Innovation and Open Source

Here’s the list of the cool things that happened at TupiLabs since last Sunday.

• Continuous Delivery support for Speak Like A Brazilian
• Started mother’s week at Speak Like A Brazilian
• A Tupi is one of the authors in this month’s MundoJ magazine, one of the top Java magazines in Brazil
• Migrated TupiLabs.com from Wordpress to PieCrust, a static website generator in PHP. Freeing about 400MB of memory (MySQL and http server)

We are working for you

Have a great week! :-)

Here’s the list of the cool things that happened at TupiLabs since last Sunday.

• Working on a hotel system. Stay tuned.
• Moved our servers to a VPS
• Updated our DNS
• Created builds.tupilabs.com to host TupiLabs projects and other OSS projects

We are working for you

Have a great week! :-)

Here’s the list of the cool things that happened at TupiLabs since last Sunday.

Probably you’ll notice we couldn’t write much code in the last weeks :-( that’s due to infrastructure changes, and meetings with project teams and clients.

• Released Speak Like A Brazilian Quiz application
• Released Speak Like A Brazilian Sentences application

We are working for you

Have a great week! :-)

Check this message in Nutch-dev mailing list archive, where Tejas Patil picked issues for beginners that are looking to contribute to Apache Nutch.

Apache Nutch is an open source web-search software project. Nutch is a project of the Apache Software Foundation and is part of the larger Apache community of developers and users

Aos poucos empresas estão começando a adotar git no lugar do Subversion e do CVS (yuck). Com o git o merge dos projetos ficou mais fácil e limpo (principalmente com cherry pick e rebase). Além da adoção da maioria dos projetos e desenvolvedores Open Source do GitHub para criação de repositórios públicos de código-fonte.

Recentemente fui em uma reunião em uma empresa em São Paulo e conheci o Nilton. Ele trabalha com infra, mas suas atividades as vezes parecem chegar bem próximas às de desenvolvedores e testadores. Ele mantém o site LinuxPro e publica sempre o que conteúdo atualizado e simples de entender sobre Linux, infra e outros assuntos geek :D

Em um dos últimos posts ele fez um ótimo screen cast sobre como instalar o GitLab, uma ferramenta que, além de instalar um servidor git, ainda possui interface Web muito semelhante à do GitHub. A instalação é bem fácil, apesar que tive problemas ao instalar o GitLab em uma rede sem acesso SSH externo (algumas gems usadas pela GitLab precisavam clonar repositórios do GitHub via SSH).

Quando mencionei pro Nilton o problema que tive ele ainda lembrou que há na BitNami uma versão ainda mais fácil de rodar já com tudo pronto. Então, o que está esperando? Adiciona o LinuxPro nos seus favoritos e bora aprender mais git e mostrar pros desenvolvedores do seu time.

TupiLabs website uses Wordpress as backing CMS and blog, and here we have a modified Roots theme with bunch of plug-ins and settings. The pagination of the posts is done with WP-Pagenavi plug-in. When we received a message from José saying that the pagination was broken, we thought it had something to do with WPML, a translation plug-in, used to maintain both languages (en and pt_BR) in the site.

Later we figured out what was causing this error. The error happened only in pages that list categories (blog, news and ideas). In these pages, the function query_posts was being used, and after googling a while we found out that there was an extra argument for pagination.

queryposts(array('categoryname' => 'blog'));

Changing the line above to:

queryposts(array('categoryname' => 'blog', 'paged' => getqueryvar('page')));

Fixed the issue. We also use plug-ins for cache, so we had to purge the pages from the cache in order to see the results. Hope that helps you, in case you have similar problem. Cheers.

From https://groups.google.com/forum/?fromgroups=#!topic/stacks-users/2HwY2dDyeSM.

Stacks processes RAD DNA sequences and produces output, that is displayed on the Web with a bundled PHP+MySQL application. This application is quite useful, especially since it helps developers to analyze large amounts of data.

One limitation though, is that some management tasks may need manual execution of SQL’s in the database. Someone could add few scripts or new features to the Web interface (like a management section?). This could help researchers on their research.

Já existem projetos semelhantes, porém seria interessante uma solução um pouco modular, onde usuários pudessem escolher dicionários diferentes, bem como algoritmos também.

Exemplo: #caiunarede seria segmentado em [‘caiu’, ‘na’, ‘rede’].

Links uteis:

http://www.usna.edu/Users/cs/nchamber/courses/nlp/f12/labs/lab1.html

http://trec.nist.gov/pubs/trec20/papers/NEMISISTICNR.microblog.update.pdf

http://www.cs.colorado.edu/~martin/csci5832/assignments_2/assignment-2-optional-proba.html

http://dl.acm.org/citation.cfm?id=2398410

This idea happened after Andrew Gray’s objectified-solutions. With Objectified Solutions one can extract meta information from a .NET solution. This information includes:

• Global properties of the solution
• List of projects in solution
• Global properties of each project
• List of source files in each project
• List of System references that each project depends on
• List of Project references that each project depends on
• Basic properties of each source file

The code is hosted at GitHub too. Maybe someone could even think about a “.NET solution project” job type for Jenkins.