December 2013 - Pharo-users

Re: [Pharo-users] Roassal3D : MessageNotUnderstood:NBMSAAOffscreenDisplay

by Ronie Salgado

Hello, Try with version 118, from Monticello. That problem is produced because I had to made some changes to NBOpenGL, for which I don't have commit privileges so I posted them on the pharo-dev mailing list. Greetings, Ronie Salgado 2013/12/14 Volkert Barr <volkert(a)nivoba.de> > Hello, > > i loaded Roassal3D from the Configuration Browser into a fresh Pharo 3.0 > Image. When selecting "3D IDE" (Roassal3D Easel??) from > the tools-menu i got to the following > error: MessageNotUnderstood:NBMSAAOffscreenDisplay>>bindFramebuffer. > > Probably i forgot to load some other base packages. > > Pharo3.0 > Latest update: #30644 > Host: OS X 10.8.5 > > Any hints? > Volkert > >

Dec. 14, 2013

Re: [Pharo-users] ConnectionClosed issues with Zinc

by StÃ©phane Ducasse

On Dec 13, 2013, at 7:15 PM, Esteban A. Maringolo <emaringolo(a)gmail.com> wrote: > Stef, > > About what exactly? About the server setup or the connection issues? things :) > I plan to write down everything once this goes live and, most > importantly, succeeds. :) Excellent! > > Regards! > > 2013/12/13 StÃ©phane Ducasse <stephane.ducasse(a)inria.fr>: >> esteban >> >> it would be really great if you could extract from your experience a kind of >> blog, book chapter whatever to share it with other developers around the world. >

Dec. 14, 2013

Roassal3D : MessageNotUnderstood:NBMSAAOffscreenDisplay

by Volkert Barr

Hello, i loaded Roassal3D from the Configuration Browser into a fresh Pharo 3.0 Image. When selecting "3D IDE" (Roassal3D Easel??) from the tools-menu i got to the following error: MessageNotUnderstood:NBMSAAOffscreenDisplay>>bindFramebuffer. Probably i forgot to load some other base packages. Pharo3.0 Latest update: #30644 Host: OS X 10.8.5 Any hints? Volkert

Dec. 14, 2013

Re: [Pharo-users] ConnectionClosed issues with Zinc

by Esteban A. Maringolo

Stef, About what exactly? About the server setup or the connection issues? I plan to write down everything once this goes live and, most importantly, succeeds. :) Regards! 2013/12/13 StÃ©phane Ducasse <stephane.ducasse(a)inria.fr>: > esteban > > it would be really great if you could extract from your experience a kind of > blog, book chapter whatever to share it with other developers around the world.

Dec. 13, 2013

Re: [Pharo-users] Pharo on Docker

by Sven Van Caekenberghe

On 13 Dec 2013, at 15:13, Esteban A. Maringolo <emaringolo(a)gmail.com> wrote: > Branch from thread: "ConnectionClosed issues with Zinc" > > 2013/12/13 Sven Van Caekenberghe <sven(a)stfx.eu>: >> On 13 Dec 2013, at 05:01, Esteban A. Maringolo <emaringolo(a)gmail.com> wrote: > > >>> The second time everything went smoother and faster. I'm thinking >>> about setting up a Docker [3] container for Pharo, and then use it in >>> Digital Ocean as well [4]. >> >> Is docker not 64-bit ? I am not sure running a 32-bit Pharo VM inside it will be easy, but maybe it is. > > Can't Pharo run on 64bit host OSes if you install ia32-libs? Of course it can. My question is: does docker allow you to install ia32-libs ? > The VM will sill be 32 bits, with its known limitations. > > But the idea of setting up the environment locally and then deploying > the same environment to a production server sounds really interesting. > "Containerization is the new virtualization" they say. > > Regards! > > Esteban A. Maringolo >

Dec. 13, 2013

Pharo on Docker

by Esteban A. Maringolo

Branch from thread: "ConnectionClosed issues with Zinc" 2013/12/13 Sven Van Caekenberghe <sven(a)stfx.eu>: > On 13 Dec 2013, at 05:01, Esteban A. Maringolo <emaringolo(a)gmail.com> wrote: >> The second time everything went smoother and faster. I'm thinking >> about setting up a Docker [3] container for Pharo, and then use it in >> Digital Ocean as well [4]. > > Is docker not 64-bit ? I am not sure running a 32-bit Pharo VM inside it will be easy, but maybe it is. Can't Pharo run on 64bit host OSes if you install ia32-libs? The VM will sill be 32 bits, with its known limitations. But the idea of setting up the environment locally and then deploying the same environment to a production server sounds really interesting. "Containerization is the new virtualization" they say. Regards! Esteban A. Maringolo

