era's Feed

YAML Ain't Markup Language

era — Mon, 19 Jun 2006 01:25:00 -0400

Human-readable data serialization language for Perl, Python, Ruby, Java, etc. Tagged by era under 03a, algorithm, data, format, module, perl, programming, rubric, 20060619-0123,

X11 Compound Text Encoding (ctext)

era — Mon, 19 Jun 2006 01:25:00 -0400

Compound Text is a format for multiple character set data, such as multi-lingual text Images missing, get the PDF from http://www.x.org/X11R6.8.1/docs/CTEXT/ctext.pdf or the sources from http://cvs.freedesktop.org/xorg/xc/doc/specs/CTEXT/ctext.tbl.ms?rev=HEAD&view=markup Tagged by era under 03a, character, data, reference, standard, unicode, x11, 20060619-0123,

World timezones map with current time- 24 hour format

era — Mon, 19 Jun 2006 01:25:00 -0400

Tagged by era under 03a, internationalization, locale, site, time, tool, www, 20060619-0123,

Wikipedia Class Action :: Lawsuit [www.wikipediaclassaction.org]

era — Mon, 19 Jun 2006 01:25:00 -0400

Most moronic of 2005? Very late in the year, a candidate appears on the scene and challenges all the other moronic events of the year. Go on, visit the page; it's too silly to quote here. The fact that their form letter is only available in M$ Void format speaks volumes. And I would imagine that their use of the wikipedia logo is also a license violation. Tagged by era under advocacy, blog, erablog, fud, opensource, wikipedia, 20060619-0123,

Push to Create Standards for Documents - New York Times [Printer friendly]

era — Mon, 19 Jun 2006 01:24:00 -0400

30 companies, trade groups, academic institutions and professional organizations are announcing today the formation of the OpenDocument Format Alliance Tagged by era under advocacy, article, blog, erablog, opensource, 20060619-0123,

mguesser: language guesser (mnoGoSearch.org)

era — Mon, 19 Jun 2006 01:24:00 -0400

C reimplementation of TextCat, open source More solid language models than TextCat, and they use a similar format, so you can use the mguesser models with TextCat and vice versa. It's written in C, so it's faster, too. The web page is hideous, but the tool is good. This is available as a Debian package as well. See also Tagged by era under 02a, download, language, language.identification, opensource, tool, 20060619-0123,

Comparing Linux/UNIX Binary Package Formats

era — Mon, 19 Jun 2006 01:24:00 -0400

Joey Hess (of Debian fame) compares .deb, .rpm, .tgz, .slp, .pkg Tagged by era under advocacy, debian, linux, packaging, reference, unix, 20060619-0123,

http://www.iki.fi/~era/tmp/entries.yml.bz2

era — Mon, 19 Jun 2006 01:24:00 -0400

delirioussitedump back up again (for the time being) The delirioussitedump is back! The dump was performed on Sunday, up through entry #49242. I'll probably start running a daily update next time I have a few minutes to work on this, but for now, that's what's there, hand updated only. Yes, it's officially in YAML now, and it's bz2 compressed. The compressed version is some 2 and a half megs, whereas the uncompressed file is 13 megs (!). Substitute .xml for .yml if you want my ad-hoc XML instead of my ad-hoc YAML version. The uncompressed files are also available, at least for now, although I would vastly prefer that you download the compressed version. See also One of these years I'll also publish some statistics, most popular URLs, etc. (I actually already have this information extracted, just not in a convenient format.) Tagged by era under blog, bugs, delirious, delirioussitedump, erablog, 20060619-0123,

