Add a bit more explanations on the meaning of OBSOLETE and EXPERIMENTAL.
author"Yann E. MORIN" <yann.morin.1998@anciens.enib.fr>
Mon Apr 20 21:30:25 2009 +0000 (2009-04-20)
changeset 13007acdd1de5b50
parent 1299 3448ac3f1a5d
child 1301 2597ecea561c
Add a bit more explanations on the meaning of OBSOLETE and EXPERIMENTAL.

-------- diffstat follows --------
/trunk/config/global/ct-behave.in | 22 13 9 0 +++++++++++++---------
/trunk/docs/overview.txt | 8 6 2 0 ++++++--
2 files changed, 19 insertions(+), 11 deletions(-)
config/global/ct-behave.in
docs/overview.txt
     1.1 --- a/config/global/ct-behave.in	Mon Apr 20 21:10:03 2009 +0000
     1.2 +++ b/config/global/ct-behave.in	Mon Apr 20 21:30:25 2009 +0000
     1.3 @@ -8,9 +8,13 @@
     1.4      default n
     1.5      help
     1.6        If you set this to Y, you will be able to select obsolete features.
     1.7 -
     1.8 -      Such obsolete features are the use of old kernel headers, old
     1.9 -      gcc versions, etc...
    1.10 +      
    1.11 +      Such obsolete features are the use of old kernel headers, old gcc
    1.12 +      versions, etc... for which maintaining support in crosstool-NG
    1.13 +      would be very costly.
    1.14 +      
    1.15 +      It does not however mean that the specific feature or version has been
    1.16 +      marked obsolete by the upstream team.
    1.17  
    1.18  config EXPERIMENTAL
    1.19      bool
    1.20 @@ -19,7 +23,7 @@
    1.21      help
    1.22        If you set this to Y, then you will be able to try very experimental
    1.23        features.
    1.24 -
    1.25 +      
    1.26        Experimental features can be one of:
    1.27          - working, in which case you should tell me it is!
    1.28          - buggy, in which case you could try patching and send me the result
    1.29 @@ -32,7 +36,7 @@
    1.30      prompt "Debug crosstool-NG"
    1.31      default n
    1.32      help
    1.33 -      Say 'y' here to get some debugging options
    1.34 +      Say 'y' here to get some options regarding debugging crosstool-NG.
    1.35  
    1.36  if DEBUG_CT
    1.37  
    1.38 @@ -51,10 +55,10 @@
    1.39      help
    1.40        If you say 'y' here, then you will be able to restart crosstool-NG at
    1.41        any step.
    1.42 -
    1.43 -      It is not currently possible to restart at any of the debug facility.
    1.44 +      
    1.45 +      It is not currently possible to restart at any of the debug facilities.
    1.46        They are treated as a whole.
    1.47 -
    1.48 +      
    1.49        See docs/overview.txt for the list of steps.
    1.50  
    1.51  config DEBUG_CT_SAVE_STEPS_GZIP
    1.52 @@ -65,7 +69,7 @@
    1.53      help
    1.54        If you are tight on space, then you can ask to gzip the saved states
    1.55        tarballs. On the other hand, this takes some longer time...
    1.56 -
    1.57 +      
    1.58        To lose as less time as possible, the gzip process is done with a low
    1.59        compression ratio (-3), which gives roughly 70% gain in size. Going
    1.60        further doesn't gain much, and takes far more time (believe me, I've
     2.1 --- a/docs/overview.txt	Mon Apr 20 21:10:03 2009 +0000
     2.2 +++ b/docs/overview.txt	Mon Apr 20 21:30:25 2009 +0000
     2.3 @@ -277,13 +277,17 @@
     2.4    Show obsolete versions or tools. Most of the time, you don't want to base
     2.5    your toolchain on too old a version (of gcc, for example). But at times, it
     2.6    can come handy to use such an old version for regression tests. Those old
     2.7 -  versions are hidden behind CT_OBSOLETE.
     2.8 +  versions are hidden behind CT_OBSOLETE. Those versions (or features) are so
     2.9 +  marked because maintaining support for those in crosstool-NG would be too
    2.10 +  costly, time-wise, and time is dear.
    2.11  
    2.12  CT_EXPERIMENTAL:
    2.13    Show experimental versions or tools. Again, you might not want to base your
    2.14    toolchain on too recent tools (eg. gcc) for production. But if you need a
    2.15    feature present only in a recent version, or a new tool, you can find them
    2.16 -  hidden behind CT_EXPERIMENTAL.
    2.17 +  hidden behind CT_EXPERIMENTAL. Those versions (or features) did not (yet)
    2.18 +  receive thorough testing in crosstool-NG, and/or are not mature enough to
    2.19 +  be blindly trusted.
    2.20  
    2.21  Re-building an existing toolchain |
    2.22  ----------------------------------+