Dec. 13, 2013

Re: [Pharo-users] Glorp PostgreSQL Drivers in Ubuntu 12.04

by Esteban A. Maringolo

2013/12/13 Pierce Ng <pierce(a)samadhiweb.com>: > On Thu, Dec 12, 2013 at 12:53:51AM -0200, Esteban A. Maringolo wrote: > I recommended SpsSplitPasswordStore: > > http://samadhiweb.com/blog/2013.08.11.splitpasswordstore.html Thank for your suggestion. Even when it's security through obscurity I like the approach. However, the pass is not stored in the application. Also, the database server only accepts connections from localhost. Host to which I can only connect by means of RSA keys (no password). So it's a risk I assume for convenience. :) Regards! Esteban A. Maringolo

Dec. 13, 2013

Re: [Pharo-users] ConnectionClosed issues with Zinc

by StÃ©phane Ducasse

esteban it would be really great if you could extract from your experience a kind of blog, book chapter whatever to share it with other developers around the world. Stef On Dec 13, 2013, at 1:36 AM, Esteban A. Maringolo <emaringolo(a)gmail.com> wrote: > I'm having issues with a connection coming from an Android app > connecting both trough an nginx proxy, but also using some REST > testing tools: > > nginx error log: > 2013/12/12 23:23:17 [error] 701#0: *49 upstream prematurely closed > connection while reading response header from upstream, client: > XX.XX.XX.XX, server: app01.mydomain.com, request: "POST > /app-api/v1/auth/login HTTP/1.1", upstream: > "http://127.0.0.1:8080/app-api/v1/auth/login", host: > "app01.mydomain.com:80" > (END) > > ZnZincSeasideAdapter log: > 2013-12-12 23:23:17 897330 D Executing request/response loop > 2013-12-12 23:23:17 897330 I Read a ZnRequest(POST /app-api/v1/auth/login) > 2013-12-12 23:23:17 897330 T POST /app-api/v1/auth/login 200 114B 42ms > 2013-12-12 23:23:17 897330 D ConnectionClosed: Cannot write data while > writing response > 2013-12-12 23:23:17 897330 D Closing stream > > I'm using Basic HTTP Auth in each request. > > But what's weird is that with curl it works perfectly fine. > > Again, I think this is not strictly Zinc related, but it doesn't work > connecting directly to the Zinc server either: > > 2013-12-13 00:17:02 939035 D Executing request/response loop > 2013-12-13 00:17:02 939035 I Read a ZnRequest(POST /app-api/v1/auth/login) > 2013-12-13 00:17:02 939035 T POST /app-api/v1/auth/login 200 114B 12ms > 2013-12-13 00:17:02 939035 D ConnectionClosed: Cannot write data while > writing response > 2013-12-13 00:17:02 939035 D Closing stream > > I don't know if this has to do with the wireless provider doing some > proxying behind the wall (which they do), but given the fact that if I > try using some REST testing tools from my computer and they also fail, > I think it has to do with something else. > > At nginx I'm forwarding the Authorization header to the Zinc Server: > > proxy_set_header Authorization $http_authorization; > proxy_pass_header Authorization; > > Is there a way I can log the complete request headers at the Zinc > side, and if possible also at nginx? > > Regards, > > > Esteban A. Maringolo >

Dec. 13, 2013

MooseDay schedule

by Marcus Denker

>> De : Anne Etien <anne.etien(a)univ-lille1.fr> >> >> Hi, >> >> Please find bellow the schedule of the MooseDay >> http://www.lifl.fr/~etien/MooseDay.html >> The idea is really to present works developed on top of (or around) Moose and to have fruitful discussions. So I planned 30 mn per topic, but the schedule is quite flexible. Lot of time is foreseen for discussions. >> >> 10h: welcome >> 10h30: APIEvolutionMiner - Andre Hora >> 11h: FamixDiff - Nicolas Anquetil >> 11h30: Orion - Jannik Laval and Anne Etien >> >> 12h: Lunch offered by ESUG. >> >> 14h: First works on Version - Yuriy Tymchuk >> 14h30: Roassal - Alexandre Bergel >> >> 15h-15h30 break >> >> 15h30: GraphET - Alexandre Bergel >> 16h: Use of GraphET - Serge Stinckwich >> 16h30: Around tests - Alain Plantec >> >> Reminder, the MooseDay will take place in Inria B building room B31. Directions to come can be found there (http://www.inria.fr/en/centre/lille/overview/locations) >> >> See you next week in Lille ;o) >> >> Anne >>