Temporary dump of de.lirio.us data

era — Mon, 19 Jun 2006 01:24:00 -0400

Just for a quick review, here is the dump I got by scraping the site the other night (and/or day) This is temporary, and will disappear. Soon. It's just for you to comment on before I do the next round (tentatively, I'd do this weekly). Please review the details of the YAML formatting in particular, if you think you know how it should look. *** This is a 770,167-byte file on a slow server [now up to 1,3 Mb]. Don't click if you're impatient. I deleted entry #247 altogether. Other than that, the data is as I got it off the server. I obtained the data by going through the /entry/nnn for each nnn up to number 2519 [now 4259], which for some reason I believe was the highest entry number at the time. Then I extracted the actual data out of the web page for each entry, and converted it to (some sort of) YAML. The /entry/ page doesn't show the modification time, so I don't have that. I have reserved a field for it in the format nevertheless. I also have this in something resembling XML, of course without any formal DTD or anything. Let me know if you think getting that would be more useful (instead, or as well). For the context impaired, the "and/or day" refers to time zone confusion. Tagged by era under blog, data, delirioussiteblog, delirioussitedump, erablog, 20060619-0123,

IBM developerWorks XML Watch: Describe open source projects with XML, Part 3

era — Mon, 19 Jun 2006 01:24:00 -0400

An introduction to DOAP with a working example Tagged by era under article, format, ontology, opensource, rdf, xml, 20060619-0123,

DOAP: Description of a Project

era — Mon, 19 Jun 2006 01:24:00 -0400

DOAP is a project to create an XML/RDF vocabulary to describe open source projects Tagged by era under format, ontology, opensource, rdf, xml, 20060619-0123,

