fixed
D'oh. Just realized that the download link for the fixed AmphetaOutlines was at the bottom of the page, so I was puzzled that I kept getting emails about it doing the same broken things. I seem to be missing many things lately. :) Anyway, the correct download is here: AmphetaOutlines-20020822.tar.gz: This should include the corrected "URL=" redirect. [ ... 76 words ... ]
-
2002 August 24
-
Comments rise from the ashes
Well, that sucked. I managed to get an all but complete dump of my comments.db file with db_dump on my server, but then nothing would parse or load that file after that. Banged around with it for awhile until I finally realized how to parse the dump file myself with a perl one-liner and rebuilt the comments DBM file that way. Immediately after that, I migrated everything to MySQL. Not a silver bullet, but it seems a better idea than relying on those DBM files. They seemed neat & clever at first - but now that I have a bit invested here, they're an annoyance. I think I managed to recover the last few comments I got while things went wonky, but if I happened to miss one of yours, I'm sorry. I'm glad I recovered the comments, though, because I tend to think that comments left here are often more valuable than whatever dross I may have spewed in the first place. [ ... 164 words ... ]
-
2002 August 23
-
Dead comments, they keep calling me
Grr again. As a few of you have told me in email, and as I noticed toward the end of yesterday, my comments feature here is dead. Seems that as the disk filled up on this server, someone tried leaving a comment, and as Murphy came into play, the comments DB file got corrupted. The odd thing, and fortunate thing for me anyway, is that I'm still getting the comments emailled to me. So if/when I get this DB file recovered, or if I wipe it and start over (not happy about that option), I might try re-posting the comments. Got some good pointers on some hosting options and may be checking out one or another of them soon. Also looks like there are still issues with my AmphetaOutlines, even after a re-release. More soon. [ ... 136 words ... ]
-
2002 August 22
-
A warning about MTXmlRpcFilterPlugin use
A few words of warning, which should be pretty obvious, with regard to MTXmlRpcFilterPlugin use: Every appearance of the tag makes a separate web service call. This could be painful if you stick it in your individual archive item template, and then rebuild your site. (ie. at least one hit on the filter service for each item in your weblog) :) [ ... 62 words ... ]
-
AmphetaOutline fix. No, really.
Ack! I just realized, even though it took two emails make the light go on, that my tarball of AmphetaOutlines for download in the wiki has still had the broken click-count page bundled for the last month or so. For some reason I thought that I'd fixed that and re-uploaded the tarball, but no! So: It's there now, hopefully this is a fix for any of you who tried it and saw the insane refresh loop happen with you clicked on a news item. Hop on over to AmphetaOutlines and grab yourself a new copy. Next... I screw around with Bayesian filters as applied to incoming news items, categories of interest, and alert levels. Maybe. [ ... 288 words ... ]
-
Of hosting woes and hippy nerd commune servers
Grr, shame on me for getting cheap hosting again. For the most part, PHPWebHosting has been just fine - they give me SSH access to a shell, about 120MB of space, a few MySQL databases, and generally leave me alone for US$10 per month. But at this point, the leaving me alone bit isn't working out so well, since the server I'm on over there has had its main disk fill up regularly, thus bringing down all the sites hosted on it (including mine). I'd joked a little while ago about getting myself an XServe and finding some affordable colocation. Now, though, I'm not laughing so much. I should probably just find another webhost, but it's tempting to have my own server. In the past, I had a few friends at an ISP, so they let me stash a little linux server on their network for occasional favors. I miss that. Didn't cost me anything, and my impact on their network was negligible. Meanwhile, I had a stable, semi-reliable box on the net at which I could point a domain. Services could catch fire and other things could tank, and no one got hurt. The problem now, though, is that shelling out for a server and colocation costs just to play around is a just a smidgen outside my budget. As I've pondered before, maybe I could actually host a money-making service on it with some of these hacks I'm percolating through. On the other hand, a co-worker put the idea in my head that it might not be unreasonable to think that I could invite some tenants onto my server to share some of the cost. I figure I'd take the burden of buying & owning the server, but I'd like to spread the service cost out. I wouldn't want to make a profit, I'd just like to make a kind of nerd commune. It'd be nice if I could get the cost for everyone under US$10-15, though I imagine my cost would be higher given a server payment. If I could get mine below $40-50 a month, I think I'd be happy, though this seems like a pipe dream. The devil being in the details, however, I don't really know how one would go about such a thing, considering service agreements and contracts and taxes and blah blah blah. I'm a financial moron, and I'm also certainly not into being a 24/7 available customer service rep, or to perform tech help. Basically, I'd like to run a server to host mad scientist experiments performed by a relatively small cadre of mad scientists who can mostly clean up after themselves. I'd like to run an IRC server, maybe a Jabber server, maybe a few random half-baked server ideas. On the other other hand, is anyone out there doing this already and want a new tenant? :) [ ... 821 words ... ]
-
2002 August 21
-
Whew, take a breath, get some lunch.