Dec. 13, 2013

Re: [Pharo-users] ConnectionClosed issues with Zinc

by Sven Van Caekenberghe

On 13 Dec 2013, at 05:01, Esteban A. Maringolo <emaringolo(a)gmail.com> wrote: > Answering myself again (I knowâ¦). Thanks for posting all this, it is useful. > There is some issue in how the Basic Authentication is handled in > certain Android devices [1], which by chance (fortunately) is the same > I use to develop with, so I had to change the way the header was set > in the request, which basically I setup manually [2] > > So, only good news. The server is up and running. > In fact it is the second server, I created a second Droplet at Digital > Ocean because the first had Ubuntu 12.04 LTS and I wanted a newer > release to have the newer version of nignx as well. So Ubuntu 13.10 > 32bit it was. > > The second time everything went smoother and faster. I'm thinking > about setting up a Docker [3] container for Pharo, and then use it in > Digital Ocean as well [4]. Is docker not 64-bit ? I am not sure running a 32-bit Pharo VM inside it will be easy, but maybe it is. > It was a long day, but worth it. > > Best regards, > > Esteban A. Maringolo > > > [1] https://github.com/loopj/android-async-http/issues/262 > [2] http://stackoverflow.com/a/20558752/644918 > [3] http://www.docker.io (anybody using it?) > [4] https://www.digitalocean.com/community/articles/how-to-use-the-digitalocean… > > 2013/12/12 Esteban A. Maringolo <emaringolo(a)gmail.com>: >> Good news, bad news. :) >> >> I decided to test using another great tool bundled with Pharo itself: ZnClient. >> >> >> "Going through nginx" >> ZnClient new >> url: 'http://myserver.com/app-api/v1/auth/login'; >> setBasicAuthenticationUsername: '24603458' password: '1234'; >> logToTranscript; >> post; >> response. >> >> And... as expected... it works! >> >> The same if connect directly to Zinc (:8080). >> >> So there is something wrong in the rest of the tools around... or I >> really don't know. >> >> But the HTTP client library we're using at the Android (Java) side is >> one of the most used (http://loopj.com/android-async-http/) but also >> fails with org.apache.http client. So... it's not a matter of client. >> >> I just need to understand the reason for the "Closed Connection". >> >> >> >> >> Esteban A. Maringolo >> >> >> 2013/12/12 Esteban A. Maringolo <emaringolo(a)gmail.com>: >>> I'm having issues with a connection coming from an Android app >>> connecting both trough an nginx proxy, but also using some REST >>> testing tools: >>> >>> nginx error log: >>> 2013/12/12 23:23:17 [error] 701#0: *49 upstream prematurely closed >>> connection while reading response header from upstream, client: >>> XX.XX.XX.XX, server: app01.mydomain.com, request: "POST >>> /app-api/v1/auth/login HTTP/1.1", upstream: >>> "http://127.0.0.1:8080/app-api/v1/auth/login", host: >>> "app01.mydomain.com:80" >>> (END) >>> >>> ZnZincSeasideAdapter log: >>> 2013-12-12 23:23:17 897330 D Executing request/response loop >>> 2013-12-12 23:23:17 897330 I Read a ZnRequest(POST /app-api/v1/auth/login) >>> 2013-12-12 23:23:17 897330 T POST /app-api/v1/auth/login 200 114B 42ms >>> 2013-12-12 23:23:17 897330 D ConnectionClosed: Cannot write data while >>> writing response >>> 2013-12-12 23:23:17 897330 D Closing stream >>> >>> I'm using Basic HTTP Auth in each request. >>> >>> But what's weird is that with curl it works perfectly fine. >>> >>> Again, I think this is not strictly Zinc related, but it doesn't work >>> connecting directly to the Zinc server either: >>> >>> 2013-12-13 00:17:02 939035 D Executing request/response loop >>> 2013-12-13 00:17:02 939035 I Read a ZnRequest(POST /app-api/v1/auth/login) >>> 2013-12-13 00:17:02 939035 T POST /app-api/v1/auth/login 200 114B 12ms >>> 2013-12-13 00:17:02 939035 D ConnectionClosed: Cannot write data while >>> writing response >>> 2013-12-13 00:17:02 939035 D Closing stream >>> >>> I don't know if this has to do with the wireless provider doing some >>> proxying behind the wall (which they do), but given the fact that if I >>> try using some REST testing tools from my computer and they also fail, >>> I think it has to do with something else. >>> >>> At nginx I'm forwarding the Authorization header to the Zinc Server: >>> >>> proxy_set_header Authorization $http_authorization; >>> proxy_pass_header Authorization; >>> >>> Is there a way I can log the complete request headers at the Zinc >>> side, and if possible also at nginx? >>> >>> Regards, >>> >>> >>> Esteban A. Maringolo >

