Nilson Santos Figueiredo Junior

• About: I enjoy Perl.

Recent Actions

• Commented on So apparently this is creepy
  I'm not an American, but I would never consider it creepy. I'd consider it pointless, but hey, that's what Acme::* is for, right? Anyway, as a foreigner, it seems like people in the US are constantly freaked out about anything...
• Commented on A distributed design challenge
  Can't you reserve the budget for a bid whenever you make one? Then servers would decide whether to bid or not based on this available budget. If a bid isn't successful you free the reserved amount for that bid, if...
• Commented on Indexing MooseX::Declare modules
  The simple solution right now would be to also look for "class XXXXX {}" statements when looking for "package XXXXX" fails. Sure, if someone comes up with another syntax, it would still be an issue, but at least this will...
• Commented on Perl, Perl 5, Perl 6, and names
  I also agree that Perl 6 should be called something else. For us, inside the Perl community, it's easy to distinguish them. For everyone else, it's a mess. Also, nowadays, people only seem to notice major version increases. Firefox has...
• Commented on Cacheing Plack Proxy - speed up development
  Thanks a lot for the LWP::Protocol::PSGI tip. I just came up with a transparent Twitter API caching solution by combining Plack::App::Proxy, Plack::Middleware::Cache and LWP::Protocol::PSGI. Awesome....
• Commented on Cacheing Plack Proxy - speed up development
  Very clever. :)...
• Commented on Introducing metacpan.org - A better search for the CPAN
  There's something I've always wanted on CPAN search and which also seems to be missing on the new metacpan.org: sorting by last release date. This would be awesome to have. Often, the most recent module ends up being the best...
• Commented on The October that will never be
  The same sort of thing happens from time to time on Brazilian Perl mailing lists. I guess it's unfortunately unavoidable and, yes, I think it does significantly impact Perl's general acceptance. Personally, I don't get offended, but not everyone is...
• Commented on YAPC::Europe 2010 - Thoughts Pt 2/3 - Promoting A YAPC
  Just a quick fix, there's a typo in your post. It's: http://www.yapcbrasil.org.br/2010 (Brasil vs Brazil)...
• Commented on SQL scripts? Just do it.
  Maybe you should try DBIx::MultiStatementDo....
• Commented on Grepping exact values
  I guess ~~ is really all you need. If you're just testing in boolean context. However you may want to count occurrences so I still think a "smart grep" would be useful....
• Commented on Grepping exact values
  Maybe there should be a "smartgrep" or "sgrep" in Perl 5? It would work just like grep, but use smart matching instead. Probably doable right now with Devel::Declare....
• Commented on L. Peter Deutsch hates Perl
  Well, I've never seen anyone with anti-Perl feelings before 2000 or so as I've been programming in Perl since 2000. I chose Perl as a language back then as I wanted to do some web development and everyone basically recommended...
• Commented on Adding gettext support to Git
  I actually just installed it on another Windows box (also with Strawberry Perl). So maybe the issue is a bit more complicated than that. Maybe I just have two identically broken environments (for some reason). sigh...
• Commented on Marpa is now O(n) for Right Recursions
  Very nice, I'll keep an eye on your project. :)...
• Commented on Marpa is now O(n) for Right Recursions
  It's very nice to see this sort of cutting edge "academic" work being done in Perl. I was interested in Marpa when I first heard about it and I'll definitely take a look at it again in the future for...
• Commented on ffi_call( 'Wolf!' )
  I think for Win32 you would want to either: a) bundle your external library dependencies or; b) setup a Alien::libffi or something like that. First choice is probably more convenient for users. I think some of the wrappers for external...
• Commented on Adding gettext support to Git
  Actually, the module compiles on Strawberry Perl. The problem occurs when testing: it fails (badly) with a Perl crash. So just keep in mind that, for now, whenever you use anything from libintl-perl, you won't be writing cross-platform code....
• Commented on Adding gettext support to Git
  One of the reasons could be that, unfortunately, even using Strawberry Perl (which is compiled using GCC), libintl-perl can't be installed on Windows. Perl code is generally supposed to be cross-platform....
• Commented on Hello, World. /cliché
  Just one quick tip. The more modern and recommended way of opening files is using 3-argument open() with a lexical filehandle, like this: my $success = open my $log, '>>', 'logfile';...
• Commented on What my experienced business manager would ask about Perl
  Well, I guess that in a situation like this, the best thing you could do is actually giving an answer to each one of these concerns in a way that the business manager would understand. I'll try to answer each...
• Commented on L. Peter Deutsch hates Perl
  According to this page: http://www.codersatwork.com/l-peter-deutsch.html The last time the guy wrote real code was in 2002. So, come on, we can't really blame him for thinking that Perl was a steaming pile of garbage back then. The large majority of...
• Commented on Introduction and API spec
  I just looked at the Python docs for ctypes and the API seems extremely sane. I would suggest maybe you should just use the Python API. It seems very "dynamic" and easy to use. At this point, it probably has...
• Commented on Measuring Rakudo performance
  Are these benchmark results available anywhere? Is there any sort of comparison between Rakudo and, e.g., Perl 5?...
• Commented on Microsoft Attack CPAN Testers (Again)
  That's one additional reason to boycott Bing. Even though Google is apparently losing its focus on search (with that hideous sidebar) they're still the most ethical ones around....
• Commented on perlmv: Renaming files with Perl code
  It seems very useful. Unfortunately, I tried installing it on Windows and the tests failed, so I couldn't try it myself. I could send you the detailed failure log in case you're interested in trying to fix it....
• Commented on Things I Can't Have
  This isn't exclusive to code formatting. People only seem to trust automatic tools (in general) if they make no mistakes. That's why we're not going to see driverless cars for a while. I do think, though, that forcing a code...
• Commented on threads::lite and the chameneoses
  Maybe you should try submitting your entry if it's 3 times faster than the original one? Or did you mean it's 3 times faster than the 6 times slower version? As you've said, it's not like these benchmarks show real-world...
• Commented on Is cpanminus the future?
  I just tested it again and it works fine. I guess I probably made an unrelated mistake when I first tested it and assumed it was cpanm's fault....
• Commented on Is cpanminus the future?
  Miyagawa: well, I basically set HTTP_PROXY and the regular CPAN client worked while cpanm didn't work. I just assumed it was something unsupported by cpanm. I can perform additional tests on Monday....

