[Scons-users] Checking for specific versions of visual studio

Fri Oct 30 19:59:03 EDT 2015

On Fri, Oct 30, 2015 at 11:35 PM, Bill Deegan <bill at baddogconsulting.com>
wrote:

> +1 on keeping the warning.
> If you ask for version x and it's not available then you've gone to the
> trouble to request a specific version, you should be notified.
> Arguably that should be an error.
>

I think we agree on this; it should be an error.

>
> Now if you wanted to be able to silence the warning in your SConstruct, or
> via command line, that would be in keeping with SCons's general philosophy.
>
>
> -Bill
>
> On Fri, Oct 30, 2015 at 11:59 AM, William Blevins <wblevins001 at gmail.com>
> wrote:
>
>>
>>
>> On Fri, Oct 30, 2015 at 5:50 PM, Brian Cody <brian.j.cody at gmail.com>
>> wrote:
>>
>>> Follow up! (now using SCons 2.4.0)
>>>
>>> This particular error goes away if in msvc.py, we handle exists this way:
>>> def exists(env):
>>>     return msvc_exists(env['MSVC_VERSION'])
>>>
>>> We could check for the presence of that key and send a None in the case
>>> the key isn't set. I don't think this should break any existing behavior.
>>> Maybe I'll look into how to run scons internal tests, see if I can expose
>>> the bug, and then submit a fix?
>>>
>>>
>>>
>>> Also in vc.py line 428 we issue a nagging warning that doesn't help
>>> anyone. This warning basically says "you created an environment for visual
>>> studio 9.0 but you only have visual studio xx installed'. Well that's
>>> factual information, but who cares? That's the point of exists. So because
>>> I created an environment for a target I may or may not be building, I'm
>>> going to get nagged? The other SCons warning printout in this file is very
>>> different. It tells you if you've created an environment with conflicting
>>> MSVC and MSVS versions. This to me feels proper.
>>>
>>> So this one is less clear to me, but I'd petition nuking the warning. If
>>> we don't, then there's no way with the aid of scons to query for the
>>> presence of a tool before getting scolded. Would a removal of this warning
>>> be accepted, or would it be shot down? It could go into that debug log
>>> instead if it was really desirable to someone somewhere.
>>>
>>>
>> My personal opinion is that SCons lives, sleeps, and breathes
>> correctness.  If a warning implies that your build may give unexpected
>> results, then you should get an error.  If the warning doesn't imply
>> anything, then it shouldn't exist.
>>
>> Let's see what the others say.
>>
>> V/R,
>> William
>>
>>
>>> Thanks!
>>>
>>>
>>> On Thu, Sep 17, 2015 at 7:39 AM, Brian Cody <brian.j.cody at gmail.com>
>>> wrote:
>>>
>>>> Bill,
>>>>
>>>> It reports "9.0".
>>>>
>>>> On Wed, Sep 16, 2015 at 10:39 AM, Bill Deegan <
>>>> bill at baddogconsulting.com> wrote:
>>>>
>>>>> Brian,
>>>>>
>>>>> Looks like SCons is only finding MSVC 11 and 11.0Exp installed on your
>>>>> system.
>>>>> Do you have MSVC 9.0 installed?
>>>>>
>>>>> What's the value of env['MSVC_VERSION'] after you initialize your
>>>>> environment?
>>>>>
>>>>> -Bill
>>>>>
>>>>> On Wed, Sep 16, 2015 at 8:02 AM, Brian Cody <brian.j.cody at gmail.com>
>>>>> wrote:
>>>>>
>>>>>> Bill,
>>>>>>
>>>>>> Here's the file you requested. I never realized the option for it
>>>>>> existed! Thanks for that.
>>>>>>
>>>>>> http://pastebin.com/0pqdMYzv
>>>>>>
>>>>>> -Brian
>>>>>>
>>>>>>
>>>>>> On Tue, Sep 15, 2015 at 9:31 AM, Bill Deegan <
>>>>>> bill at baddogconsulting.com> wrote:
>>>>>>
>>>>>>> Can you run with:
>>>>>>>
>>>>>>> set SCONS_MSCOMMON_DEBUG=c:\somefile.log
>>>>>>> scons <whatever>
>>>>>>> and pastbin the log file?
>>>>>>>
>>>>>>> -Bill
>>>>>>>
>>>>>>> On Tue, Sep 15, 2015 at 9:01 AM, Brian Cody <brian.j.cody at gmail.com>
>>>>>>> wrote:
>>>>>>>
