Launchpad 1.0 Beta Released

Posted Apr 2, 2007 21:30 UTC (Mon) by jwb (guest, #15467) [Link]

That isn't really my point. Specifying your base font size as 80% is simply antisocial. If the browser's default fonts are assumed to be 20% too large, then the browser should be shipped with a different set of default fonts.

The reality is someone designed this page on Windows, where all the default fonts really are too large. Everybody hates the Times New Roman 12, which is why just about every single web designer in the Windows world starts out thinking "These fonts need to be smaller, and sans-serif!"

Posted Apr 2, 2007 21:35 UTC (Mon) by maney (subscriber, #12630) [Link]

I've always wanted to ask someone who espouses this sort of theory why the monitor size and/or dot pitch has anything whatsoever to do with the brain damaged practice of setting the page's main text size - the size of text that is presumably intended to be read, not the disclaimer or other stuff that's not - to be smaller than the size the user has chosen for readable text? What, are you thinking?

---

In high-resolution print typography, designers enjoy considerable freedom and control over the articulation of this [font size] range. In low-resolution screen typography, designers don't. -- Todd Fahrner

Posted Apr 2, 2007 21:45 UTC (Mon) by ringerc (subscriber, #3071) [Link]

P.S. Don't get me wrong - fonts look too small here on my 1200x1920 95dpi LCD and lines are too long, but standard Internet's monitor today is 1024x768, 85dpi (50% or close to it) so... If you are using old 72dpi or new 120dpi monitor - you should know where the knobs are...

This website overrides the most important knob - my browser's font size setting. I set my font size for a reason - so why does it want to give me only 80% of it for general reading text, and even smaller for most of the navigation and the rest of the site?

The depressing thing is how much better the ubuntu site is than average (excessive use of px units but still less than average; actually uses em for some padding measures etc; avoids hopelessly broken `pt' unit; nav menus etc can cope with at least a small amount of scaling), given how bad it still is.

Having to force scaling of the site manually each time I visit it is absolutely hopeless design - but unfortunately almost universal. I force a minimum font size to override the worst of web designer stupidity, but said stupidity prevents it from being a comfortable reading size because of their love of using incredibly tiny text in layouts that break horribly if that text is rendered at a readable size instead.

If they focused on usability, accessibility, and respecting the user's preferences over pixel-perfect rounded-corner graphics crud, maybe the website would be easier to use for people with good hardware and/or bad vision. lwn.net is one of the depressingly few sites I don't have to hit CTRL-+ to read comfortably even AFTER my min font size setting and proper display resolution setting.

Lots of people at work just can't read many websites because of problems like this. They are AMAZED when I show them page scaling, because they can now get around the web without leaning in towards their display and squinting. These are people on what you describe as "standard" 1024x768 displays or 1280x1024 panels, and most are middle aged with average vision for that age group. Is this a web design success?

What about the girl we had at work until last year, who could read type 3cm high comfortably, but needed to squint for anything smaller, or resort to a slow and clumsy screen magnifier? I landed up giving her a Linux box so she could use a user interface that didn't explode into overlapping unusable goo at those font sizes. Web sites were still a constant pain because of designers' love of specifying frame sizes in px and not scaling the site layout to the user's font sizes (which they override anyway). I fail to see how this is in any way acceptable when modern HTML+CSS provides designers with tools to make wonderfully accessible, user friendly sites that look good while respecting user's settings. They clearly just want that little extra bit of bling more than a usable site, and that drives me insane.

Designers need to be forced to actually use their product, and to do so on a variety of systems. Including (*shudder*) 56k connections.

Posted Apr 3, 2007 11:47 UTC (Tue) by endecotp (guest, #36428) [Link]

> standard resolution (1024x768)

It's a long time since I've used a 1024x768 monitor. (Maybe some laptops are still this small?)

But curiously, it seems that a lot of web users do have such small screens - or at least they have such small browser windows. A couple of years ago I was having to make decisions about page layout for a rather complex web site and I installed a bit of javascript (yes, "spyware"!) on the beta that would tell me the inner dimensions of the user's browser window. The results were astonishing: the most common size was 1000 pixels wide by about 550 pixels high! The height in particular was vastly less than I was using for development, and it caused me to re-asses the whole design. But why were the windows so small? I had two theories:

1. Users' browsers have become lumbered with "toolbars" that waste space that could be used for content.

2. Users have set their screen resolution to less than the native resolution of their screen in order to make text more readable.

So there could even be some sort of "arms race" here:

- Web designers use small fonts so that they can fit everything that they want onto the screen. (Yes, I also find very many web sites use too-small fonts. Worst example I've seen recently: try viewing dpreview.com zoomed in.)

- Users reduce their screen resolution to make the pages legible. (On windows with IE, right-clicking on the desktop and reducing the resolution is the only practical way to make pages bigger.)