Dec. 13, 2013

Re: [Pharo-users] Glorp PostgreSQL Drivers in Ubuntu 12.04

by Pierce Ng

On Thu, Dec 12, 2013 at 12:53:51AM -0200, Esteban A. Maringolo wrote: > sampleLogin > ^Login new > database: (PostgreSQLPlatform new characterEncoding: 'utf8'); > username: 'user'; > password: 'secret'; <shamelessplug> I recommended SpsSplitPasswordStore: http://samadhiweb.com/blog/2013.08.11.splitpasswordstore.html ^Login new database: (PostgreSQLPlatform new characterEncoding: 'utf8'); username: 'user'; password: (SpsSplitPasswordStore readFrom: 'user.dat'); ... </shamelessplug> I'm thinking to remove the code that takes random strings from /dev/random and do it in a more portable way without weakening security too much. -- Pierce Ng http://samadhiweb.com/blog/

Dec. 13, 2013

Re: [Pharo-users] ConnectionClosed issues with Zinc

by Esteban A. Maringolo

Answering myself again (I know...). There is some issue in how the Basic Authentication is handled in certain Android devices [1], which by chance (fortunately) is the same I use to develop with, so I had to change the way the header was set in the request, which basically I setup manually [2] So, only good news. The server is up and running. In fact it is the second server, I created a second Droplet at Digital Ocean because the first had Ubuntu 12.04 LTS and I wanted a newer release to have the newer version of nignx as well. So Ubuntu 13.10 32bit it was. The second time everything went smoother and faster. I'm thinking about setting up a Docker [3] container for Pharo, and then use it in Digital Ocean as well [4]. It was a long day, but worth it. Best regards, Esteban A. Maringolo [1] https://github.com/loopj/android-async-http/issues/262 [2] http://stackoverflow.com/a/20558752/644918 [3] http://www.docker.io (anybody using it?) [4] https://www.digitalocean.com/community/articles/how-to-use-the-digitalocean… 2013/12/12 Esteban A. Maringolo <emaringolo(a)gmail.com>: > Good news, bad news. :) > > I decided to test using another great tool bundled with Pharo itself: ZnClient. > > > "Going through nginx" > ZnClient new > url: 'http://myserver.com/app-api/v1/auth/login'; > setBasicAuthenticationUsername: '24603458' password: '1234'; > logToTranscript; > post; > response. > > And... as expected... it works! > > The same if connect directly to Zinc (:8080). > > So there is something wrong in the rest of the tools around... or I > really don't know. > > But the HTTP client library we're using at the Android (Java) side is > one of the most used (http://loopj.com/android-async-http/) but also > fails with org.apache.http client. So... it's not a matter of client. > > I just need to understand the reason for the "Closed Connection". > > > > > Esteban A. Maringolo > > > 2013/12/12 Esteban A. Maringolo <emaringolo(a)gmail.com>: >> I'm having issues with a connection coming from an Android app >> connecting both trough an nginx proxy, but also using some REST >> testing tools: >> >> nginx error log: >> 2013/12/12 23:23:17 [error] 701#0: *49 upstream prematurely closed >> connection while reading response header from upstream, client: >> XX.XX.XX.XX, server: app01.mydomain.com, request: "POST >> /app-api/v1/auth/login HTTP/1.1", upstream: >> "http://127.0.0.1:8080/app-api/v1/auth/login", host: >> "app01.mydomain.com:80" >> (END) >> >> ZnZincSeasideAdapter log: >> 2013-12-12 23:23:17 897330 D Executing request/response loop >> 2013-12-12 23:23:17 897330 I Read a ZnRequest(POST /app-api/v1/auth/login) >> 2013-12-12 23:23:17 897330 T POST /app-api/v1/auth/login 200 114B 42ms >> 2013-12-12 23:23:17 897330 D ConnectionClosed: Cannot write data while >> writing response >> 2013-12-12 23:23:17 897330 D Closing stream >> >> I'm using Basic HTTP Auth in each request. >> >> But what's weird is that with curl it works perfectly fine. >> >> Again, I think this is not strictly Zinc related, but it doesn't work >> connecting directly to the Zinc server either: >> >> 2013-12-13 00:17:02 939035 D Executing request/response loop >> 2013-12-13 00:17:02 939035 I Read a ZnRequest(POST /app-api/v1/auth/login) >> 2013-12-13 00:17:02 939035 T POST /app-api/v1/auth/login 200 114B 12ms >> 2013-12-13 00:17:02 939035 D ConnectionClosed: Cannot write data while >> writing response >> 2013-12-13 00:17:02 939035 D Closing stream >> >> I don't know if this has to do with the wireless provider doing some >> proxying behind the wall (which they do), but given the fact that if I >> try using some REST testing tools from my computer and they also fail, >> I think it has to do with something else. >> >> At nginx I'm forwarding the Authorization header to the Zinc Server: >> >> proxy_set_header Authorization $http_authorization; >> proxy_pass_header Authorization; >> >> Is there a way I can log the complete request headers at the Zinc >> side, and if possible also at nginx? >> >> Regards, >> >> >> Esteban A. Maringolo

