Older blog entries for bwh (starting at number 20)

19 Jul 2005 (updated 21 Jul 2005 at 03:47 UTC) »
Naming inkscape

Mental blogged about back when we named Inkscape, and how eery it is seeing the name everywhere, and hearing everyone talk about it.

I don't remember all the details, but I remember brainstorming names with Mental. It was really important to me to pick a good names, because I know that for the name of a project can be key to its success or failure. A lot of people pick names that are silly, or inside jokes, or boring, meaningless acronyms, but I think these are limiting; they make your project sound too unprofessional, or too hard to remember, or too boring. We cracked the thesaurus and went through a LOT of variations of various words and combinations. Just about everything art related was already used. I remember when the name 'inkscape' came up, it was like, "Aha, that's it!" :-)

15 Jul 2005 (updated 15 Jul 2005 at 20:47 UTC) »
Inkscape in Ottawa

Next week at the Desktop Developers' Conference, I'll be giving a presentation on Inkscape. Good timing too, since we're also within a week of having the new 0.42 out. I'm hoping to be able to directly demo the new features, but using Inkscape with my laptop's touchpad is a bit cumbersome, and the mouse connector is not ps2 or usb so I'm not certain if I can get a mouse working with it. But I have a few days to figure this out so who knows...

Speaking of presentations, I was down at DreamWorks yesterday giving a presentation about NFSv4 on Linux to the film industry, and a guy from Novell (named Guy) noticed I'd been hacking on the Inkscape code in one of my xterms, and excitedly mentioned he'd just been playing around with it earlier that day. It's great to see how far Inkscape's spread through the Open Source world.

I'm also continually amazed at how well the Inkscape bug fixing process works. Before we started the release I was rather concerned about the quantity of bugs that built up since our last release. At first after starting the bug fixing phase, progress seemed slow, but I remembered from prior releases that it always takes time to build up momentum, so just waited. Holy cow! After a couple weeks it really got going, and in fact the last couple weeks has been the largest bug massacre I think we've ever had. I'm also impressed at how much teamwork there's been in the bug fixing this time through; it's common for five or six people to be involved in generating the fixes for the bug, integrating them, and validating them.

Render Test

I've started a tool to allow testing SVG renderers. Here's an Example. This allows you to take a set of .svg's and render them with a variety of svg renderers, and then generate a "diff" image to show how their SVG rendering varies. I'm hoping to extend this with more renderers, better reports, and more controls over the rendering process. I also want to run this over the W3C testsuite, and to start tracking it against the Inkscape codebase so we can see how rendering changes over time.

Inkscape and Google Summer of Code

If any students out there are interested, Inkscape's been accepted as one of the mentor projects for Google's Summer of Code. More info is at http://code.google.com.

20 May 2005 (updated 20 May 2005 at 22:43 UTC) »
Kernel Build Script

Jonathan posted a script he wrote to help automate his kernel builds today.

For the testing I do at OSDL, I also needed to automate installing new kernels on Gentoo on my client and server. The NFSv4 guys put out a new patch every week, so it's a significant amount of time to have to do it manually.

Anyway, I threw together a build_kernel bash script a few weeks back. Fairly similar to Jon's script, but I also added logic to make it append to lilo.conf or grub.conf, and to invoke mkinitrd.

Also, I fiddled around with the kernel config options, and instead of using make oldconfig, found I could automate things better by copying my config into arch/i386/configs/config.default and then using make defconfig.

Note that if you use this script, doublecheck the hd devices used in the lilo/grub stuff; /dev/hda3 is my root device, and you'll want to change that if yours is different.

One of the next options I want to add is if using grub, have it attempt to boot to the new kernel, and default back to the old one if failed.

Jon, you've got some interesting things in your script. It looks like you've put more attention into the messages, and symlinking. It'd be interesting to see if your work and mine could be combined into an overall better script. :-)

20 May 2005 (updated 20 May 2005 at 22:42 UTC) »
NFSv4

LinuxWorld Magazine (I know, I know) is running a cover story I wrote on NFSv4 testing. Next, working on a paper about "Open Source Testing for NFSv4" for PNSQC.

At the encouragement of several other NFSv4 guys, I've set up an NFSv4 Wiki, using the MediaWiki software instead of the UseModWiki code that I've used for wiki's in the past. I'm pretty impressed with MW; easy to set up and configure, the default look is a bit nicer, and it has quite a few more handy features than UMW. The one downside is that it requires a database for its back end, but even there they made it reasonably easy to set up.

