work

Cosmo VM Application?

December 2nd, 2008 Comments Off

Both Parallels and VMWare support “applications”, which are pre-configured snapshots.

I wonder how hard it would be to set up an appliance for Chandler Server?

Chandler 0.7.4 released

February 14th, 2008 Comments Off

I’m not being paid by OSAF any more, but I’m still following Chandler and am poking at the code (it’s what open source is all about).

Version 0.7.4 of Chandler has been released.

Mac Tiger Intel
Mac Tiger PowerPC
Mac Leopard
Windows
Linux tar.gz
Ubuntu Linux .deb

With Leopard and Ubuntu Chandler is now using the OS’s version of Python, which is interesting.

Chandler 0.7.3 released

December 7th, 2007 No Comments »

Chandler 0.7.3, sporting the month view

It’s official, version 0.7.3 is out. Get your Intel Tiger or Leopard version, or the PPC build.

Let’s not forget those on Windows or Ubuntu 7.10.

Be sure to check out the Release notes!

Design iteration on Multiweek View

November 24th, 2007 No Comments »

So I’ve been working on a new design for parts of the Multiweek view. Indeed, I knew I would have to do this slight rework, and made sure not to write too much throw-away code on the first draught (or “draft” for the USians in the audience

). There is some, since we needed something stable and functional, but it is a bit rickety.

UML of Multiweek and some associated classes

wxMultiWeekCanvas bits

Part of reworking the code involved having good documentation fo the current setup. I used a tool written in wxPython called PyNSource to analyze the source and build a UML diagram (seen above; click for larger version).

But all that was too dense for my poor eyes to read on paper (to print I captured the screen to a JPEG screenshot and printed that — I should try the native printing, as deficient as it may be). So I cropped it a bit and printed in bigger type, with the shot you see to the right. (Again, click for larger version).

The main change I am thinking of is making the basic UI rendering unit be the week instead of the day. That makes drawing lozenges across multiple days pretty straight forward. Also, it allows arrangement of said lozenges in a rational manner.

Another element is to keep multi-day events separate from single-day events. I’ve started to wonder if even a “normal” event that spans two days should have some sort of different rendering from the one-line text we use for “normal” events.

I don’t know, sometimes simple things can go from say 11pm to 3am and you wouldn’t want that to look different. Personally, I would prefer everything up to 5am to go in the previous day. e.g. tonight, I am up late after playing board games (Power Grid with the new expansion pack tonight!) but I wouldn’t want my “gaming” event to be considered as spanning two days.

5am seems arbitrary, and should be a user preference. Alas, Chandler is weak in the area of UI for user preferences. But the code can be written in such a way as to make it easy to change/preference-ize later.

Just out of curiosity, if you read this please leave a comment saying what you think about this proposition: events that start before midnight and last beyond 5am the following day should be considered as “multi-day” normal events, and should be rendered differently. You can just say “+1″ (like it) or “-1″ (don’t like it) or “0″ (don’t care either way).

Right now, I am only thinking of some sort of rectangular shading, but if Mimi gets into this I’m sure we’ll have quite a snappy treatment.

svn merge problems

October 23rd, 2007 2 Comments »

I seem to be having problems merging from the trunk to the multiweekview branch. Specifically, a file that should be there on the branch, isn’t.

The last time I merged trunk changes into the multiweekview branch was at r15506, when I merged from r15455 through 15504. That went sort of ok (I did the commit of the merge from the multiweekview/chandler directory, and not the multiweekview directory, so I had to do two commits. Oops.).

In r15531, Bryan Stearns checked in a change that deleted a bunch of files and created a bunch of new ones that replaced them. I know, because I reviewed the patch! (r=rae).

I will choose one file as an example: chandler/Chandler.egg-info/resources/images/TriageNow.png. This file existed in r15530, but it was gone in 15531.

On the multiweekview branch, TriageNow.png continued to live, since we had not merged in Bryan’s change. Recently, I wanted to merge in that change, along with all the others that have happened since the last merge. So I executed this command:

% svn merge -r 15505:15564 svn+ssh://svn.osafoundation.org/svn/chandler/trunk

This seemed to work well, until the final line of output from svn, where it prints a series of dots “………….”. My computer hung for about ½ hour.

It was Friday, and I had to get going, and I didn’t want to block Jeffrey from checking in his changes to the branch, so I hit control-C to kill the commit and told Jeffrey to go ahead, resigning myself to figuring it all out later.

Alas, later has lasted a long time.

When I did an svn diff to check how things were, svn said there were no changes. That was a bit strange, since the commit hadn’t finished. Or, maybe it had?

I had checked in all my changes to the branch before merging, so there was no uncommitted work on the branch hanging around, so I was able to freely deleted the entire directory and check out the branch again. After checkout I ran Chandler, and I saw a few changes from the trunk[!]. Specifically, the new wx toolbar appearance.

