3 posts tagged “3.10”

It has been a while.  I could blame the holiday season partially, but ultimately I've just been distracted and somewhat lazy.  I've done the minimum of course but little more in the last couple of weeks.

Back in August of last year (2008) Klaus Witzel emailed me asking about all the open Mantis issues that appeared to have been addressed in 3.10.  On his prompting, at that point I took on the job of checking through all of the updates harvested for 3.10, 3.10.1, and 3.10.2 and closing the relevant Mantis reports where I was confident the proposed fix was indeed harvested and released.  Well nearly 5 months later I report that I'm pretty much done.  I closed the vast majority of the issues, I left a few open where further attention was needed, and there are a couple omore that may be closed after I get some feedback.

Hopefully this sort of thing will not be needed in the future.

Well, after the promise of Monday this week has not been particularly productive Squeak-wise.  I did nothing really in the last couple of days.  I did manage to come up with a little time today but only closed one issue.  But I learned a few things with this one.

The issue I closed is #5285.  I had looked at this one previously and gotten really confused.  I went back to it today with fresh eyes and started doing a manual version history bisection with the handy Dual Change Log tool for Monticello from Keith Hodges.  The EToys history was really looking strange.  I won't go into the details but suffice it to say that my conception of the version history as it appears in the tool was mistaken.  Also I was reminded that I have a very poor memory and mentioned that versions had been left out of 3.10 in the past.

Also I had been ignoring the fact the entries in the list can be in plain type, bold type, or underlined.  Bert Freudenberg kindly clued me in.  It turns out that bold entries are versions not in the ancestry of the current working copy (the version in the image).  And underlined versions are the current version, or more accurately (thanks to Matthew Fulmer) the immediate ancestor of the working copy.  Anyway this information all clued me into the fact that I was comparing the wrong versions.  Once I got that sorted out it was clear and I confirmed the code was harvested.

I also need to get out my monthly Box-Admins Team report and a new Roundup...

[edited: typo corrected]

The Squeak 3.10 release team was one of the first to start basing their work on reports entered into the Squeak Bug Server.  However for some reason they failed to close most of the reports that they addressed.  This was pointed out to me some months ago and off and on I have been going through all of the updates harvested for 3.10, checking the associated reports, comparing the reports to the code that was harvested, verifying that they match, and closing the reports with appropriate notices.  This is a slow and often tedious task.  Recently I've been trying to spend one hour per day on this task.

Today's hour had some distractions but most of the reports today were fairly straightforward and I managed to close seven of them.  I'm beginning to see the light at the end of the tunnel, it's possible I could finish this task this week or maybe next.