Yikes. I think that was the longest entry I've posted here - I usually reserve my rants, opinion posts, and various longer prose for my LiveJournal. Maybe I'll actually start showing a little more personality around these parts too. :) [ ... 41 words ... ]
-
On grousing about Dave's exit from a mailing list
Scanning news today, I see in my referrers a link from the new aggregators discussion group to my outliner AmphetaDesk skin. Curious, I go check it out and see what looks like a bit of grumbling between Dave Winer and Morbus Iff. I make my way back to near the head of the thread to see what's what, where I find a link to Morbus' post of opinioned and somewhat ranty disagreement toward Dave's definition of the modern News Aggregator. What disturbs me is that the grumbling quickly ends with:Bill -- I'm going to unsubscribe from this list now. The neener boys are in charge. Take care. Dave Now first of all, I too disagree with Dave's definition, and have posted as much. Although I did so in a less rant-mode manner, both Morbus and I pointed to the fact that the news aggregator definition served the goals of Radio 8.0 marketing, being that the definition very specifically identified the features of his own product. Rant or not, the point is made. What disappoints me about Dave's response on-list, though, is that instead of addressing the content of the disagreement, he goes after the ranting and hits back with a smidgen of ad-hominem of his own while calling for the "high road". Then, once he's tagged on the ad-hominem he unsubscribes and dismisses the list altogether, never actually acknowledging the original point made. Now, though I don't have a link, I've read several times where Dave has disparaged discussion groups and offered linked weblogs as the superior alternative. I've read his essay on Stop Energy. I'm aware of his position toward the grousings and grumblings and flamings that go on within a discussion group. And I've nodded my head almost every time he's written about these things - freeform discussions on the net can sometimes - but not always - be a clusterfuck of morons and Stop Energy. Maybe it's just because Dave is so intolerably fed up with any hint of a clusterfuck that he dropped off the list. I wouldn't put up with much shit either, especially after heart surgery. But to me it looked like pretty defensive behavior right from the get-go. Morbus even gave somewhat of an apology (for him :) ) and moved on to politely hit the issues. I guess this bugs me a bit, along with other UserLand-related thoughts I've had lately, for this reason: Here's a discussion group devoted to the very nature and future of one of the headline features of RadioUserLand. And the people there are some of the most prevalent names I've seen appear time and time again in relation ot the subject. I would think that this would be a dream group for him, in terms of driving and benefiting from innovation. Instead, it takes only a small amount of rant to drive him away. See, I've never met him, but I think I'd like Dave personally. Yes, I know his writings often have one foot or both feet in marketing for his products. But, I can't recall his denying this, and I see it as the natural behavior of someone who really thinks he's got something hot. He gives favor to his customers in his posts, and he has his own opinions. But I know these things, and I still like his software, and I still buy it. On the other hand, it's clear that sometimes he doesn't give equal time to points which may be contradictory or unfavorable to his side of things, and sometimes he gets things wrong. But he's not a journallist - he's a blogger who writes from the hip and sees what flies. Whether he lies outright is another story I haven't the first inclination to follow up on, but the very nature of his product stands to check and balance to him there. He can't get away with too much with his own customers calling him on things. Anyway, I'm starting to ramble. What really bothers me about the quick dismissal of the list seems like a "See?! This is what I mean!" demonstration on-list, and I almost expect a post later again promoting weblogs over DG's altogether. But I didn't even see him give the list a chance, and I think that's sad given the brain power in that roster. Granted, it's his perogative to have slim thresholds for annoyance, but I'd think this would be value for his company and product. As I said, I like his software, but he's refusing free help and consultation on the future of his product. I see those involved in groups like this taking their products beyond his, even those that were originally direct clones of his. Maybe this ends up tying into his grousing on Open Source versus commercial software eventually. On the other hand, maybe he just thinks he doesn't need any help of this sort and has the sheer magnitude of innovational mojo to leave us all in the dust. If so, rock on. :) As for me, I just joined the list. [ ... 988 words ... ]
-
And I didn't even know 0xDECAFBAD had soul!