Reflections on Thunderbird 1.5.0.2

era — Mon, 19 Jun 2006 01:23:00 -0400

I broke down and installed a "modern", "graphical" email client Much as I like Gnus in many respects (but not all ... but that's the topic for another time) I don't particularly like to use it as my secondary email client when I'm on the road with a borrowed laptop etc. I tried Evolution once, but it basically sucked rocks. (There will probably still be open bugs in the Debian BTS which I reported then, some obviously hard to repro and/or due to cosmic rays and/or bad karma.) Incidentally, it still pops up and wants to intrude in my affairs from time to time, such as when I accidentally click on something in the Gnome calendar thingy which pops up when you click on the date and time on the lower Gnome menu bar, and then you're left with a spyware process which lingers in your system and waits for a reason to pop up meeting reminders even when you have never ever put any meetings in your Gnomelution calendar. But I digress. Things are smooth sailing so far; I only have a bunch of minor warts and wishlist items to put here. * When you create a new account, why the heck isn't there an SSL checkbox in the account wizard? (I think I saw the same thing when I had to touch Outlook [2000?] with a ten-foot pole the other day.) You have to go back after the account is created and open the Account Properties to click SSL, even when you know in advance that you are going to need SSL (or, worse, that you are going to want to use SSL, but you are going to forget to go and change it, and so you will transmit your password in the clear until you go and fix it). * Similarly, when I set up a second IMAP account, it forces me to use the same outgoing SMTP server as for the primary account. The wizard even explains that I can change it later, but this is how it is going to be for now. Why? * So here I go into "Account settings" immediately after I specify the second account, and want to change the outgoing SMTP server, and what are my options? "Default server", that's it! I'm not blind so I did notice that there is a place where I can go and configure additional outgoing SMTP servers, so that I can come back and change to the one I want, but this is getting very very tiresome to write about. Why can't I just add a new server there and then if that's obviously the task I am trying to complete? * In "Account Settings", it's possible to tick the "Attach this signature:" box without specifying which signature file to use. I wonder what happens when I leave it empty? (Guesstimated outcomes, with guesstimated probabilities: Nothing, 80%; spectacular buffer underflow security CVE crash boom bang, 20%? Maybe I'll try it out once I have all the things configured that I want to configure.) * Oh, and WIBNI one could just type in a short .signature into a text box, rather than select a text file to attach to messages? What if the file disappears? Anyway, isn't it kind of crude to expose the file system etc. in this particular setting, when basically everywhere else, Thunderbird is configured from its own interface? (Also, having people create free-form text files sounds to me like encouraging them to make the .signature bigger than it ought to be. I don't expect there would be a condescending lecture about McQ signatures if your put one which is too big?) * Why, oh why does Thunderbird repeat the mistake of every earlier IMAP client to default to some particular naming scheme for Trash, Sent, and Drafts? Whatever you default to, it will likely be wrong for the majority of existing IMAP accounts. It would be an improvement if it would look for an existing structure of "Trash" or "Deleted" or "Deleted Items" or "Deleted items", and "Sent" or "Sent Items" or "Sent items" or "SENT" or "Outbox". The consensus for "Drafts" I think is "Drafts" everywhere. (Of course, most accounts probably contain several of the "de facto standard variations", as people have switched between clients with different "de facto standard variations" hard-coded into them. I imagine some twit might have decided to localize these names in some locales, too, although I haven't bumped into that anywhere personally; but then I tend to stay away from localized software, for reasons which might probably seem ironic to some of you out there in the audience.) * The security defaults make sense for the most part, but return receipts are enabled. This, to me, is rather iffy. On one hand, nobody will ever want to enable return receipts if it's not the default, and they spent all this time implementing this hugely useful and repeatedly requested (not) feature. On the other hand, it's lame, and a security/privacy problem. "But Outlook has it." Maybe the initial set-up wizard could offer a choice of security profiles? "Lame newbie, turn _everything_ off", "Lame newbie who wants to explore what might happen if you turn all the security hazards on", and "Useful default for everyone else" maybe? This could be extended to other profile types, too -- like, you could put "Lame newbie who wants the transition from Cthulhu Outlook to be as smooth as possible" as one choice, for example, and "1337 mutt user who will not use a mouse voluntarily, or send anything in the clear" as another profile. (Gnus is arguably too esoteric, but I would have no problem identifying myself as more like "1337 mutt user" than "lame Outlook victim".) * Ick, why do the "Composition & Addressing" options default to "Compose messages in HTML format"!? (I guess it makes sense historically, but, erm, no.) * At least they have the sense to default to "reply below quoted stuff". But it would be more usable if the label wasn't so long, IMHO; here's what it looks like: Composition [X] _Compose messages in HTML format [X] Automatically _quote the original message when replying Then, [start my reply below the quote [v]] (greyed out) and _place my signature [below the quote (recommended) [v]] I thought dialog boxes were supposed to be terse and informative rather than