Dec. 13, 2013

Re: [Pharo-users] ConnectionClosed issues with Zinc

by Esteban A. Maringolo

Good news, bad news. :) I decided to test using another great tool bundled with Pharo itself: ZnClient. "Going through nginx" ZnClient new url: 'http://myserver.com/app-api/v1/auth/login'; setBasicAuthenticationUsername: '24603458' password: '1234'; logToTranscript; post; response. And... as expected... it works! The same if connect directly to Zinc (:8080). So there is something wrong in the rest of the tools around... or I really don't know. But the HTTP client library we're using at the Android (Java) side is one of the most used (http://loopj.com/android-async-http/) but also fails with org.apache.http client. So... it's not a matter of client. I just need to understand the reason for the "Closed Connection". Esteban A. Maringolo 2013/12/12 Esteban A. Maringolo <emaringolo(a)gmail.com>: > I'm having issues with a connection coming from an Android app > connecting both trough an nginx proxy, but also using some REST > testing tools: > > nginx error log: > 2013/12/12 23:23:17 [error] 701#0: *49 upstream prematurely closed > connection while reading response header from upstream, client: > XX.XX.XX.XX, server: app01.mydomain.com, request: "POST > /app-api/v1/auth/login HTTP/1.1", upstream: > "http://127.0.0.1:8080/app-api/v1/auth/login", host: > "app01.mydomain.com:80" > (END) > > ZnZincSeasideAdapter log: > 2013-12-12 23:23:17 897330 D Executing request/response loop > 2013-12-12 23:23:17 897330 I Read a ZnRequest(POST /app-api/v1/auth/login) > 2013-12-12 23:23:17 897330 T POST /app-api/v1/auth/login 200 114B 42ms > 2013-12-12 23:23:17 897330 D ConnectionClosed: Cannot write data while > writing response > 2013-12-12 23:23:17 897330 D Closing stream > > I'm using Basic HTTP Auth in each request. > > But what's weird is that with curl it works perfectly fine. > > Again, I think this is not strictly Zinc related, but it doesn't work > connecting directly to the Zinc server either: > > 2013-12-13 00:17:02 939035 D Executing request/response loop > 2013-12-13 00:17:02 939035 I Read a ZnRequest(POST /app-api/v1/auth/login) > 2013-12-13 00:17:02 939035 T POST /app-api/v1/auth/login 200 114B 12ms > 2013-12-13 00:17:02 939035 D ConnectionClosed: Cannot write data while > writing response > 2013-12-13 00:17:02 939035 D Closing stream > > I don't know if this has to do with the wireless provider doing some > proxying behind the wall (which they do), but given the fact that if I > try using some REST testing tools from my computer and they also fail, > I think it has to do with something else. > > At nginx I'm forwarding the Authorization header to the Zinc Server: > > proxy_set_header Authorization $http_authorization; > proxy_pass_header Authorization; > > Is there a way I can log the complete request headers at the Zinc > side, and if possible also at nginx? > > Regards, > > > Esteban A. Maringolo

