What's the oldest MS VC version Parrot should built with?

[Geoffrey Broadwell]

On Fri, 2008-12-19 at 21:29 +0000, Vadim Konovalov wrote:
> 在 Friday 19 December 2008 17:06:25，Geoffrey Broadwell 写道：
> > 2.  Anyone not running a current SP is probably running an insecure
> > system.  I consider it a net win for the world to strongly encourage
> > users to at least *minimally* secure their systems.
> 
> installing SP3 does not mean securing the system.

Of course not.  But the group of people who do not keep up to date with
service packs are the same ones that do not keep up to date with all of
the other security updates (and antivirus packages, and firewall rules,
and ...).  So I think it's valid to say that people backrev on the OS
are probably running an insecure system.  It's a matter of
*correlation*, not (necessarily) *causation*.

(And as a side note, once an operating system is officially de-supported
by its maker, it can be pretty much *assumed* insecure, because there
are no more security patches being made for it, and newer, more secure
versions of applications won't run on it.  No amount of anti-virus
software is going to cover up that mess.)

> I prefer other ways on securing.

Yes, but by the very fact of knowing that there are other security knobs
to turn, you've just proven that you are not in the group I'm talking
about.

> Or you actually believe all the MS advertising?

Of course not!

> Whenever I see the message "you turned off automatic updates. your computer 
> might be at risk" I understand that this is rather declaring that they are 
> about security, rather than actual security!
> 
> Some thing MS do quite good (MSVC, MS-Office are examples) but not security 
> solutions.

That is very much missing my point.

As much as I hate to admit it, there are still people who insist on
using old revs of Windows, and will most likely never switch away until
the machines literally go up in smoke.  (Heck, I'm related to some of
them.)  I can at least try to alter their behavior to cause less damage
to all of the rest of us, or tell them they don't get *support* for
acting that way.

> > *PROVIDED SOMEONE ACTS AS MAINTAINER*, we support VC6 for users to
> > manually build with, but it's not the compiler we use for default
> > prebuilt Windows binaries.  For these, we can use VC9, and require Vista
> > or WinXP SP3. [1]
> >
> > Is that an acceptable compromise?
> 
> I thought there is already such a compromise accepted.

Good then, I must have missed that.

> Except requiring WinXP-SP3 or Vista is catastrophic from marketing POV.
> At least, providing an explanation on how to install certain missing CRT  is 
> much better solution.

That's what I was saying.  The big green download arrow at the top of
the page points to the new-OS-required version.  Below it in
(relatively) small print is a link saying "Running an older version of
Windows?  Use this installer instead", pointing to the back-compatible
build, or perhaps just a larger install file with the missing CRT
included.

> I very much doubt that you will ever see WinXP-SP3 requirement for ruby, 
> python, tcl, etc.

There's a difference between "absolute requirement to run at all" and
"required for optimized fully supported version with smallest installer
-- but larger and/or slower and/or not fully supported version still
available".  I'm not suggesting we go with the "absolute requirement"
version; that would be an overreaction.

Also:  Many of the major *applications* are starting to require new
versions of their respective OSen.  Many Mac apps require OS 10.4 or
later to run the latest rev.  Many Windows apps require WinXP (or
sometimes WinXP SP2) to run properly.  Even in the Linux world apps,
libraries, and system daemons can only be backported so far.  I doubt
you can run a modern desktop hardware manager on an ancient kernel ....

> Best regards,
> Vadim.


-'f