320449 - Make properties of GRE_GetGREPathWithProperties useful

Status: RESOLVED FIXED

(Toolkit Graveyard :: XULRunner, defect, P1)

Description

[Benjamin Smedberg]

The "properties" of GRE_GetGREPathWithProperties is currently unused, and it should be because we're probably going to end up with XULRunner GRE 1.8 and seamonkey GRE 1.8.

So I've added two properties:

xulrunner=true (this GRE has a xulrunner binary)
javaxpcom=1 (this GRE supports javaxpcom version 1)

I know that there are other useful properties that could be added (ABI, graphics toolkit), but I couldn't find a quick and simple way to get all the data I wanted so I'd like to stick with these two for 1.8.

Comment 1

[Benjamin Smedberg]

Attachment: xulrunner=true and javaxpcom=1

Comment 2

[Darin Fisher]

Comment on attachment 206010 [details] [diff] [review]
xulrunner=true and javaxpcom=1

How about "javaxpcomversion" or "javaxpcom-version" or something along those lines to make it clear that the property's value is a version field.  what are the rules for parsing that version field?  do we assume dotted notation?  why not "1.0.0.0" to make that very clear?

Comment 3

[Benjamin Smedberg]

We do not parse the field (currently): it's just a string. I'm happy with javaxpcom-version if you think that's better.

Comment 4

[Darin Fisher]

Comment on attachment 206010 [details] [diff] [review]
xulrunner=true and javaxpcom=1

r=darin

I think NS_ARRAY_LENGTH should be moved into nscore.h ;)

Comment 5

[Darin Fisher]

> We do not parse the field (currently): it's just a string. I'm happy with
> javaxpcom-version if you think that's better.

It doesn't really matter that much to me.  I was just thinking that someone
else might try to parse these values :-/

Comment 6

[Benjamin Smedberg]

Attachment: urgh, windows glue is wrong

This was a logic error in the windows glue that I didn't catch until testing these properties... it won't affect ffox/tbird because they don't use the GRE/glue.

Comment 7

[Benjamin Smedberg]

Comment on attachment 206010 [details] [diff] [review]
xulrunner=true and javaxpcom=1

xulrunner-only

Comment 8

[Benjamin Smedberg]

Fixed on trunk. I did not move NS_ARRAY_LENGTH because I want to get this on branches with as little pain as possible, that can be done later.

Comment 9

[Mike Schroepfer]

Comment on attachment 206085 [details] [diff] [review]
urgh, windows glue is wrong

Please consider for 1.8.1 - 1.8.0.1 is for major security and crash issues only.

Comment 10

[Benjamin Smedberg]

Comment on attachment 206085 [details] [diff] [review]
urgh, windows glue is wrong

Re-requesting approval, this does not affect ff/tbird which do not use the GRE-finding functions, it only affects xulrunner GRE-finding.

Comment 11

[Benjamin Smedberg]

Comment on attachment 206010 [details] [diff] [review]
xulrunner=true and javaxpcom=1

a=drivers

Comment 12

[Benjamin Smedberg]

Comment on attachment 206085 [details] [diff] [review]
urgh, windows glue is wrong

a=drivers

Comment 13

[Benjamin Smedberg]

Fixed on MOZILLA_1_8_BRANCH and MOZILLA_1_8_0_BRANCH.