grammatically correct. Those who have a problem will hover over the text to discover if there's a tooltip there ... which will inevitably say something like "click this checkbox to check or uncheck this checkbox" anyway. * Well, luckily (or not?), there are no tooltips for the account preferences dialog. No help either, as far as I can tell. Oh, right, there's the Help menu item in the main window which opens up the support web site. I guess that's good enough, although it's not context-sensitive at all. * When selecting folders to subscribe to, why on Earth is the hierarchy collapsed by default? I also could not immediately come up with a keyboard shortcut for expanding a collapsed folder, but it just turned out to be too obvious -- the left/right arrow keys _without_ any control or alt modifiers do this. Also, incidentally, shift-arrow down the list and press space to toggle all (alas, not subscribe all, but since basically everything is unsubscribed except the inbox, that's close enough for the first-time user. I expect I will swear over this later, though. "Subscribe all" button, pretty please?) Shift-selecting a bunch of collapsed folders and pressing right arrow will expand -- the last one only. ... And, gag!, when I go back and revisit that dialog, it's all collapsed _again_! * Gnus has a useful "browse foreign server" which allows you to look in folders etc. on a server without changing their subscription status. I occasionally want to look in, or file a message into, a folder which I don't want to be subscribed to -- definitely a wishlist item for me. There's the option to uncheck "View only subscribed folders" for a particular server but ... it's somehow just not the same thing. Does the Quick File extension allow me to file into folders which are not currently subscribed, by the way? * Gnus allows me to list the folders in any order I find convenient, including "priority" and "any damn order I put them in". So far, I have only been able to list folders alphabetically. Haven't tried too hard, though. Phew, finally I'm at the point where I can actually try to read some email. * When I click on a message, I most likely will want to move it to the spam folder, not preview it! I couldn't even find a place where I could modify the details of message selecting and viewing, let alone key bindings etc. I imagine it's all possible via some super secret undocumented about:discordianism settings pane. * I can't help but continue to think that it's awkward that "mark message read" is a toogle. Shouldn't the menu entry's text at least reflect this, so that it changes to "unread" when you're operating on a message which is marked as read? * Right-clicking on a message entry in the inbox brings up a context menu ... but it doesn't contain "view message source", which is more or less the only thing I would like to see there. * Inexplicably, "View message source" brings up the "You are about to view stuff over a secure connection" dialog box. I already checked "Don't show this to me ever again, WOULD YOU MIND" so I don't recall the exact text, but you know the one, it's what you get the first time you view an https URL in Firebird too. Can't say whether this was specific to my using an SSL-protected connection or not, but what business does this have popping up for "view message source"? I already viewed the rendered message, ferchrissake, so it's moot at this point to throw up this (oh so necessary) warning (oh, is it a "notification"? It looks like a warning). * I can open several folders in separate windows, but I can't move a message between folders by dragging from one window to another. Why is that? * I really really like the Fastmail.fm web interface in that it will show a preview of the first few lines (or first line, or none at all, for unread messages or all messages, respectively) -- this helps immensely in weeding out the non-spam messages from my spam-infested folder. (I know, I should use more aggressive filters, but I have professional reasons for actually wanting to see what's coming into my account before discarding it. Anyway, I'm having high hopes for the bayesian filter in T'bird.) * Hey, nice, the junk filtering is working fabulously already! 211 spam messages and no ham (so how does it know when something is ham?) and it's already marking my incoming messages as junk. ... Update: In fact it doesn't necessarily know when something is ham, but it goes over old messages (when I open the folder they're in?) and marks them too, so I get to tell it which ones are false positives, so it should be learning pretty quickly. * I had written this here: "What's more, it doesn't display the little notification alert about a new message when the new message is marked as junk. That's more polished than I expected." But, alas, in fact it does beep at me about incoming junk mail. I guess I had the inbox focused when the previous ones came in, and that's why it didn't beep. * The hierarchy of the Junk Mail Controls dialog box is weird. I don't want to have messages automatically put into the junk folder because, like I said, I like to see what's getting all the way into my inbox, and besides, I want to scan for false positives. But I have to select this option to get to say which folder the messages should be moved to when I do request them to be moved. Then in a completely separate setting, I can only say whether I want messages manually marked as junk to be moved into "the" junk folder when I mark them. So which junk folder is this now? The default one which I just tried to override, or the one I have specified above, although I have now unchecked that part of the "handling" options? ... Turns out it's the latter, although the processing takes long enough that I had already written a rant about "none of the above" here. But it works as expected, even when I select messages in the inbox which are already marked as junk and mark them again. * So maybe there should be a clearer indication about what is happening. All I see is a progress