Well, imagine that! Someone doing something in the spirit of my site! Neat! :) [ ... 29 words ... ]
-
2002 August 20
-
Note to self
Quick mental notes: Ditch the blogroll or make better efforts at updating it. Publish my news aggregator subscription list in addition to, or instead of the blogroll. Examine referrer records for readers of my RSS feed to find readers of whom I'm unaware. Visit all the sites listed under my stats at the blogging ecosystem site. Write more, both here, and over at LiveJournal. Possibly revive The Spirit of Radio and my RadioUserLand hacking efforts. (ie. Make Radio run blog entries through my Wiki filter) Reskin this place, simplify and make accessible and less ugly. Reduce number of toys, or at least give them all off switches. Get back to work. [ ... 112 words ... ]
-
First semi-satisfied plugin customer
Well, it looks like Josh Cooper has gotten the MTXmlRpcFilterPlugin and XmlRpcToWiki working on his blog. Yay hooray! Now, I just have to see about improving this thing if I can, and to make this place look better. This sudden easy abundance of links makes me think that they draw too much attention from the surrounding text. This is desirable, but not to the degree that they do now. [ ... 70 words ... ]
-
Protect your customers from the RIAA
This ISP has the right idea: treat the RIAA hacking threat like any other hacking threat. Let's see how long before this is made illegal. In the meantime, I'm trying to come up with a way to convince myself that it's worth buying an XServe and colocating with them. But first, I've got to come up with a business model to make some money to afford the toys. :) This will likely never happen. [ ... 75 words ... ]
-
2002 August 19
-
Movable Type plugin for Wiki Formatting and XML-RPC Filtering Pipelines
Last week, after reading what Mark Pilgrim had to say about macros in MovableType, I made a mental note to finally circle back around to hacking together my WeblogWithWiki now that MovableType has plugin features. Turns out it was so much easier than I thought it would be. MovableType's plugin scheme is dead simple, which hopefully means that plugins will flourish like mad. First, I hacked together MTWikiFormatPlugin. This plugin simply implements a new container tag, MTWikiFormat, which runs the contents of the tag through CPAN:Text::WikiFormat. This doesn't actually integrate with any existing wiki, but it is very simple to install and does bring some wiki-ness to blog entries, including some limited formatting and Wiki:WikiWords. This doesn't provide everything Mark had posted in a lil wishlist comment to me, but it's a start. Maybe I'll look into tearing the formatting guts out of some wiki to make a Text::WikiFormat replacement, or maybe I'll submit patches to the original module. The second plugin though, MTXmlRpcFilterPlugin, is what I'm really happy about. Whereas MTWikiFormatPlugin filters content through one perl module, MTXmlRpcFilterPlugin can filter content through one or more XmlRpcFilteringPipe interfaces. I have a handful of these filters available on my site right now, and in a little while I will catalog them in the wiki. For now, I'm just filtering this entry through DecafbadWiki. In the future, I may get more adventurous with my content filtering pipeline. One drawback to using MTXmlRpcFilterPlugin for the purposes of a WeblogWithWiki is that I've only got support for TWiki so far in my XmlRpcToWiki project. Other wikis still need some hacking before they can provide filters. Some assembly required, fellow AlphaGeeks. So, ShareAndEnjoy. Time for bed. [ ... 1317 words ... ]
-
Working on a Wiki format MT plugin
Please excuse the noise and dust - I'm working on a few new MT plugins to support Wiki formatting and XmlRpcFilteringPipe. Are these wiki words working? PipeFilters, MovableType, ShareAndEnjoy, etc. How about unordered lists? * One * Two * Three How about ordered lists? 1. One 2. Two 3. Three And ''how'' '''about''' '''''this'''''? This is another test. PipeFilters, ShareAndEnjoy. One One a two three four two a How about a table? this is a test table format [ ... 166 words ... ]
-
2002 August 16
-
I love it when Dave invokes Murphy.
