The two issues below have been fixed and a few other issues which were resolved in the mean time have been backported to 0.26.1 as well. I was able to build the binaries using the abcl.release target (instead of using the separate abcl.binary.* targets), so it seems all our issues are now fixed and we can move to 0.26.1 next week.
Is there anybody with any remaining issues or fixes up his sleeve?
Anybody wants to report an issue with the current 0.26.x branch?
If not, I'm planning to roll the 0.26.1 release mid next week.
Bye,
Erik.
On Mon, Jul 11, 2011 at 10:18 PM, Erik Huelsmann ehuels@gmail.com wrote:
Hi,
On Mon, Jul 11, 2011 at 7:51 PM, Mark Evenson evenson@panix.com wrote:
On 7/10/11 12:53 AM, Erik Huelsmann wrote:
I've created the 0.26 release branch after updating CHANGES.
If there are no further objections, I'll continue the release process by signing the release.
It'd be very helpful if one of the other committers can update the webpages, as I won't have time to do that today or tomorrow.
Unfortunately, the abcl-0.26.0-src release won't build because that archive is missing the newly added org.armedbear.lisp.protocol package.
[Fixed and backported to the abcl-0.26.x branch][1].
My vote would be to figure out the stack overflow reported in [ticket #154][2] as well as whatever is causing the ansi-interpreted errors that Ville is noticing, and move to releasing an abcl-0.26.1.
Erik is on his first day of vacation, and probably has limited upload bandwidth, so we may be a bit constrained here.
Yea, it'd be very convenient for me to fix any remaining issues and go for a 0.26.1 in 2 weeks. Also because my signing key is most easily operated at home.
Let's take some time to solve these regressions and scrap the brown paper bag release :-)
Bye,
Erik.
On Jul 21, 2011, at 11:44 PM, Erik Huelsmann wrote:
The two issues below have been fixed and a few other issues which were resolved in the mean time have been backported to 0.26.1 as well. I was able to build the binaries using the abcl.release target (instead of using the separate abcl.binary.* targets), so it seems all our issues are now fixed and we can move to 0.26.1 next week.
Is there anybody with any remaining issues or fixes up his sleeve?
Anybody wants to report an issue with the current 0.26.x branch?
If not, I'm planning to roll the 0.26.1 release mid next week.
I believe we are still out two ANSI regressions due to the #\Unnnn syntax added to the reader for Unicode characters, but I haven't really had time to check in the past few days (vacation!), but I should be able to get to this over the next few days.
-- "A screaming comes across the sky. It has happened before, but there is nothing to compare to it now."
On 22 July 2011 08:27, Mark Evenson evenson@panix.com wrote:
Is there anybody with any remaining issues or fixes up his sleeve? Anybody wants to report an issue with the current 0.26.x branch? If not, I'm planning to roll the 0.26.1 release mid next week.
I believe we are still out two ANSI regressions due to the #\Unnnn syntax added to the reader for Unicode characters, but I haven't really had time to check in the past few days (vacation!), but I should be able to get to this over the next few days.
According to a test run I did last night, we don't have regressions compared to 0.25. Our current failure count is 22, not counting ensure-directories-exist because I didn't remove the scratch directory, and the previous count was 30, so we got a pretty nice improvement from the printing fixes. We did try and rush the release a bit, I suppose, but we picked up nicely.
I think we should just release it, and look at the next milestones, which in my mind would be roughly
1) strive for sub-20 failures, there's a couple that should be pretty straightforward although not trivial (clos, make-condition) 2) find out why our build nowadays takes 7 minutes, we used to be able to do it in 2 minutes. Something(s) has increased the build time quite a bit. 3) take a look at defstruct, it looks visibly slow when looking at the tests progressing
And of course any other improvements that scratch people's itches. :) All in all it looks to be in a good shape.
As discussed on #abcl last night, I have back ported the following to the 0.26.x branch
1) Make the [reader accept the #\Uxxxx syntax as a synonym but not make it canonical][1]
2) Include [asdf-2.017][2].
After [updating the CHANGES file][3], I am finished with all the tasks I had in mind for the release.
[1]: http://trac.common-lisp.net/armedbear/changeset/13416 [2]: http://trac.common-lisp.net/armedbear/changeset/13418 [3]: http://trac.common-lisp.net/armedbear/changeset/13419
-- "A screaming comes across the sky. It has happened before, but there is nothing to compare to it now."
armedbear-devel@common-lisp.net
-
Erik Huelsmann -
Mark Evenson -
Ville Voutilainen