Russel Winder Posted in reply to Paulo Pinto signature.asc (This is a digitally signed message part) Attachments:



Permalink Reply

On Thu, 2013-07-04 at 11:35 +0200, Paulo Pinto wrote: […] > Every time I see the usual shell and vi/emacs combo it brings me back memories of when my home computer was a Timex 2068. > > Or as someone puts it, > > http:// andrewbro okins.com/ tech/ one-year- later- an-epic- review-of- pycharm- 2-7-from- a-vim- users- perspective/ Every time I see the usual shell and vi/emacs combo it brings me back memories of when my home computer was a Timex 2068.Or as someone puts it, > > And neither of these groups worry about test coverage that much. > > To be honest in the fortune 500 companies very few do. We always > try to push for it in our consulting projects, but not all > customers > buy into it, specially if the project requires interaction with > the > in-house developers. Perhaps stock prices should be linked to code coverage statistics for that company. It is about as good a measure as the gamblers, sorry traders, use – which seems to rely most on whether the CEO sneezed at the time the traders specified they should. > I think Microsoft plays a big role here. By sponsoring OCaml, Haskel and F# development, as well as, introducing FP to the enterprise via LINQ. Agreed, and I hate to give Microsoft credit for anything positive, but here they definitely have. Sad they dropped IronPython and IronRuby though. > The main issue is that this is very human related. > > Sometimes you really need a few generations to make people adopt new ideas. Not if ideas are reified in the programming language. Java brought shared memory multi-threading front and centre (or should that be center?) and it remains the biggest barrier to quality software we have. Programmers see the feature and use it despite the fact that it had been known for 20 years prior to 1995 that it was the wrong way forward for applications development. Actors, dataflow, CSP, data parallelism, masses of great high level models for concurrency and parallelism all ignore. And now C++11 nearly did almost the same thing. At least thanks to the UK vote C++ includes asynchronous function call and futures. All the rest of C++ threads stuff, like Java, is already 40 years behind well tried and trusted approaches. -- Russel. ========== ========== ========== ========= ========== ========= ========== ========= Dr Russel Winder t: +44 20 7585 2200 voip: sip:russel.winder@ekiga.net 41 Buckmaster Road m: +44 7770 465 077 xmpp: russel@winder.org.uk London SW11 1EN, UK w: www.russel.org.uk skype: russel_winder