Just read this on Scripting News:Heads-up, some time in the next few hours (Murphy-willing) we're going to release tcp.im, which allows Radio and Frontier to be an instant messaging client or server (either can be either). Okay, this is insanely great, or at least has gigantic potential to be great. I just hope ActiveBuddy doesn't swoop in and claim to have invented the whole kit and kaboodle. Barring that, I can see a whole suite of new P2P apps with desktop servers, kicked off by the UserLand crowd. This makes me want to find a decent machine for Radio to run on and give it a fair amount of attention again. I wish it weren't so demanding of my poor iBook. But I do need excuses to convince myself to invest in a new ?TiBook... [ ... 135 words ... ]
-
Bayesian filters, the ultimate anti-spam weapon?
Do not stand in our way - we will walk around you. Spam is in our way. We will use the inhuman aspects inherent to your sales pitch to walk around you. Cluetrain on statistic-driven autopilot. :) [ ... 186 words ... ]
-
2002 August 15
-
It's about the dots that connect themselves
Further self-directed dot connection via Sam Ruby:I talked Sanjiva Weerawarana into creating a Radio weblog. He's one of the driving forces behind many important Web services standards including WSDL and BPEL4WS. Welcome to the Blogosphere, Sanjiva! :) [ ... 37 words ... ]
-
ChoiceMail, the ultimate anti-spam weapon?
I just read about ChoiceMail, a whitelist-based email service that "turns your inbox into the equivalent of your front door" where people have to knock and identify themselves as human via a quick web form (hello, Cluetrain) before their message is allowed in. This is exactly the idea I've been tossing around in my head and reading about for some time now. I would pay for this service in an instant, but unfortunately it apparently requires Windows. So, I just got SpamAssassin and Razor working on home IMAP server, so I think I might just poke at finally implementing a service like ?ChoiceMail on my own. Hell, maybe I'll even have it send me Jabber messages for permission confirmation and accept messages for whitelist additions, but keeping everything in the context of email seems better. [ ... 344 words ... ]
-
Java Web Start r0xx0rs
Another reason why I need to poke around at Java Web Start development, found via the rebelutionary: jlGui, a 100% Java music player that looks great, seems to work nicely, and launched right from my Mac's JWS panel. [ ... 67 words ... ]
-
My god, it's full of connected dots
Sam Ruby helped in connecting the dots between me and someone responsible for my recent Bean Markup Language obsession. Thanks Sam! :) [ ... 23 words ... ]
-
Living la vida regex
Better living through regular expressions from Mark Pilgrim. This has got me thinking of two things:I need to hack together some MT macros or a plugin that finds & links wiki-words in blog entries to my wiki, and I need to start writing in the wiki again.I really dig his blog's skin, and I need to finally read though all of his accessibility tips and re-skin this site. This place is butt-ugly. [ ... 288 words ... ]
-
Bean Markup Language creator speaks!
Wow, that was fast. I got a direct response from one of the two original creators of BML, Sanjiva Weerawarana:...I still believe BML has a useful role in life. We still have some of our original trusty users who periodically email us asking for status / updates etc.. What's the right thing to do? I am quite certain that we can get it open-sourced (I mean from an IBM process point of view), but I haven't yet been convinced that there'd be a willing community for it yet. This tells me that there's still some life left in the project, so I certainly won't be running off to release my unauthorized decompilation. Besides respect of ownership and authorship, it'd be nice to see a sanctioned release of the real code, comments and all. Sanjiva asks what to do and who wants it. I say: Gimme, and I want it! But of course there's more, in terms of effort and consideration, to an Open Source project than that. So, where's the interest (besides me) to make it worth doing? I'm not sure. Seems very nifty to me, but again, I see barely a mention on Google. But, I look at Thinlet buzz that recently bubbled through my news scan, with raves about the XML wiring for GUI and quickness in assembling apps. It's very cool. And then I wonder how cool might it be to combine the two, or use BML instead of Thinlet. Is the lack of interest a product of a lack of word getting out? Or, again, have I missed something in my return to the world of Java? [ ... 418 words ... ]
-
2002 August 14
-
Bean Markup Language revealed, sorta!