Dec. 13, 2013

Re: [Pharo-users] Is OpenDBX alive?

by Pierce Ng

On Thu, Dec 12, 2013 at 07:01:07PM -0200, Esteban A. Maringolo wrote: > I don't know how good is to have a pure Smalltalk driver either, not > in terms of maintenance*, but in terms of performance, memory, etc. Linking to C libraries using FFI brings its own performance considerations with respect to blocking the VM. Speaking of which, what is the state of the art? Does the Cog+NB combo allow, say, to write an SQLite user-defined function in Smalltalk that is invoked by the C library per row of result? -- Pierce Ng http://samadhiweb.com/blog/

Dec. 13, 2013

ConnectionClosed issues with Zinc

by Esteban A. Maringolo

I'm having issues with a connection coming from an Android app connecting both trough an nginx proxy, but also using some REST testing tools: nginx error log: 2013/12/12 23:23:17 [error] 701#0: *49 upstream prematurely closed connection while reading response header from upstream, client: XX.XX.XX.XX, server: app01.mydomain.com, request: "POST /app-api/v1/auth/login HTTP/1.1", upstream: "http://127.0.0.1:8080/app-api/v1/auth/login", host: "app01.mydomain.com:80" (END) ZnZincSeasideAdapter log: 2013-12-12 23:23:17 897330 D Executing request/response loop 2013-12-12 23:23:17 897330 I Read a ZnRequest(POST /app-api/v1/auth/login) 2013-12-12 23:23:17 897330 T POST /app-api/v1/auth/login 200 114B 42ms 2013-12-12 23:23:17 897330 D ConnectionClosed: Cannot write data while writing response 2013-12-12 23:23:17 897330 D Closing stream I'm using Basic HTTP Auth in each request. But what's weird is that with curl it works perfectly fine. Again, I think this is not strictly Zinc related, but it doesn't work connecting directly to the Zinc server either: 2013-12-13 00:17:02 939035 D Executing request/response loop 2013-12-13 00:17:02 939035 I Read a ZnRequest(POST /app-api/v1/auth/login) 2013-12-13 00:17:02 939035 T POST /app-api/v1/auth/login 200 114B 12ms 2013-12-13 00:17:02 939035 D ConnectionClosed: Cannot write data while writing response 2013-12-13 00:17:02 939035 D Closing stream I don't know if this has to do with the wireless provider doing some proxying behind the wall (which they do), but given the fact that if I try using some REST testing tools from my computer and they also fail, I think it has to do with something else. At nginx I'm forwarding the Authorization header to the Zinc Server: proxy_set_header Authorization $http_authorization; proxy_pass_header Authorization; Is there a way I can log the complete request headers at the Zinc side, and if possible also at nginx? Regards, Esteban A. Maringolo

Dec. 13, 2013

Re: [Pharo-users] Is OpenDBX alive?