>>>>>>>> Hey guys, a while ago I asked about how exists() is used for tools
>>>>>>>> and I was told it pretty much isn't. We have an scons.workarounds we use to
>>>>>>>> get.. around things like this. Today it looks like this:
>>>>>>>>
>>>>>>>> _already_warned_about_tool = set()
>>>>>>>>
>>>>>>>> def does_tool_exist(env, tool):
>>>>>>>>     return SCons.Tool.Tool(tool, [str(SCons.Script.Dir(path)) for
>>>>>>>> path in env['toolpath']]).exists(env)
>>>>>>>>
>>>>>>>> def verify_tool_exists(env, tool):
>>>>>>>>     exists = does_tool_exist(env, tool)
>>>>>>>>     if not exists:
>>>>>>>>         if tool not in _already_warned_about_tool:
>>>>>>>>             _already_warned_about_tool.add(tool)
>>>>>>>>             sys.stderr.write('*** Tool "' + tool + '" does not
>>>>>>>> exist; targets skipped.\n')
>>>>>>>>             if env['fail_if_tools_are_unavailable']:
>>>>>>>>                 raise
>>>>>>>> SCons.Errors.BuildError(errstr='fail_if_tools_are_unavailable is enabled.')
>>>>>>>>     return exists
>>>>>>>>
>>>>>>>> This is used because we have some developers who don't have all
>>>>>>>> tools who still want to be able to "build all that I can". Now the issue
>>>>>>>> we're seeing is that we create an environment specifically for visual
>>>>>>>> studio 9.0. i.e.
>>>>>>>> Environment(tools = ['msvc', 'mslink', 'masm' , 'mslib'],
>>>>>>>> MSVC_VERSION='9.0', ......)
>>>>>>>>
>>>>>>>> With this environment, the above function will successfully tell
>>>>>>>> you if visual studio is installed or not, however it will return True even
>>>>>>>> if the version is 12.0. At the moment I'm not sure that we can go to 12
>>>>>>>> with our version of the compact framework. Our easier solution would be
>>>>>>>> that this function should just return false. Deep in the bowels of the
>>>>>>>> visual studio support I see that the "msvc_exists" takes a version number,
>>>>>>>> but the exists() in the tool does not attempt to use it. In "msvs_exists"
>>>>>>>> it does not allow the same version check.
>>>>>>>>
>>>>>>>>
>>>>>>>> So my leading question: How am I supposed to offer this
>>>>>>>> functionality? If this is a half-implemented in SCons, any suggestion on
>>>>>>>> how to solve this generically from our standpoint without injecting changes
>>>>>>>> into the SCons package?
>>>>>>>>
>>>>>>>> Thanks
>>>>>>>>
>>>>>>>> _______________________________________________
>>>>>>>> Scons-users mailing list
>>>>>>>> Scons-users at scons.org
>>>>>>>> https://pairlist4.pair.net/mailman/listinfo/scons-users
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> Scons-users mailing list
>>>>>>> Scons-users at scons.org
>>>>>>> https://pairlist4.pair.net/mailman/listinfo/scons-users
>>>>>>>
>>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> Scons-users mailing list
>>>>>> Scons-users at scons.org
>>>>>> https://pairlist4.pair.net/mailman/listinfo/scons-users
>>>>>>
>>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Scons-users mailing list
>>>>> Scons-users at scons.org
>>>>> https://pairlist4.pair.net/mailman/listinfo/scons-users
>>>>>
>>>>>
>>>>
>>>
>>> _______________________________________________
>>> Scons-users mailing list
>>> Scons-users at scons.org
>>> https://pairlist4.pair.net/mailman/listinfo/scons-users
>>>
>>>
>>
>> _______________________________________________
>> Scons-users mailing list
>> Scons-users at scons.org
>> https://pairlist4.pair.net/mailman/listinfo/scons-users
>>
>>
>
> _______________________________________________
> Scons-users mailing list
> Scons-users at scons.org
> https://pairlist4.pair.net/mailman/listinfo/scons-users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://pairlist4.pair.net/pipermail/scons-users/attachments/20151030/77fc3b18/attachment-0001.html>