Hmm... The plot thickens a bit for me with regard to the Bean Markup Language project. On a whim, I Google'd for a Java decompiler, having remembered that I'd gotten some use and enlightenment back in the day from Mocha. So I found, downloaded, and tried Mocha on the BML class files. It choked a bit, produced some source files, but the collection as a whole was not trivially easy to recompile. Pretty much what I expected, given that Mocha is a relic these days. But then, I noticed JODE, a few search results down the page. Google's description said, "Java decompiler with full source code. Supports inner/anonymous classes..." Well, inner/anonymous classes are new to me since I was last very active with Java. (Yes, I know, that's been awhile.) So I figured I'd check JODE out. Besides, the last release of JODE looked newer than the last release of BML. Much to my surprise, JODE consumed the BML jar file directly and gave me a directory full of source without a complaint. For the hell of it, I compiled everything there, and made a new jar file from the results. I cleaned my project out, replaced the original BML jar with my new decompiled-recompiled archive and... everything worked just fine. Skimming through the various source files JODE gave me, things look surprisingly less ugly than I'd expected. Of course comments are nowhere to be seen, and variable names are nothing like the original would have been, but the source was is still readable and I can follow it. So this means I have the source code to BML now, after a fashion. So my question now is... I read the IBM ALPHAWORKS COMMERCIAL LICENSE AGREEMENT which came with the original download, and I see these paragraphs:Permission to reproduce and create derivative works from the Software ("Software Derivative Works") is hereby granted to you under the copyrights of International Business Machines Corporation ("IBM"). IBM also grants you the right to distribute the Software and Software Derivative Works. You grant IBM a world-wide, royalty-free right to use, copy, distribute, sublicense and prepare derivative works based upon any feedback, including materials, error corrections, Software Derivatives, enhancements, suggestions and the like that you provide to IBM relating to the Software. Does this mean, basically, that it's okay for me to distribute this source as a derivative work, with various potential enhancements, &etc, as long as IBM is still able to grab a copy of it back from me with no strings attached? If so, that's great. My only other question remaining, though, is whether or not my naivete toward Java has me yet to find what the community at large considers the Right Way to work with ?JavaBeans. Because, this BML thing seems great to me, but it seems to have gotten next to zero attention. This usually tells me that I'm missing something, but the thing itself works nicely. Any thoughts out there in the blogosphere? [ ... 494 words ... ]
-
Where art thou, Bean Markup Project?
Ack. Okay, so I started tinkering with Bean Markup Language as the configuration glue for my Java app. So far it seems to be working very nicely and doing everything I want it to do. The problem, though, is that as I look around for examples and things I notice that it seems as if this project is dead. I look further, and I realize... I can't find any source code to this thing. And here I am working it into the core of my app. Well, I'd hate to start over and reinvent this wheel, since it seems to be such a nice wheel. Between the clean definition of beans and their properties, and the ability to toss little bits of script (using the BSF) in there to further tweak things, it seems perfect. But the apparent abandonment of the project and my lack of source gives me the willies. So... Does anyone out there know anything about what's happened with the Bean Markup Language project? Could anyone get me the source? Or, could someone point me to an equivalent alternative? Because, if I can't find source or an equally powerful alternative, I'm tossing around the notion of re-implementing BML with a clone project if the license doesn't prohibit it. Or am I completely crazy and barking up the wrong tree to begin with? Is there a better, more elegant way to create and connect Java components together than this? And is it free? [ ... 945 words ... ]
-
Further adventures in Java!