This was weird. I checked the svn logs as mailed out and didn’t see any record of my merge being committed. Again, “svn diff” reported nothing. So, I did the merge again, same command line as before. This caused output that had errors in it:

A    chandler/Chandler.egg-info/resources/images/Triage.Now.Rollover.Right.png
A    chandler/Chandler.egg-info/resources/images/Markup.Later.StampedRollover.Left.png
A    chandler/Chandler.egg-info/resources/images/Triage.Done.Left.png
A    chandler/Chandler.egg-info/resources/images/Markup.Done.StampedPressed.Right.png
Skipped missing target: 'chandler/Chandler.egg-info/resources/images/MarkupNowStampedPressed.png'
Skipped missing target: 'chandler/Chandler.egg-info/resources/images/MarkupNowStamped.png'
Skipped missing target: 'chandler/Chandler.egg-info/resources/images/TriageLater.png'
Skipped missing target: 'chandler/Chandler.egg-info/resources/images/TriageDone.png'
Skipped missing target: 'chandler/Chandler.egg-info/resources/images/MarkupNowStampedRollover.png'
Skipped missing target: 'chandler/Chandler.egg-info/resources/images/TriageLaterRollover.png'
Skipped missing target: 'chandler/Chandler.egg-info/resources/images/TriageDoneRollover.png'
Skipped missing target: 'chandler/Chandler.egg-info/resources/images/MarkupLaterStampedPressed.png'
Skipped missing target: 'chandler/Chandler.egg-info/resources/images/TriageNowMousedown.png'
Skipped missing target: 'chandler/Chandler.egg-info/resources/images/MarkupDoneStampedPressed.png'
Skipped missing target: 'chandler/Chandler.egg-info/resources/images/MarkupDoneStamped.png'
Skipped missing target: 'chandler/Chandler.egg-info/resources/images/MarkupDoneStampedRollover.png'
Skipped missing target: 'chandler/Chandler.egg-info/resources/images/MarkupLaterStamped.png'
Skipped missing target: 'chandler/Chandler.egg-info/resources/images/TriageDoneMousedown.png'
Skipped missing target: 'chandler/Chandler.egg-info/resources/images/MarkupLaterStampedRollover.png'
Skipped missing target: 'chandler/Chandler.egg-info/resources/images/TriageLaterMousedown.png'
Skipped missing target: 'chandler/Chandler.egg-info/resources/images/TriageNow.png‘
Skipped missing target: ‘chandler/Chandler.egg-info/resources/images/TriageNowRollover.png’
A    chandler/util/triagebuttonimageprovider.py

What this is saying is that svn expected there to be files there, including our trusty TriageNow.png, but they weren’t there. Now, since these files were deleted in r15531, svn should be able to handle that as part of the merge. So I’m thinking the svn database is a bit screwed up.

Then I hear from Bear that my Friday-night-never-ending svn process is still running on the server. He kills it for me, and I try again. Exact same results.

So I **could** try adding the above files myself and see if that makes svn happy. Or maybe we need to do some svn admin work to fix things. I’m not sure.

Chandler Sharing video

October 10th, 2007 No Comments »

I’ve uploaded the Chandler Sharing video to YouTube, Vimeo and Viddler. The YouTube video is much lower quality, and is only 720k, while the Vimeo is higher quality and is 3.9 MB; Viddler is in the middle at 2.1 MB. The Viddler player has the most features and is very community-friendly.

The fullscreen option for Vimeo is really nice. YouTube’s (at least on my Mac) seems buggy. Actually YouTube’s player is generally not the best, and doesn’t “seek” very well.

Vimeo

Chandler - Share via Hub from Reid on Vimeo.

YouTube

Viddler

Chandler Preview is out!

September 12th, 2007 No Comments »

Be sure to get a copy and give it a whirl:

Intel Mac OS X
PPC Mac OS X
Windows
Linux

“repository view” is just a database ref

July 28th, 2007 No Comments »

Something that I find often confuses code I am looking at is the use in Chandler of the term ‘view’ to refer to a database cursor/reference. In most apps out there, ‘view’ refers to a very specific piece of UI.

I wish we could globally substitute something like ‘dbRef’ wherever we use ‘view’ in Chandler’s code. It would make reading the code a lot easier clearer.

And actually, while we’re at it, replace all uses of ‘repository’ with ‘repo’. ‘repository’ doesn’t mean very muchand ‘repo’ stands out more.

Ya, I seem to be in a complaining mood.

Chandler migration - happy surprise

May 22nd, 2007 No Comments »

I contributed to the Chandler migration testing we were doing today and ran across a happy surprise. The idea of the exercise was to migrate data from Chandler 0.6alpha4 to the latest checkpoint build. In the course of this, I exported my calendars both as .ics files and via sharing on the server.

When I loaded up the latest chandler, I first connected to the server and made sure I could see all my calendars, which I could, and then I imported one of the .ics files. Lo and behold, it noticed that I already had the same items from the server and merged the items as they were read in. I ended up not with two copies of every item, but instead, each item was referenced by two calendars. (I had renamed the calendar before importing the .ics file to see if it would create a collection with the same name - and it did).

I was quite pleased at this bit of smarts, as you can tell!

Performance

April 26th, 2007 No Comments »

Last night before heading to bed, I left my Mac running two series of performance tests. One with my wx.GraphicsContext changes turned off, and another with them turned on. The command line was:

date
cp -f parcels/osaf/framework/blocks/calendar/CollectionCanvas{-off,}.py \
&& perftest off \
&& cp -f parcels/osaf/framework/blocks/calendar/CollectionCanvas{-on,}.py \
&& perftest on
date

The tests started at Thu Apr 26 02:59:25 EDT 2007 and finished at Thu Apr 26 05:28:06 EDT 2007. So they ran for about 2½ hours.

You don’t want to do this kind of thing during the day! This ended up being a long post with code, etc, so click to see more.

Chomping at Chandler