179118 - Drop IBM VisualAge compiler on OS/2

Status: RESOLVED FIXED

(SeaMonkey :: Build Config, defect)

Description

[John Keiser (jkeiser)]

There is a creeping number of things we can't do with IBM's VisualAge 4.0
compiler on OS/2.  The compiler became unsupported several revisions ago
(http://www1.ibmlink.ibm.com/cgi-bin/master?xh=Fi7dVYO8vOv$7X1USenGnN9332&request=announcements&parms=H%5f901%2d013&xhi=usa%2emain&xfr=N).
 We should remove support and find a replacement.

Comment 1

[John Keiser (jkeiser)]

One more thing we will probably be able to do with VACPP gone is use bool and
true/false.  This is a potentially worrisome thing though, since as mkaply
points out some people assign true and false into PRBool when they should not. 
We almost certainly will not be able to replace PRBool with bool or typedef it
because it's not guaranteed to be the same size as PRInt32 and that would break
compatibility.  We *might* be able to just use bool and true/false internally,
assuming we feel the risk of people assigning true/false into PRBool is not high.

Comment 2

[John Keiser (jkeiser)]

setting dependent on bug 188249, gcc on OS/2.  And it feels goood.

Comment 3

[Evgen]

And what we see now with gcc ?
IMHO
Why IBM drop VAC 3.08/3.65 ? It is the most stable till now....

Comment 4

[John Keiser (jkeiser)]

It Is Done.  1.4 is the last shipping version with VACPP according to sicking
and mkaply.  Many thanks!