So I've been in the underworld these past weeks, head down in battle with this project in my return to Java, as I'd written last month. It's been great fun, and I've not missed perl at all along the way. (Although, I did miss CPAN.) I've raided the stocks of Jakarta and other shores, and I've flagged a few new people to watch on my daily scan. These past 4 weeks or so, it's amazing to me during how many new or long-neglected (to me) technologies at which I've thrown myself, and managed to use productively. Among them are XML, Xerces, XSLT, XPath, Xalan, SVG, Batik, Xindice, JUnit, Log4J, Ant, BML, BSF, BeanShell, NetBeans, and a slew of others my poor sore brain won't produce at the moment. I've plucked each of these from the ether and carefully tied them all together into a loosely confederated beast that, surprisingly, works as I'd intended thus far. My biggest regret during this voyage is that this beast will likely never roam beyond the fences of my employer's realm. But, I still have the new experience with building the beast. And besides, you'd probably all laugh at the poor thing anyway since its the product of wobbling new steps. Though, actually, that would be a good thing for my company and myself since some of you might tell me why you're laughing. But, no matter - that's a political fight I'm not yet prepared to enter here at work, so maybe I can think of other things with which to tinker, out here in the open. The Java support in OS X makes me very happy, as does having Java Web Start come installed out of the box. And so, today I was also happy when I saw ThinRSS and managed to launch it with a click. I'd been musing about maybe doing a news aggregator in Java, but AmphetaDesk works fine and I've got some hack time invested in the fun outliner skin. But this really makes me want to do something with Thinlet. Possibilities abound. I've also started toying with an XPath-to-RSS scraper to replace my ?RssDistiller usage. I know that there's a Python-based one out there, xpath2rss, but I've yet to quite figure it out and I feel like tinkering on my own anyway. So far I've pulled in JTidy, Xalan, Xerces, and JUnit to help. Amazingly, to me, everything seems to be working pretty well. Anyway, back to the day's work, but I figured I'd ramble a bit and share my recent fun since I've been largely quiet around here. Hopefully I'll be back soon with updates to the AmphetaDesk outline skin, and with a new toy scraper project. [ ... 451 words ... ]
-
2002 August 09
-
Referrers and the Fish
When I see links in my referrers like this from Niutopia, it makes me wish I had translator microbes instead of the Babelfish. But, at least I have the Fish, and for free! (Thanks ?AltaVista!) The translation is sketchy and random at best, but it's decent enough for me to get the gist of things. Now if only I could respond to things in a foriegn language and not accidentally start a war or insult someone's relations, or at best sound like a moron. It is nice, though, to be reminded that English is not the only language of the web. [ ... 269 words ... ]
-
2002 August 08
-
Summer's almost over
So are any of you guys getting itchy? I've had a great Summer all around, but without much hacking or progress on little projects. And I've noticed that there's been relative quiet and little hacking amongst my reading list's authors. Could be that I'm just not looking in the right places - I did lose a few subscriptions in my bounce from Radio to Ampheta for daily use. But I just don't see as many daily innovations as I did toward the end of Winter and beginning of Spring. As for me, as the nights get longer and things get colder, I'm starting to feel an itch to start tinkering again. Maybe it has to do with not-yet-dead back-to-school reflexes - summer vacation's over, and its time to get back to learning and experimenting. Starting to look again at my neglected wiki, and of course I've started tinkering with my AmphetaDesk outliner skin again. I've also been thinking about my other little projects, referrer links, whitelist-based spam filtering, XML-RPC pipelines, and other things I've not touched in months. The fact that I'd not felt like playing with these toys for so long worried me a bit, so the return of the itch is reassuring. So, I should get back to playing soon. And, now that my financial situation has greatly improved, I also need to figure out where to start and to whom I should talk on really heading back to school. So Summer has been just a nice, long nap in a hammock. Time to start making the coffee. (Not decaf, of course.) [ ... 316 words ... ]
-
2002 August 07
-
Ampheta + Windows + Outlines: Once More with Feeling
