Mantis - Squeak
Viewing Issue Advanced Details
6571 Graphics crash sometimes 07-22-07 01:51 10-09-07 16:43
assigned 3.10  
0006571: In 3dot10 running font test can crash system
See Ralphs letter on v3dot10. [^]
See my peliminary comments on v3dot10 [^]

Basically squeak has hit the babel limit. Too many cooks and the font and font theme preferences were added but not reconciled in 3dot9.

Now that the emphasis is on tests, these problems are being detected.

First suspects are the small land themes.

Instead of setting systemfonts (and therefore default fonts first) they were being set fourth.
I've uploaded "corrected" theme orders for evaluation.

Still we are at the beginning and not the end of this puzzle.

Yours in curiosity and service, --Jerome Peace

related to 0006715assigned andreas An issue to collect font tests and font test knowledge 
child of 0006570assigned tim A Mother for font and font test problems 
 Menubefore.gif [^] (8,667 bytes) 07-22-07 04:08
 Menuafter.gif [^] (14,263 bytes) 07-22-07 04:09
 Fonto changeO.text [^] (1,689 bytes) 07-22-07 04:14
 SetSysFont1st-wiz.1.cs [^] (3,067 bytes) 07-22-07 04:15

07-22-07 04:19   
hard evidence:

For this one:

In a fresh 7122

evaluate "Preferences printStandardSystemFonts" in a workspace

In the sunit tests run only LangEnvBugs (the only test is >>testIsFontAvailable )

then re-evauate the expression. ( Fonto changeO.text shows the result as well as the code for the test )

The test which goes to the trouble of restoring the preference it changes still manages to get a permanent change in system fonts.

Everything ends up as a ttfont bitstreamVeraSans.

So the finger points at the 3.8 ttfont stuff. ( I think the ttfonts delete the other fonts with prejudice.) stay tuned.

Yours in curiosity and service, Jerome Peace, bug tracker
07-22-07 05:05   
Has anyone other than Ralph actually seen a crash? I run the tests with no ill effects (other than it changing the default fonts).