- Web designers observe that the "standard" browser size is only 1024x768 and make their fonts even smaller.

- etc.

Clearly we ought to be using the default font size for body text. The main problem that I find with this is that there will always be a few elements that have pixel sizes, such as borders and graphics, and standard CSS has no way to say "width: 20em minus 2px".

Posted Apr 3, 2007 13:51 UTC (Tue) by ofeeley (subscriber, #36105) [Link]

"Not sure about why 11 words are optimal but I see 10-12 words in line here on main page. "

I think it's to do with "saccades": the eyes jump from fixation point to fixation point instead of progressing smoothly. There's a good article about it at Microsoft here: http://www.microsoft.com/typography/ctfonts/WordRecogniti...

11 words per line seems to be an experimentally determined optimum.

Posted Apr 3, 2007 11:29 UTC (Tue) by akumria (subscriber, #7773) [Link]

For those not in the know, a `pt' is a real physical measurement unit like `cm', `mm' or `in'. It's defined as 1/72 in (and nobody pipe up about the cicero, please!).

As far as I know 'cm' is centimetres, which is 1/100th of a metre. 'mm' is defined as 1/1000th of a metres. 'in' is an inch which is defined to be 25.4mm.

I don't really understand what you mean when you say that 'pt' is a "real physical measurement" and that the others are not. Perhaps you can elaborate?

Posted Apr 3, 2007 18:26 UTC (Tue) by oak (guest, #2786) [Link]

Loan them a N800 for a few weeks and ask them to preview their sites with its Opera Browser. The device has 800x480 screen at 226 DPI, see http://en.wikipedia.org/wiki/N800

(With my old eyes I really hate viewing "designer" sites on it, but I love LWN! :-))

Posted Apr 3, 2007 21:36 UTC (Tue) by k8to (subscriber, #15413) [Link]

Canonical has made multiple choices for expedience at the cost of freedom, even when they weren't really necessary.

A good example is their forum software. Now, I really do not like forums as a discussion medium, but Ubuntu has chosen the medium probably because of its broad acceptance across the users of the internet. This is the centerpiece around which its user community is expected to coalesce. They selected a completely proprietary web forum solution, when comparable free software solutions (multiple!) are already in healthy development. The rationale was, apparently, that "this one had some features we wanted". Of course they could have taken on the development work to mould a free software tool to fit their needs, thus taking charge of their own infrastructure, but I suppose the advantages of free software only apply to Ubuntu users, not Canonical?

Of course this is an aside, and does not really say whether or not the desire to free Launchpad is sincere, or taken seriously (two seperate issues). But at least it is public record that on the issue of software freedom the organization does not shrink from inconsistency, which is not what I would really hope for.

Essentially my perception is that Canonical has observably chosen against free software even when it was perfectly possible for a free software solution to be *better for Canonical* in the long run. So I do not feel this is really a position of innocent until proven guilty. I would suggest the better idiom would be that of "holding feet to a fire".

Posted Apr 4, 2007 2:47 UTC (Wed) by njs (guest, #40338) [Link]

I don't have the proper cites for this handy, but I'm pretty sure that's not the reason. They're the ones writing this software (so I'm very dubious that there are licensing issues), they have all the infrastructure for running open community projects (in fact, this is part of it :-)), they do run such projects in many many other cases, and they're basically using the same infrastructure internally as they would be externally anyway... it'd be pretty much just a matter of flipping the switch to make it public. And they could trivially find dozens of people happy to help them flip that switch.

OTOH, keep in mind that Launchpad's goal is to become a universal hub of free software development, that crosses projects and distros -- and is, incidentally, controlled by Canonical. If it becomes widely adopted, then releasing the software running it won't interfere with this vision -- the whole point is one-stop-shopping for bugs etc., so network effects will keep everyone on the well-established system. Before it becomes widely adopted, though, having just anyone setting up competing servers would be a serious threat to their gaining that kind of inertia.

Is this Evil? Not with a capital E -- the idea is that Launchpad will be so useful _because_ it's centralized, and people will use it because this makes their lives easier. Making free software development more efficient is by itself a laudable goal. Nor can they get total lock-in in any case; if they start doing really nasty things, that _will_ cause the community to turn on them, even if doing so is inconvenient. But... the whole plan does still makes me rather nervous. If nothing else, I remember the pain of trying to get a project off Sourceforge. (And last I heard, project export was a promised but "low priority" feature for Google's hosting service. Sigh.)

Again, I don't think this is deep-dark-conspiracy, more like information that the Canonical folks try not to talk about too much, but will if pressed. But I've not been following closely, so if anyone has any corrections or corroborating evidence, I'd be curious to see it.

Posted Apr 6, 2007 22:47 UTC (Fri) by gerv (subscriber, #3376) [Link]

It took Mozilla 5 years to finish properly to license all files.

But it was free software the entire time.