DMS

Posted version 0.12 of Document-Manager to CPAN a few days ago. Still not quite to where I would consider it 'usable', but it's coming along. An editor contacted me to write an article about it for The Perl Journal for next month, so I think I got my work cut out for me here! </a>

Inkscape

We had chosen to try to get the interface converted from Gtk+ to Gtkmm for the 0.42 release, but it turned out that not too many people were available to work on it, so progress was slow. I've decided to de-scope the 0.42 development efforts to the features we've implemented so far, plus the DialogManager. I've integrated the latter into the old SPDesktop code and explained to a few people how to do Gtkmm dialogs with the new system. I figure we should give through the rest of the month for development, and start getting going on the bug fixing for the release around June 1st. I'd like to get back to a routine release cycle for Inkscape; it's been too long since our last release. :-)

NFSv4 - Today we hit a nice milestone; we've completed the prioritization of all of the test matrix. Whew; this took around 20 meetings to get through, so I'm very happy to have that in the bag now. I'm looking forward to really digging into the testing and start getting some solid progress.

I've also gotten accepted to the Pacific NW Software Quality Conference conference. I'm going to be presenting a paper on "Creating an Open Testing Process with the NFSV4 Open Source Project". The intent of my talk is to explain that Open Source style testing is a valid testing methodology, and that when combined with traditional formal testing methods, has benefits far beyond either way alone.

Open Clip Art & DMS

I've set aside half of my freetime this week to focus on DMS (the other half going to a carpentry project I need to get done). Last night I tested DMS on the freedesktop.org servers, and solved an issue that's been troubling me for a couple weeks. I can once again get client/server communication from my local box to fdo, for the first time since the fdo compromise last year.

However, I notice that the operations are slow as snot, which is a bit of a disappointment. Not a total surprise - SOAP is not certainly not known for its performance.

Luckily, I've made contact with the DoXFS developer, who has an XFS-based backend system that promises to work *much* faster. The downside is that it's a bit more involved to get installed, so I haven't had a chance to play with it. My plan is to continue with the direction I've already started down, since I've got the tasks scoped out pretty well, and then later on, investigate migrating the Document::Repository portion of my code to something DoXFS-based. But my near-term goal is to get DMS to a point where OCAL can start making use of it for managing its SVG files.

Inkscape for Maps - Nicu took me up on my idea to use Inkscape for making game maps, and posted an awesome example. Nice work Nicu!! Proves my point, that map is beautiful. :-)

GOPchop - yesterday helped Kees work on GOPchop. Created the website (derived from the original at his outflux site). Also helped him lay out a development roadmap. The nature of this project is quite different from ones I usually get involved with, in that it's already had its 1.0 release, and all the principle functionality has been coded. The work from 1.0 to 2.0 will revolve around "finishing" - debugging, documenting, etc. I think having the roadmap will help, since sometimes that kind of work can feel overwhelming if you don't have a tangible goal in sight.

WikiNews - Ted discussed a talk on Monopoly, Media, & the Right to Know discussing the over-corporatization of the news, the rise of blogging, and the cost of investigative journalism today.

I have found the intersection of news and the internet to be absolutely fascinating. I first learned about the WWW back in '93, before there were any news organizations out there. Today, only a decade later, the Internet has become a primary source of news for many people.

When my dad comes and visits, we have to drive down to the minimart to get a Sunday newspaper (he has to have his 'fix', he jokes). Meanwhile, I'm off in the computer room doing basically the samething, except with mouse clicks and photons instead of newsprint and ink... I sometimes wonder which of us is getting "better" news? I'm able to quickly filter through to exactly the stories I care about, but he probably gets broader insights. I can put my finger on *more* news, but he probably has the edge on local news. It's a tossup whether either of us is getting anything resembling truth; I know some people are concerned about newspapers becoming corporate dominated, and others worry that blogs are inadequately researched. But heck, news inaccuracy is nothing new; if I remember my U.S. History correctly, truth and the news media have been fickle bedpartners at best, and bias comes with the territory. The only realistic solution is the same one we were taught in school: "Think critically."

A while back I was watching a series about the history of communication mediums (newspapers, radio, tv, etc.) and it struck me that each medium has a principle "style" or "mode". When a new media technology comes along, it takes a while before people figure out how to use that mode. For example, when TV started it was used sort of like radio-with-pictures - advertisements were just radio jingles with some dancers for show. News programs were simply a well educated guy sitting at a desk reading the day's news stories. It took them a while to work out that the principle mode with TV is as a "window on the world", with live footage and on-location journalists. Once they did, it relegated radio to a secondary role for news reporting.

The program didn't really cover the Internet much that I recall, but if you think about it, the principle mode of the Internet is "interactivity". Anyone can be a publisher. We can (often) comment directly on stories. Through tools like CVS, Wiki, Blogs, and so on we often participate in *making* the stories.

