It takes a full 3 clicks to get from the wikipedia article on lolcats to the article on cheeseburger. Or from cheeseburger to lolcat. This is not right.
Archive for the ‘web’ Category
This is not right
Wednesday, December 10th, 2008Comments reenabled
Sunday, June 1st, 2008So after only a year and two months, comments are finally reenabled here. I’m now using Akismet and Bad Behavior for spam blocking. Hopefully, it will work out well enough to leave it open (I didn’t like having to skim through 200 spams a day to find the on average 0 actual comments there).
Automatic updates
Thursday, October 25th, 2007This has got to be one of the coolest things I’ve ever seen about Mozilla as a platform.
Grey Text Considered Harmful (to your eyes)
Saturday, September 8th, 2007Dear everyone,
Please do not use grey text on a white background. Use black. Grey is harder to read, causes more strain on the eyes, and makes me less likely to read what you say. Really. I recently realized that one weblog I used to read a lot I wasn’t liking nearly as much recently, not because the content was worse, but because it was actually more work to read and that made it not worth it.
So don’t do it. Please. Or I’ll just stop reading your stuff (maybe not even consciously).
And yes, I am aware that this blog uses it, by virtue of the default wordpress theme. I only just realized that, and am fixing it now.
Edit: So a wordpress update broke this for a few days. Fixed now.
Linux TP down 10%!
Wednesday, May 9th, 2007Thank you, bug 375760:
And a big thank you to roc for writing that patch, and vlad for reviewing it too!
Comments closed (hopefully temporarily)
Tuesday, April 3rd, 2007I was getting about 200 spams a day to this blog, so I’ve turned comments off, hopefully temporarily. Sometime when I have more time, I’ll get a better solution in place (I’m out of school in a month, so probably then).
Reflow Branch!
Wednesday, December 13th, 2006Woooooooooooooo Reflow Branch!
(Proof dbaron is awesome: I didn’t even notice for 6 days.)
Thank you, OED
Monday, November 13th, 2006So today I was looking through my firefox search bar’s history, and along with things like “how to raise your iq by eating gifted children” (it’s a book apparently) and “fraught with all manner of spiky-haired cute things” (no results, this page will be the first), I found ‘quercivorous nepheliad’.
And of course, I immediately recognized what I was doing when I searched for that: “Hey, it’s a googlewhack! I wonder if it still works.”
So I tried it, and it almost worked, except the passage it was written in had been quoted a bunch of times. But I also noticed that google suggested “quercivorus” instead of “quercivorous”.
Obvious choice of action, test it out and see which one gets more search results. Which I do, and see quercivorus gets 500 results, and quercivorous gets 15300. And wow! My blog is at the top of the results for quercivorous! That’s awesome! But wait… these look like mostly just word lists… the results for quercivorus looked a lot more scholarly… yeah, they do. And oh… 15100ish of the 15300 are all one site.
GAAH. I misspelled my own favorite word, and it’s at the top of google’s results for it to boot.
So I went and edited my post. Made a note about my misspelling it, and corrected the spelling for the future, in hopes of google putting me back high in the rankings for the right spelling.
But just before submitting, I thought to myself “I wish I had my OED, so I could check what it says.” Of course, I don’t, because no one brings their OED to college with them, at least, not anyone studying computer science.
But maybe the library has one. So I walk over there, and yes, it does! And then I look it up… And it says I was right all along! It really is spelled quercivorous! Yes!
A bit more research shows that there are 4 species (probably of insect) named Something quercivorus. Removing those, there are practically no results. I cancel the edit to my old post and write this one.
The moral of the story: The OED may not know everything. But in its niche, it’s smarter than even google.
Namespace syntax in HTML
Monday, October 9th, 2006- We need namespaces in HTML.
- We need a syntax for declaring those namespaces in HTML.
- A very good, fitting syntax has already been invented for use in XML, and is understood by a significant number of authors.
- That syntax has been corrupted through misuse by many pages. Is it worth trying to save, or do we need a new one (or is one of the premises wrong)?
Enabling encryption in vsftpd
Tuesday, September 12th, 2006For the past few days I’ve been struggling with enabling SSL (actually TLS) encryption in vsftpd. I followed the directions, but every time I try to connect it gives the message
Connected to xx.xx.xx.xx:21 220 (vsFTPd 2.0.4) USER anonymous 530 Anonymous sessions must use encryption. Disconnecting from site xx.xx.xx.xx
(Several hours later)
Apparently it’s not a problem with vsftpd. ftp -z ssl xx.xx.xx.xx works fine. It’s just the ftp clients I tried (gFTP and Filezilla) don’t encrypt the connection. For gFTP I found an FAQ entry that says
You must add the public key of your self signed CA to your OpenSSL certs directory.
but I’ve done that, because I had to put it there for vsftpd to use it!
Edit: Firefox apparently doesn’t have support. That’s annoying.