Comment Threads

• Randy Stauner commented on Indexing MooseX::Declare modules

  My previous comment is probably more influenced by the goofiness of common::sense (mentioned in David Cantrell's post) than by MooseX::Declare (of which I am not familiar)...

  I am not against new syntaxes or new ways of doing things, but if it were my dist I'd fill in the META 'provides' to make sure it got indexed properly, knowing that what I was doing was unusual and it was my responsibility to package my dist correctly.

  For the record, I *do* fill in META provides (automatically using Dist::Zilla::Plugin::MetaProvides).

  Is there a reason the META file (and 'provid…

• djzort.myopenid.com commented on A distributed design challenge

  my 2-5c

  Assuming the servers are physical servers, you could move the clocking source off of the motherboard clocks and on to a single external clock for all the servers. This would be connected via serial (Symmetricom makes such a device, i dont work for them) and the servers would need to drop themselves out of the load balancer if they lose the clocking source.

  With all the servers on the same clocking source, they will all have the same time exactly.

  Another option is to use a 'layer7' loadbalancer, aka reverse proxy. The software isnt important, the function thou…

• Steven Haryanto commented on So apparently this is creepy

  Hi Sawyer X,

  I'd be interested to read your take on the matter.
  

• Sawyer X commented on So apparently this is creepy

  I personally can attest to the importance of anonymity. I don't have a Facebook, Google Plus, LinkedIn or Twitter and I use a moniker. Anonymity provides us with a way to work with others without being afraid of being public. It lets us be in public while keeping our privacy, and that's something very important to understand.

  Some groups are attacked more than others. For example, gays are attacked more than Russians. While I can see a Russian not wanting to be exposed, it's clearly understandable if a person at work won't want to reveal they're gay.

  I don't know when's the l…

• Abigail commented on A distributed design challenge

  Can't you just say campaign $C is handled by server hash($C)%36, for some appropriate hash function?

  Or you could throw in the amount of money left on the budget, and if $N units are left, you allow min($N, 36) servers to bid. Which means that if your budget is almost gone, just one server bids, but if hardly anything is gone, all servers are allowed to bid.

  Third option, if a campaign $C has $N units left in its budget, all servers flip a 36 sided die. Any server that rolls less than $N is allowed to bid. Any server that doesn't isn't allowed to bid in this campaign again. …