At last: I had a good solid hacking session with the beast under WinXP, and I think I've flushed out the showstopper bugs. So, here's yet another release (hopefully zaro boogs) of the thing for you to grab and try:AmphetaOutlines-20020806-2.tar.gz Thanks for bearing with me out there and not beating me about the head and shoulders. :) I mean, yes, it's free and its experimental... but it's still damn frustrating. Basically, the trouble started when I tried to use "use lib". Seems like a reasonable enough request, lib.pm being very much a core part of perl. But, Morbus hadn't used it anywhere in AmphetaDesk, so Perl2EXE cheerfully left it out. Well, I had to hack by other means, since I wanted this thing to be as drop-in compatible with the current release of AmphetaDesk as possible. And then, there were other bits I'd left out. And also there were the bits that I'd written at 3:30 am the other night and wasn't precisely sure what made it possible for them to actually work anywhere, let alone on my iBook. But, I think now I've got things under control. It makes me appreciate the OS X environment so much more. OS X is like driving a nice big, fully packed RV that handles like a station wagon down the hacking highway. On the other hand, Windows is a Ford Festiva with a hole in the floorboards and a nagging suspicion that I forgot something back at the house. Okay, so that was a bit contrived. I just wanted to say something smarmy after all this grumbling. :) Is this thing working for anyone yet?! Update: In case you were wondering, this release should be usable via the original instructions back here. Which are, basically:Replace templates/default with this archive's contents.Create a directory named channels_meta in your data directory.Share and Enjoy [ ... 1231 words ... ]
-
2002 August 06
-
Windows + AmphetaDesk + Outlines + Metadata
That'll teach me not to test on other platforms before I release :) It appears that Windows users are missing 'lib.pm' in their bundle of AmphetaDesk joy, which really perplexed me at first. So, if you've already downloaded my tarball from yesterday, download the following template and replace templates/default/index.html:ampheta-index-20020806.html If you've yet to grab my initial release, grab this instead:AmphetaOutlines-20020806.tar.gz This all seems to have worked on my machine at home, but that Windows machine also has a full Perl installation. I'm going to try everything out on my Virtual PC WinXP install here at work, but I seem to be having trouble downloading a copy of AmphetaDesk from ?SourceForge. Let me know if you feel the love yet. :) [ ... 262 words ... ]
-
2002 August 05
-
AmphetaDesk and Outlines: Together again
Okay, after several comments, emails, IMs, and other encouragement, I stopped adding features and cleaned up a few things in my new AmphetaDesk template and have uploaded a tarball of it:AmphetaOutlines-20020806-2.tar.gz Again, some highlights:Leaner template code on the server side, leaner HTML on the browser side.Ampheta remembers things about items now, such as # of clicks, age, and # of times seen.Ampheta can act on this memory, sorting and hiding items.Old, stale things tend to go away.Often used & visited things tend to come first. To install this thing:Rename/backup the directory default in the templates directory of your AmphetaDesk installation.Replace that directory with the contents of the above-linked tarball.Create a directory named channels_meta in the data directory of your AmphetaDesk installation.Start AmphetaDesk, and let simmer for a few days to see things start workingThe thing's still a bit rough around the corners. To change thresholds and other settings, you'll need to edit default/index.html directly for now. I plan to add some fields to the settings page to manage these settings, eventually. Also, you really want to have the Digest::MD5 perl module installed, but I have included Digest::Perl::MD5 in the template. This means that things will work, but could be much faster. As for the outlines, give thanks to Marc Barrot for his activeRenderer package from which I borrowed nearly all of my outline code. (The only ?JavaScript that's mine is the expand/collapse all logic.) Okay, I'm off to lunch. Let me know how you like it! Update: It appears that some people are having problems with their AmphetaDesk failing to find the custom modules I include with the template. If you have trouble running the template, try copying the contents of templates/default/lib into the lib/ directory of your AmphetaDesk install. Update 2: I think I've fixed the showstoppers, as I will write here in the future. So, I've taken the liberty of going back in time and replacing the tarball here with the fixed one in case anyone has linked to this story. Enjoy! [ ... 490 words ... ]
-
2002 August 02
-
Well intentioned links drive the herd to stampede
So, I was happy to discover I'd gotten a link from Doc. But then, I was not not so happy to discover it was because I'm contributing to the downfall of his website due to popularity. Gah! The link, she is a double edged sword, and of course I only meant to point, not to prick or stab. This makes me recall blog posts to the Open Content Network I'd seen awhile back. Sounds like a little bit of the HTTP Extensions for a Content-Addressable Web and Content Mirror Advertisement Specification voodoo magic would hit the spot here. [ ... 176 words ... ]