bar which apparently shows the per-message progress, so it zips like crazy from 0 to 100 over and over, with no indication (that I can see) what sort of progress it is actually trying to report to me. Right now, it's not even doing that. The two junk messages I pressed j on are still in my inbox, they just changed from unread to read (undesired side effect!) and there is no progress bar of any sort. ... Actually, I can't repro this any longer. The junk messages stay in my inbox and I have to drag them over to the spam folder manually. (For the record, the "Junk" folder name was already reserved by me for other stuff; the spam goes into Junk.spam and there are several other subfolders too.) Perhaps the "Automagically move messages to junk folder" was still checked in Thunderbird's little mind while I was mucking with that dialog, or something. * Why is the message composition window so small? It's wider than 80 characters, but it feels kind of small, and I would definitely like it to be taller. The size and location of the new message window seems to be hard-coded or something (but at least flexible enough that a second new window only partially overlaps with the one I opened before). * No custom X-headers, boo. * Signature is in the composition window where you can edit or trim it -- good. * Impressive: Options > Security > Encrypt this message, right there in the regular newbie interface. Granted, you need to set up keys etc. in order to actually use it, but I like the gesture. (Maybe it could be in a less weird place, though.) * Options > Quote message is really weird too. Apparently, I can quote an arbitrary message by selecting it in the folder and then switching to the composition window and selecting this menu item. I guess occasionally that's useful. It's still really weird. Which message? When I'm writing a new message, not responding to anything in particular? Just some message which happened to be selected when I chose "New Message", it seems. At least it would obviously belong in the Edit menu; maybe it would feel less weird already then. * I've always wondered why most clients make forwarding type a global or per-server preference. I would like to be able to select when I forward a message how exactly I want to do it (attach MIME message/rfc822, inline text/plain, inline plain old text, quoted, or what, maybe uuencode?), or at least given the opportunity to override the default. * Ouch, looks like it EHLOs as [172.16.0.4] (sic; that's the dynamic, private IP address I have on my internal network right now) when sending out a message. Not very un-spammy, is it? I would consider it to be a de facto standard to take the computer name from "My Computer" on Windows, and the obvious uname -n on *nix. * Furthermore, the Message-Id is generated based on my From: address, which also looks kind of spammy to me. Oh, maybe I'm overreacting. ... And there's that User-Agent: thingy, and that format=flowed thingy, and that charset=ISO-8859-15 but Content-Transfer-Encoding: 7bit thingy, and that MIME-Version: 1.0 being in completely a different place than the Content-Xxx headers, but ... I guess I'm getting too old. At least it has proper References: and In-Reply-To: when responding to a message (unlike, I discovered to my surprise, and that was a surprise in itself, because I should be pretty jaded by now ... unlike, I was saying, Microsoft Outlook, which completely hasn't ever heard about these headers, which have only existed since probably before the original RFC822 in 1982. Instead, Microsoft had to invent their own Thread-Index: and Thread-Topic: [identical to Subject: for all practical purposes]. Aargh, don't make me write about Outlook). Disclaimer: If you are anybody who is somebody, you are probably not at all like me when it comes to email. Heck, I would probably still be using VM, despite all its quirks, if it could cope with IMAP in a useful way. Disclaimer too: Hanging my head in shame, I confess that this was the Windows version I installed, on something like Windows XP SP2. Quite frankly, I would have expected the Linux version to crash or act erratically at least once during these initial tests. Oh well. Update / reference: I posted a teaser to the Mozillazine T'bird forums: http://forums.mozillazine.org/viewtopic.php?p=2248547#2248547 Update: I installed the Swedish spelling checker dictionary from http://dictionaries.mozdev.org/installation.html -- they don't seem to have Finnish, although e.g. Hungarian is available (technically, I would believe they are roughly as challenging to support with standard run of the mill 1970s spelling checking technologies, i.e. Hungarian must suck too). Anyway, it seems they have simply allowed unlimited synthetic compounds, meaning your average misspelling is just as likely to go unnoticed as most correctly spelled words. So "gambla" and "rambla" are not flagged (we speculate that they are synthesized from the proper words "gam" and "ram", and the onomatopoetic "bla"), nor is "tiodostonhallinta" (my old favorite "tiedostonhallinta" is flagged, but that's because "tie" -- informal for "tio" [ten] -- is not in the dictionary; this is Finnish for "file management", but it also oh so conveniently can be synthesized from 'ten' + 'dose' + 'hall' + 'occupy' [which in turn is formed from 'in' and 'take']). Oh well, it might still be marginally better than no spell checker (which is all we expect from a spell checker anyway, so). Pardon the digression, it just happens that I find this hilarious. More test cases: "terÃ¥ratack" - unflagged (thanks to Lena for that one) - syntesized from "ter" (what's that??) + "Ã¥ra" + "tack"; "gÃ¥rilla" - unflagged - synthesized from "gÃ¥r" + "illa"; "applikationsprotokollis" - flagged; "moapplikationsprotokoll" - flagged; "apapplikationsprotokoll" - flagged; "applikationsprotokollte" - flagged. Hey, maybe it's better than I thought. Tagged by era under blog, download, erablog, imap, mail, mozilla, _notes, opensource, software, testing, 20060619-0123,

