On 11/12/2013 08:33 AM, Martin Schreiber wrote: > > The C-like syntax > " > FIELDNAME ':' FIELDTYPE [':' BITSIZE] ';' > " > is accepted? Seems useful and nicely readable for me (of course I do have a strong C background). > Is a record with BITSIZE automatically bitpacked? I gather this does make sense.
> Defining of little/big-endian could be useful too. > The position of the bits needs to be documented. With "low endian" archs c starts with the lowest order bits. This seems to make sense for mseLang, as well, even though when transporting data (e.g. via TCP/IP") the usual definition is "high endian" (aka high Byte first, high bit first). I did get burned on than behalf several times. For this an alternate "high endian bit packed" keyword might come handy. Moreover: If it's really bit packed you might get numerical values that require rather complex handling by the compiler: record tr x1: int64: 2; x2: int64: 63; end var t1, t2: tr t1 := t1.x2 + t2.x2; A 32 bit compiler would need to load 64 bit values that cover three 32 words, shift them into a double register and mask out the upper bit before doing the add. Then it needs to shift, mask and "or" the stuff for storing. Nice :-) -Michael ------------------------------------------------------------------------------ November Webinars for C, C++, Fortran Developers Accelerate application performance with scalable programming models. Explore techniques for threading, error checking, porting, and tuning. Get the most from the latest Intel processors and coprocessors. See abstracts and register http://pubads.g.doubleclick.net/gampad/clk?id=60136231&iu=/4140/ostg.clktrk _______________________________________________ mseide-msegui-talk mailing list mseide-msegui-talk@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/mseide-msegui-talk
