One reason I dropped out of Microsoft beta programs years ago is that I
would raise these issues, such as “The GUI is unusable” or “This critical
feature no longer works”, and I was told “This is by design”. That is,
“beta” testing was only to find bugs in the product *as delivered*, not bugs
relative to, let’s say, bad design decisions. The product was assumed to be
perfect in design, and all they wanted to know was “does it work right when
used in the way we designed it?” rather than “does this meet your needs?” or
“is this product worth releasing at all in its current form?” It was a
given that questioning the design was outside the scope of the beta tester’s
acceptable responses, and after a few years of this, I stopped testing.
It didn’t help that I was sent the beta of, say, “Chicago”, not of “The next
MS-DOS-based Windows Release” (Yes, most of us knew “Chicago” by that point,
but, for example, I just learned that “groove” has something to do with
laptop synchronization and nothing to do with, say, Zune players). The
betas were sent out without any information describing what the product was,
what it would do for use, why we might want to test it, etc. All we got was
the internal project name! At least for WDK, we have some idea of the
purpose of the product!
Either Microsoft needs to re-evaluate what they mean by “beta testing” or
get the professional users in the loop much earlier so we don’t have to
fight the broken-as-designed (BAD) software issues after the fact.
joe
-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of David R. Cattley
Sent: Sunday, January 18, 2009 1:13 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] WDK7 custom target path
I agree whole heartedly with Peter that building IHV/ISV drivers with the
very same tool-set is critically important.
I just wish that there were perhaps a more ‘communicative’ approach to
having changes arrive in the tools. Maybe beta builds of the WDK are just
that and I should just provide feedback on the alias in hopes that the “oh,
so you were using *that*” response implies a vote for putting it back.
TARGETPATH falls into that category because *yes* I use that (principally
because I was too darn lazy to figure out BINPLACE for which I no longer
have an excuse).
But Don’s illustration of ‘preventing drop-in adoption’ because the kit
itself requires build system work that breaks using earlier kits rings true
with me as well.
My favorite illustration of this nonsense is the removal of amd64 and
replacement with x64 as an argument to setenv.bat that rolled into the WDK
rather silently. Now *that* would have been an absolute no-brainer to
support both switches and not break Mark’s or OSR’s DDKBUILD.BAT in a
multiple kit scenario.
So yeah, TARGETPATH might not be the battle to pick but it does mean *some*
amount of work to build processes will be necessary to just *test* the WDK
where it has been used. Maybe those changes are for the good. The problem
is that nobody has yet illustrated and explained why TARGETPATH is now such
a bad idea where it had survived quite usable for over a decade of kits.
And what *are* the changes in BUILD - aka, what should we be asking about
that we can only discover by being blind-sided by the new kit? We are
black-box testers here. The sanitized release notes for the kit don’t offer
much insight to the changes and the reasoning behind them.
David R. Cattley
Consulting Engineer
Systems Software Development
NTDEV is sponsored by OSR
For our schedule of WDF, WDM, debugging and other seminars visit:
http://www.osr.com/seminars
To unsubscribe, visit the List Server section of OSR Online at
http://www.osronline.com/page.cfm?name=ListServer
–
This message has been scanned for viruses and dangerous content by
MailScanner, and is believed to be clean.