Bug: UTF8 does not display correctly

era — Mon, 19 Jun 2006 01:23:00 -0400

If you require input to be UTF8, you should also serve it back as UTF8 The referenced entry contains a bit of background. Anyway, I was trying to create an entry with a body which contained the letter lowercase a with a circle on top ("Angstrom", "Swedish o") and the post was getting rejected. I finally managed to get it accepted by typing in the raw Unicode for this character, which displays as ÃÂ¥ in Latin-1 (that's uppercase A with a tilde accent, and a Yen currency symbol). But now that I view the page in Latin-1, that's still how it displays. (And the UTF-8 in this entry displays as roughly AfAY, i.e. the same with a "franc" currency sign and an uppercase A with a circonflex accent in between the two others.) The server should be configured to serve the pages as UTF-8, and/or the served pages should have a http-equiv header identifying the character set of the content (but this is not widely supported -- I believe it's disabled in Apache by default, and on the client side at least Mozilla didn't use to obey it). Update: It seems that this is merely a bug in my browser. It appears to ignore the header which is sent correctly from the server as it should. HTTP/1.1 200 OK^M Date: Mon, 18 Apr 2005 12:09:09 GMT^M Server: Apache/2.0.51 (Fedora)^M Set-Cookie: CGISESSID=1dae828430a5d71938409c99ddf93897; path=/; expires=Mon, 25-Apr-2005 12:09:09 GMT^M Connection: close^M Content-Type: text/html; charset="utf-8"^M ^M de.lirio.us :: : entries filed under { delirioussiteblog } Tagged by era under blog, deliriousbugs.worksforme, erablog, rubric_0.09, 20060619-0123,

Bug: "recent tags" listing is weird

era — Mon, 19 Jun 2006 01:23:00 -0400

The tags I see in "recent tags" on the front page are quite different from the actual content Here are the "recent tags" I get on the front page right now: 9 01a 7 blog We pause the listing here to note that the 01a links are all my own, and I haven't added anything to that category today. Actually the newest entry is from five days ago. On the whole, the "recent tags" listing looks eerily like my own personal tags, but I don't use "news", "games", "books_to_read", "wordpress", "themes", "portal", "mac", "freeware", "computers", "web", "lithuania", or "law". The rest of the listing is below. In the meantime, let me note that just now, the most frequent tags I see on the actual front page are linux, EDA (3 occurrences), software, ruimte, howto, finance, china, Wildlife, PotentialPurchases, Linux, KDE, Cams, and Birds (2 occurrences). Okay, here is the rest of the "recent tags" listing, in somewhat condensed format (all tags with the same occurrence count on the same line). 6 wiki software news games category books_to_read advocacy 00a 5 unix site programming history graphics emacs design deepsite 4 wordpress themes opensource extension erablog download development algorithm 3 portal mac linux hacker freeware delirioussiteblog deliriousbugs delirious debian css computing computers bugs 02a 2 www web society search sawfish music mail lithuania law internet Update: This bug still seems to be present in Rubric 0.09_01. Update: This looks like it's finally fixed in 0.10. Tagged by era under blog, bugs, deliriousbugs.fixed, delirioussiteblog, erablog, _fixed, rubric_0.08, rubric_0.09, 20060619-0123,

jk: Hey, it's open source - I'd run my own installation if I were you

era — Mon, 19 Jun 2006 01:23:00 -0400

If you think de.lirio.us is "private", think again. You know, there's a reason why they call it "social" bookmarking. I'm concerned about my privacy, too -- anybody who isn't paranoid is insane, these days -- but acting in the open on the public internet is something you do because the benefits outweigh the disadvantages (or at least, one would hope that you have made this risk assessment). If "jk" is too personal, use a more anonymous handle. For a thing like bookmarks, being "just another demographic data point" is anonymous enough for most of us. One could even argue that being at least somewhat candid about some things you do is a political and social statement unto itself, but that's probably a topic for another time. It's good that you bring up these concerns, though; they should probably be documented somewhere so that people know what they are signing up for when they join de.lirio.us. In the meantime, anyone who wants the distributed on-line central bookmarks without the "social" part can grab a copy of Rubric and install it on a server they control, in an obscure location where nobody will ever find it and perhaps somehow password protected and SSL-encrypted (and for the real pros, apply another level of encryption on the actual data). About the practicality of blogging on de.lirio.us -- if you want to see how well or badly it works in practice, do check out the entry I linked to before (http://de.lirio.us/rubric/entry/1925) -- it's basically the entry page for my "not really my blog" entries. It could benefit from a better user interface (a couple of panes with "about this blog" and "about me" links would take it a long way towards being a proper blog, although still without pictures, or live links in the body of an entry) but the basics are there and work fairly well. In my daydreams, I might write a wrapper for it some day -- write a "portal" which pulls in a selected group of de.lirio.us entries (or other social bookmarks, or whatever, Flickr pictures maybe?) and displays them in a nice blog-like view, with the additional navigation links and some sort of rich formatting of the content. (I don't particularly care that people can't comment on my blog entries, as most blog comment pages are not worth reading anyway, in my experience. But I guess that could be fixed as well.) Tagged by era under blog, delirious, delirioussiteblog, erablog, 20060619-0123,

