Vents are welcome. I definitely agree with "creating a library, documenting it, publishing it should be drop dead simple and just work. And it probably does, for simple things, if you're careful" and I think we need to work on this step. BUT, are you sure you want to publish the library on PLaneT not as a package?
When/if you figure out what our 40% share of problems are, please post them here. We would very much appreciate it. Thanks -- Matthias On May 27, 2013, at 8:34 PM, Ray Racine wrote: > Well today was I'd say my was third attempt to refactor an accumulation of > code into some sort of structure, create documentation, clean raco setup and > Planet publication. Third time fail. And fail badly. > > Granted say +60% is on me, but ... some stuff should just work. > > In my opinion the combinatorial explosion of Racket features intersecting > with a lacking in modularity pile of functionality results in robustness > failure where facets are coming together.[1] > > Creating a library, documenting it, publishing it should be drop dead simple > and just work. And it probably does, for simple things, if you're careful. > Probably. > > So ... I'm devoting the rest of the week, all of it, to isolating and > submitting core bread and butter bugs that are stymieing that goal until the > process works smoothly, without kludges, hacks, compromises and work-arounds. > > I'm going to start with hunting down one that's been happening for long while > now, what are the conditions that cause a simple 'require' to fail and why > for the love of god does the compiler refuse to tell you where the require > was and which module was requiring it? > > I love haystack diving and recursive grepping 10,000 lines of code in search > of what causes a simple require resolve error as much as anyone, but > sometimes you just want the darn compiler to cut you some slack and tell you > a) the error specifics in an intelligible manner b) the source line that > caused it, thank you very much. > > Don't get me wrong there is so much amazing stuff in Racket. I can't think > of a single substantive programming paradigm not contained in Racket or any > aspect of the development process not attempted by Racket, all engineered to > the nth ... yet, I think the philosopher Bilbo Baggins said it best, “Racket > feels thin, sort of stretched, like butter scraped over too much bread.” > > Sorry for the vent ... > > Ray > > [1] http://en.wikipedia.org/wiki/The_Mythical_Man-Month#Conceptual_integrity > > > > > > ____________________ > Racket Users list: > http://lists.racket-lang.org/users
____________________ Racket Users list: http://lists.racket-lang.org/users
