[prev in list] [next in list] [prev in thread] [next in thread] 

List:       ms-smartcardddk
Subject:    Re: BWT calculated by Smartcard Library.
From:       Klaus Schutz <kschutz () MICROSOFT ! COM>
Date:       1999-04-19 18:14:42
[Download RAW message or body]


Jan,

in case of the BWT, the calculated time is currently too short.
With a newer version of smclib, 'your' BWT would be twice as long
as it should be. You 'just' would wait a longer than you have to.
This means, if a card is 'dead' (does not answer within BWT),
the end user has to wait twice as long before s/he gets a timeout.
I consider this only an inconvenience.

We try to keep backwards compatibility whenever we release
a new version of smclib.sys. We run tests against several
logo'ed devices with the version we're about to release.

We will keep the 'dll' model of smclib since it allows
us to fix protocol problems with cards(that we have seen
quite often). If an 'important' card fails to work with several
readers, we only need to 'fix' smclib.sys and the customer
does not need to get new drivers.

Klaus U Schutz
Microsoft

-----Original Message-----
From: Jan Bottorff [mailto:janb@PMATRIX.COM]
Sent: Thursday, April 15, 1999 6:35 PM
To: SmartCardDDK@DISCUSS.MICROSOFT.COM
Subject: Re: BWT calculated by Smartcard Library.


At 11:56 AM 4/12/99 -0700, Klaus Schutz wrote:
>you're right, the calculated BWT is incorrect in your case.
>You have to multiply the calculated BWT by a factor of 2.
>This bug will be fixed in the next release.
>
>This does not affect readers running at 3.5MHz - 4MHz.
>
>Thanks for this 'bug report'
>
>Klaus U Schutz
>Microsoft

So a little question Klaus. Let's suppose a driver get's written and has to
work around this incorrect calculation of BWT. If Microsoft then releases a
new smart card library DLL that fixes it, won't that break the driver? Or
is the smart card library written such it will look at the library version
assumed by the driver, and be bug for bug compatable?

I'd like to suggest that if Microsoft fixing their bugs, breaks our
drivers, it's strong evidence that making the smart card library a static
library would be a better plan. The option to just keep the old smart card
library may not be viable either. Imagine the case where a user has two
card readers, from different vendors. One was built and tested to work
around Microsoft's bugs in an older shared library, and the other didn't
need the workaround because it was written later, using a fixed version of
the library. This may be especially annoying if a user has a reader already
installed, written to work around current smclib.sys bugs. They install
some new reader, which updates the version of smclib.sys. They now find
their new reader works, but their old one is broken. They uninstall the new
reader and STILL find their old reader is broken, as the updated smclib.sys
will not get rolled back to the old version.

- Jan

___________________________________________________________________
            Paradigm Matrix Inc., San Ramon California
   "video products and development services for Win32 platforms"
Internet: Jan Bottorff janb@pmatrix.com
          WWW          http://www.pmatrix.com
Phone: voice  (925) 803-9318
       fax    (925) 803-9397
PGP: public key  <http://www-swiss.ai.mit.edu/~bal/pks-toplev.html>
     fingerprint  52 CB FF 60 91 25 F9 44  6F 87 23 C9 AB 5D 05 F6
___________________________________________________________________

[prev in list] [next in list] [prev in thread] [next in thread]