Re: scraping of de.lirio.us data

era — Mon, 19 Jun 2006 01:23:00 -0400

Some questions to Steve, and some brief statistics from the first pull Steve: Just a clarifying question. There is exactly one email address in the data I scraped, apart from a few occurrences of the address you have in the "Feedback, baby!" link on every de.lirio.us page. This is in entry #247, which IMHO looks rather plainly like it was supposed to be visible and public. For now, I have the data in an ad-hoc XML-ish format I just made up. I guess it would be better if I used YAML but I'm not very familiar with how to do things with YAML in practice. Any chance you could show an example of how you expect your own public format to look once you get started on that? (Update: I am now a bit more familiar with YAML, and regard my previus question as a bit silly. There are a couple of open issues but on the whole, it's not very unclear how to do it. See tag delirioussitedump.) Here are some statistics from the data, as of some 16 hours ago: Top 20 users (number is count of entries in the database): 113 comforteagle 102 nic 81 xsteadfastx 71 johnpogy 53 peakoil 49 rmutt 46 Razor 42 era (how appropriate with a Magic Number!) 41 ab9 40 umm 39 frenchtouch 36 julien 34 macmanx 33 theanomaly 31 dez 29 LibraryStuff 27 gorbag 25 Masa 25 aliko 24 krazykiwi 23 a77ila 23 ellis 23 hernanm 23 phraje Most popular tags: 217 blog 101 linux 98 news 74 software 56 delirioussiteblog 49 programming 46 music 44 design 43 opensource 40 de.lirio.us 40 web 39 blogs 36 comics 35 security 34 News 34 site 33 development 32 mac 31 search 30 daily 30 reference 29 lslinkblog 28 internet Most popular page titles: 19 Site Submission MultiTool- Alan's Marklet Maker 17 Slashdot: News for nerds, stuff that matters 17 de.lirio.us :: Social Bookmarking, Tagging, Blogging & Notes. Mmmmm, Notes. 11 Slashdot 10 Google News 6 Google 5 User Friendly the Comic Strip - The Daily Static 5 de.lirio.us :: Social Bookmarking, Tagging, Blogging 4 Gmail 4 Google Gulp 4 Jokes and humour 4 Penny Arcade! 4 Ubuntu Linux 4 del.icio.us 3 1911 - DESIGN YOUR OWN ANTI-PROCRASTINATION PLAN 3 Boing Boing: A Directory of Wonderful Things 3 Internet message headers - quick reference 3 Lessig's Code v.2 3 RealWorldTrading - Forums,Message Boards,Trading,Trading Education,Trade rs Education,Trading News,Canslim,Eminis,Futures,Stock Trading,Trading Chat 3 Scuttle: de.lirio.us 3 The Art of Unix Programming 3 The GNU Operating System 3 The Linux Game Tome 3 heise online 3 slashdot I made a frequency table of the URLs too but it's on my laptop and I'm not too eager to rewrite the scipt, even though it's trivial. Anyway, that list is not particularly surprising. Tagged by era under blog, delirious, delirioussiteblog, delirioussitedump, erablog, 20060619-0123,