With this view in mind, I'd conclude that the cnn.com style news sites (basically just a newspaper with hyperlinks), is a quaint attempt to adapt the principle mode of the printed world to the online world, and that things like Blogs, that allow the audience room to interact with the news, were inevitable.

We're already seeing the power of this new mode in other areas of culture/society/knowledge. The Open Source software movement is an easy example, but it's even having an impact outside of software circles. Wikipedia is probably the classic example of how tapping the principle mode of the Internet medium (interactivity) can be very powerful. It's weird that it worked so well; writing an online encyclopedia seems like the proverbial Sisyphean task, yet today it seems to have become the online world's primary source of fact. It's imperfect, but info in it is probably more reliable than you'd get from a random googling, more relevant than that antique set of tomes sitting on my bookshelf, and more up-to-the-minute than the local library.

Anyway, with all of the above in mind, I ran across WikiNews late last year. Basically the premise was simple - do for news what Wikipedia had done for fact. Even I thought it was a bit crazy - and I'm definitely no stranger to crazy ideas. But I'd just seen Outfoxed, and felt like we needed to start experimenting with some new ideas for news. I pitched in with a few articles, but found it a bit more time consuming than I was ready for. I figured it'd fizzle out eventually - I mean, who has time to write news articles every day?

Yet here we are, half a year later, and WikiNews is not only going strong, but I'm finding myself starting to use it as my primary source of news. They don't always have every article that cnn.com might have, but I actually like that; the bulk of stuff on the big news sites like CNN just seem too biased and corporate to me. But it does cover the important stories, with more factual integrity than you could expect from a blog. I think the thing I appreciate most about the stories is that I know they're getting scrutinized by many lovingly obsessive people with a passion for stomping out bias. I'm sure each of the news authors have their own axes to grind, but I'm also confident that the news that's there is (by definition) representative of things that real people care about.

Some day soon someone's going to figure out a way to combine Wikinews (for accurate up-to-the-minute news), blogs (for editorials and personal journalism), and Wikipedia (for peer-reviewed factual background). The public will be able to get involved in adding to the stories and correcting biases, share their own opinions directly on editorials, and when something newsworthy happens to them, be able to report on it in detail, directly. If corporate news is worrying about the impact of plain old blogs, wait 'til they see this...

Inkscape for Game Map Editing

Slashdot is running a story on Map-Making Software for RPG Campaigns. I found it odd that there was no mention of Inkscape; one of the main reasons I started using SVG way back when was for game maps. Anyway, here's the comment I posted:

I'd suggest looking into using SVG for game map creation, because there's getting to be a lot of Open Source tools out there (like Inkscape, that I help develop) that can edit, convert, etc. them. I've done some map making with it and while it lacks many of the advanced features that commercial map tools have, it's got the basics, plus if you can code, you gain the option of adding the feature in yourself. ;-)

Making maps with Inkscape / SVG is different than using CAD-style software like Campaign Cartographer, but you can achieve pretty much the same things. With features like alpha blending, text-to-shape, layers, grouping, shape fills, tiling, and infinite zoom, you can make much "prettier" maps in much less time than it'd take to do in a CAD-like program. See the screenshots to get some ideas of what can be done with these features. It has a fancy calligraphy mode that could be quite handy if you need to hand-write calligraphic text on a map. There's also a nifty bitmap-to-vector tracing tool that might help in converting hand-drawn maps to vectors. Also comes with several useful tutorials (in the Help menu).

There's also a site for sharing SVG clipart (like map symbols), the Open Clip Art Library. Not a lot of RPG art yet, but there's some and it's likely going to grow a lot. Plus, since all of its content is Public Domain, there's no restrictions at all placed on your maps if you use it. I could *easily* imagine this being a way for RPG mappers to collectively build an open library of RPG map symbols and artwork.

Task Management

I wrote earlier about task management, and chatting about it with Kees and Mike Day.

One of the aspects I'm interested in is generalizing a list of tasks that implement something, and having those 'procedures' be published in a way that lets people add to them. I've set up an OpenProcedures Wiki for doing this.

Inkscape Senior Projects

Ted posted an idea for an SVG difference tool as a project for college students to do with Inkscape. Here's some other projects that I think might be useful to consider:

  • Adding Chatroom capabilities to Inkview
  • Expanding on the Inkview code to make it into a more powerful presentation tool
  • Creating a framework for testing the rendering capabilities of a variety of SVG applications.
  • Create more Inkscape Extensions
  • Multi-page editing
  • Add SVG import capability to OpenOffice.org

There are more detailed writeups of the above in the Inkscape feature request tracker, or drop me an email.

A couple oldies but goodies: herding cats, and ballmer sells Windows.

11 older entries...

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!