Do NOT attach “target” through script!
Friday, July 28th, 2006Using target in (X)HTML Strict. If you need to use the target attribute, either stick with Transitional DOCTYPEs, or add the target attributes dynamically using semantic scripting, e.g. based on semantic class names.
(from http://tantek.com/log/2006/07.html#d27t1218)
NO NO NO NO NO! Your page is just as invalid if you add target in with script as if you put it in your markup.
If you absolutely must open a new window, make it a normal <a href=""> and use window.open() and cancel the default action for the link. But if you weren’t providing any size or chrome info (which are impossible with target anyway), then you don’t have any reason to be opening a window in the first place. Just don’t do it.
A good font package approaches…
Thursday, July 27th, 2006The STIX Fonts have all been made, and just need to be packaged now! Wooo!
Back to the 1.8 branch, part 2?
Wednesday, June 28th, 2006I’m considering switching back to the 1.8 branch again. There’s a giant memory leak on trunk that it seems only I can see (yes, I’ve tried safe mode) that I’m having trouble narrowing down. Cairo is still incredibly slow on Linux. The tab scrolling thing that just landed is incredibly stupid, in my at least partly-qualified opinion, and interrupts my workflow a lot.
And to add to all that, ever since going to a music festival for 4 days and not using a computer at all, I’ve stopped really wanting to spend so much time on it in the first place, and I just want it to do what I want.
So yeah, I might switch back, to take some of the hassle away. That would also mean I’d file fewer bug reports…
Bad Behavior
Monday, June 26th, 2006I’ve just enabled Bad Behavior here. If it gives you any problems, please contact me.
Extension reqest: animate one image
Monday, June 19th, 2006I browse the web with image.animation_mode = once, because I, you know, value my sanity. However, there is the occasional image that I do want to see animate. Does anyone know of an extension that lets me temporarily animate one particular image? If not, I think a context menu item (only visible on animated images, of course) would be a good solution for anyone that feels like making one.
HTML is not an “Application Language”
Tuesday, June 13th, 2006Part of the reason is of course that HTML was originally a document language and is slowly evolving in being both that and an application language.
I’ve said it before and I’ll say it again: It shouldn’t be. HTML is fundamentally a language for displaying information. Accepting input should not be a part of HTML because a) It’s out of scope, and b) it’s useful in places that HTML isn’t. Manipulating information is also out of scope (and thankfully has never been mushed in). An application requires bits from all three: Accepting, manipulating, and displaying (it also benefits from a fourth: styling).
Back to the 1.8 branch.
Thursday, June 8th, 2006I’m no longer running firefox trunk nightlies, having moved back to the 1.8 branch. It was nice having the new features, and I wish I could keep testing them, but the performance was just too bad. I was able to stand it for a long while, but then I had to test something on 1.8, and after that I just couldn’t go back. It’s amazing how a fast browser doesn’t seem all that much faster, but a slow one feels awful.
The perf bugs (mostly on Linux, from what I gather) have been sitting there untouched for a while, as people focused on the 1.8 branch, and went to XTech, and had vacations, and the like. Vlad says there will be work on them soon, though, within “the next few weeks” (on the 30th in mozilla.dev.tech.gfx). And indeed, one small fix went in yesterday, though not enough to make me switch back.
Hopefully “next few weeks” from the 30th turns out to mean starting at the end of this week and picking up next, so I don’t have to stay away too long.
Update: So it seems that that in fact, it wasn’t just psychological, there was a huge regression that happened on trunk while I was using branch. I’ve moved back to about half and half for now, because testing trunk is important and I know there aren’t enough people that do it, but I hope the cairo bugs get fixed soon.
Web2.0 / CSS 2.0
Friday, June 2nd, 2006Ironically, Web2.0 is much like CSS 2.0. Although one was specced in advance and the other just happened, they both had/ve a lot of new things, some of which were/will be adopted, some of which were/will be unused, and some of which need/ed slight revision. Like CSS, when 2.1 comes out Web2.1 will be a more polished, mature technology. And like CSS, Web2.0 has major deficiencies that won’t be corrected until 3.0 ;)
Firefox Tip #1
Thursday, May 25th, 2006A cool thing in Firefox that I figured out not too long ago:
- Ctrl-R refreshes
- F5 refreshes
- Ctrl-Shift-R clears the entire cache and refreshes
- Ctrl-F5 clears only the cache of that page (and images, etc. on it) and refreshes.
Hope someone finds this useful.
Getting rid of unused option cruft in search queries
Friday, April 21st, 2006So I have a form with a lot of options. It’s a search form, so it’s a GET, meaning it goes into a URL and is reusable, as it should be. However, at any one time most of those options are unused. This means I get lots of &option1=&option2= cruft in the url.
So, what’s the best method of getting rid of that? Client-side removal of the extras on submission? HTTP 302 Found? HTTP 303 See Other?