Bug: Duplicate tags result in an ugly error message

era — Mon, 19 Jun 2006 01:23:00 -0400

Entering the same tag twice gets me a hideous SQL error message The code should probably just remove duplicates before passing it all on to SQL. Update: This is supposed to be fixed with the upgrade to Rubric 0.08 Here's what I get (formatting obviously elided, this is just a simple cut & paste): Software error: Error executing run mode 'post': Can't insert new Rubric::EntryTag: DBD::SQLite::st execute failed: columns entry, tag are not unique(1) at dbdimp.c line 401 [for Statement "INSERT INTO entrytags (entry, id, tag) VALUES (?, ?, ?) "] at /usr/local/lib/perl5/site_perl/5.8.6/DBIx/ContextualFetch.pm line 51. at /usr/local/lib/perl5/site_perl/5.8.6/Class/DBI/Relationship/HasMany.pm line 89 at /var/www/vhosts/Rubric-0.06/rubric line 4 For help, please send mail to the webmaster (steve@de.lirio.us), giving this error message and the time and date of the error. Tagged by era under blog, bugs, delirious, deliriousbugs.fixed, delirioussiteblog, erablog, _fixed, rubric_0.07, 20060619-0123,

Wishlist item: RSS feed for actual entries

era — Mon, 19 Jun 2006 01:23:00 -0400

An alternative RSS feed with links to my entries would be nice The live bookmarks I have now allow me to easily visit the sites I have added to my de.lirio.us links. But as often as I wish to visit those sites, I wish to revisit the actual entry on de.lirio.us, perhaps to read the description, or add a tag, or otherwise edit the entry (especially these blog-type entries need an update every once in a while!) At least in Firefox, it seems quite possible to have multiple RSS feeds for a page; you can have different titles for each feed, and the "live bookmark" button down in the right corner expands to a menu where you pick the one you want to subscribe to. In concrete terms, I would suggest adding a &type= parameter to the ?format=rss operation, where allowable values might be links (the current kind, and the default if this parameter is missing) and entries (where instead of links to the uri= of each entry, you would get links to the entries themselves). Tagged by era under blog, bugs, delirious, delirioussiteblog, deliriouswishlist, erablog, rubric_0.10, 20060619-0123,

Bug: Missing and mislabelled entries in RSS feed

era — Mon, 19 Jun 2006 01:23:00 -0400

For example, a couple of oddities in the live bookmark for delirioussiteblog Here are the top 10 items on "site blog/forum" when I view it via the regular web interface @private filter broken? jpbarto @ freeshell.org Re: Converting Del.icio.us to De.lirio.us jpbarto @ freeshell.org Wishlist item: @any private tags Wishlist item: RSS feed for actual entries Re: Converting Del.icio.us to De.lirio.us Re: Converting Del.icio.us to De.lirio.us Bug: leading space in tags field not allowed Re: Delirious site dump temporarily offline and here's what's displayed as the top 10 items in my corresponding live bookmark jpbarto @ freeshell.org jpbarto @ freeshell.org jpbarto @ freeshell.org Wishlist item: RSS feed for actual entries Re: Converting Del.icio.us to De.lirio.us jpbarto @ freeshell.org Bug: leading space in tags field not allowed Re: Delirious site dump temporarily offline de.lirio.us/popular Re: displaying image links inline Notice that two of the actual top 10 are missing, so the RSS top 10 contains entries which are number 11 and 12 on the web page. (The RSS feed has more than 10 entries, of course, but the top 10 are illustrative of the problem.) Off the cuff analysis: The entries which are not links to anywhere are simply omitted from the RSS feed. This is perhaps working according to spec, but is a bit counter-productive (in case you want the RSS feed so you can see when there's new stuff to read in the blog, not particularly because you want the Latest Links to Random Places). This would be easily fixed if the entries RSS were available When there are several links to the same URL in the RSS feed, they are all labelled identically. The labels in the RSS feed seem to correspond to either the title of the page they link to (sort of understandable, but I think technically not feasible ...?) or just the latest (or earliest?) label assigned to that particular URL (in which case this is definitely a bug). Tagged by era under blog, bugs, deliriousbugs, delirioussiteblog, erablog, rss, rubric_0.10, 20060619-0123,

Top open source lawyer blesses new terms on Microsoft's XML file format

era — Mon, 19 Jun 2006 01:23:00 -0400

Rosen declares Microsoft's XML formats kosher Tagged by era under advocacy, history, licensing, opensource, xml, 20060619-0123,