Check out: NIST Digital Library of Mathematical Functions (companion to the NIST Handbook of Mathematical Functions)
The Return of OS/2?
Is IBM considering an OS/2 redo?
Should IBM consider a return of OS/2? A systems integrator close to the computing giant swears there’s a move afoot inside IBM to reintroduce the operating system — revered by some, reviled by many –before it faded out more than 10 years ago. He has heard this from inside Big Blue itself.
This will mean nothing to young’uns who do not remember the great OS war waged between IBM and Microsoft after their joint development of OS/2 fell apart. Running on both PCs and servers, OS/2 was object-oriented and CORBA-compliant, could multitask and sported a neat graphical user interface. People liked the OO Workplace Shell GUI very much.
…
According to the integrator, IBM’s nascent plan is to repurpose OS/2 “services” atop a Linux core.
The obvious question is why? And the answer is despite Windows’ spanking, IBM shops still run quite a bit of OS/2 — not that they’ve advertised that fact. In theory these companies can drop this Linux-OS/2 amalgam in to replace aging installations with minimal disruption.
I thought Workplace Shell was the best thing about OS/2, otherwise my interest in OS/2 at this point is pretty minimal.
OpenDDS 2.1.2
(via OpenDDS)
(23 March 2010) We are pleased to announce the release of OpenDDS version 2.1.2!
- This is a bugfix release to account for issues introduced by the previous version.
- Various bugfixes for locking issues during sample removal.
- Fixed a bug in obtain_buffer() that did not return the TIMEOUT error code when max_blocking_time is 0 or a very small period.
- Fixed a deadlock in the new IP multicast transport caused when data is received while a reliable session is handshaking.
- Fixed an issue which caused spurious synch threads to be created for connectionless transports (i.e. UDP/IP, IP multicast).
- Enhanced the experimental UDP/IP transport to support multiplexing samples to unique endpoints rather than unique entities.
See the Release Notes for details.
OpenDDS 2.1.1
(via OpenDDS)
(8 March 2010) We are pleased to announce the release of OpenDDS version 2.1.1!
- This is a bugfix release to account for issues introduced by the previous version.
- Various bugfixes have been applied to the new IP multicast transport and the Extensible Transport Framework (ETF) to improve stability.
- Added IDL for ContentFilteredTopic and MultiTopic, but they are not yet implemented (operations will return NULL or RETCODE_UNSUPPORTED).
- A Wireshark 1.2.x compatible dissector is included in this release. See $DDS_ROOT/tools/odds_dissector/README for details.
See the Release Notes for details.
OpenDDS 2.1
(via OpenDDS)
(5 February 2010) We are pleased to announce the release of OpenDDS version 2.1!
- The OpenDDS-Bench performance testing framework has been enhanced to include support for user-submitted performance results.
- A new monitor library introduces an instrumentation mechanism for reporting OpenDDS runtime information.
- A new instrumentation application is available to view executing OpenDDS service Entities.
- A new IP multicast transport is available which replaces the older ReliableMulticast and SimpleMcast transports.
- A new UDP/IP transport is available which replaces the older SimpleUdp transport.
- Numerous bugs in the Extensible Transport Framework (ETF) have been addressed to improve the stability of new and existing transports.
See the Release Notes for details.
It Proved Just Too Difficult
(via an email discussion)
VC8 compiler: definition mis-interpreted as a function declaration
Posted by Microsoft on 2/6/2007 at 10:06 AM
Basically this is a known problem with Visual C++ – we just don’t handle this style of initialization when pointers are involved. I have looked at this issue a couple of times and have always given up as it proved just too difficult to shoehorn this feature into our compiler.So when we looked at this bug in triage yesterday we decided that it didn’t meet the bar – mostly because of the difficulty and also because there is a workaround – use the ‘=’ form of initialization (which also, in my opinion, has the advantage of being clearer).
Jonathan Caves
Visual C++ Compiler Team
It must be nice to be able to use the “too difficult” excuse…
Bridge Simulators
I used to develop software for shiphandling simulators:
Continue reading ‘Bridge Simulators’ »