by Esteban A. Maringolo

I know it's not extinct, its webpage even has news from June this year. :) But what I mean is that it "didn't catch" the audience that maybe was expected. I don't know how good is to have a pure Smalltalk driver either, not in terms of maintenance*, but in terms of performance, memory, etc. I'm happy to have it, and until something replaces it, it's better to keep it maintaned, but I was thinking in the long term. Regards! Esteban A. Maringolo 2013/12/12 Guillermo Polito <guillermopolito(a)gmail.com>: > Well, first, I don't know the original motivation for opendbx... Maybe > Norbert wanted to (sort of) decouple his C application from a specific > driver. The thing is that it's not simple to track its users. > > Then, alive as "it compiles and it works in latest versions of OSs", it is > :). I did the whole compilation for the three platforms and drivers about a > month ago. > > On the other side, alive as "it is adding new features from time to time", > it's not. Maybe because database drivers do not evolve also too much, nor > they change their API. The latest changes, from about a year ago were some > bugfixes to ODBC and fixing some compilation stuff for windowze. > > Following EstebanL's comment, we would like to get completely rid of it, but > so far it's useful to connect to databases that otherwise we couldn't, like > oracle. For other vendors, such as Mysql or PGsql, there are the pure > smalltalk versions of the drivers, which are probably better from the > maintenance point of view. > > Guille > > > > On Thu, Dec 12, 2013 at 3:35 PM, Esteban Lorenzano <estebanlm(a)gmail.com> > wrote: >> >> is not incredibly used, but it is not completely dead :) >> >> anyway... one of our many many projects is to drop opendbx (by moving that >> layer to pharo itself)... but that's for the future (probably far future). >> >> Esteban >> >> >> >> >> >> >> On Thu, Dec 12, 2013 at 2:30 PM, Esteban A. Maringolo >> <emaringolo(a)gmail.com> wrote: >>> >>> As a side question about my dependency issues... >>> >>> Is the OpenDBX project still active? Not its Smalltalk side, I mean... >>> the core project. >>> >>> To me it seems like it was a great attempt to have something similar >>> to ODBC but for all platforms. >>> But I'm not seeing a wide usage of it in the wild. >>> >>> Seems like everybody else is happy with their JDBC, DBI and similar >>> "DB layers" on the language side. >>> >>> >>> Regards, >>> >>> Esteban A. Maringolo >>> >> >

Dec. 12, 2013

Re: [Pharo-users] Glorp PostgreSQL Drivers in Ubuntu 12.04

by phil＠highoctane.be

I loaded that one a while ago. Works. --- Philippe Back Dramatic Performance Improvements Mob: +32(0) 478 650 140 | Fax: +32 (0) 70 408 027 Mail:phil@highoctane.be | Web: http://philippeback.eu Blog: http://philippeback.be | Twitter: @philippeback Youtube: http://www.youtube.com/user/philippeback/videos High Octane SPRL rue cour Boisacq 101 | 1301 Bierges | Belgium Pharo Consortium Member - http://consortium.pharo.org/ Featured on the Software Process and Measurement Cast - http://spamcast.libsyn.com Sparx Systems Enterprise Architect and Ability Engineering EADocX Value Added Reseller On Thu, Dec 12, 2013 at 4:31 PM, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > > On 12 Dec 2013, at 16:24, Esteban A. Maringolo <emaringolo(a)gmail.com> > wrote: > > > 2013/12/12 Sven Van Caekenberghe <sven(a)stfx.eu>: > >> The class NativePostgresDriver is in a package called > GlorpDriverPostgreSQL. > >> It seems to have come from > >> MCSqueaksourceRepository > >> location: 'http://squeaksource.com/PostgresV2' > >> user: '' > >> password: â' > > > > SqueakSource... how quaint! ;-) > > Yeah, PostgresV2 exists on StHub as well, with some updates from Mariano > it seems. > > > Thanks Sven. It works now. Your pragmatism outsmarts most of us :D > > ;-) > > > Best regards! > > > > > > Esteban A. Maringolo > > > > > >

Dec. 12, 2013

Re: [Pharo-users] Is OpenDBX alive?

by Guillermo Polito

Well, first, I don't know the original motivation for opendbx... Maybe Norbert wanted to (sort of) decouple his C application from a specific driver. The thing is that it's not simple to track its users. Then, alive as "it compiles and it works in latest versions of OSs", it is :). I did the whole compilation for the three platforms and drivers about a month ago. On the other side, alive as "it is adding new features from time to time", it's not. Maybe because database drivers do not evolve also too much, nor they change their API. The latest changes, from about a year ago were some bugfixes to ODBC and fixing some compilation stuff for windowze. Following EstebanL's comment, we would like to get completely rid of it, but so far it's useful to connect to databases that otherwise we couldn't, like oracle. For other vendors, such as Mysql or PGsql, there are the pure smalltalk versions of the drivers, which are probably better from the maintenance point of view. Guille On Thu, Dec 12, 2013 at 3:35 PM, Esteban Lorenzano <estebanlm(a)gmail.com>wrote: > is not incredibly used, but it is not completely dead :) > > anyway... one of our many many projects is to drop opendbx (by moving that > layer to pharo itself)... but that's for the future (probably far future). > > Esteban > > > > > > > On Thu, Dec 12, 2013 at 2:30 PM, Esteban A. Maringolo < > emaringolo(a)gmail.com> wrote: > >> As a side question about my dependency issues... >> >> Is the OpenDBX project still active? Not its Smalltalk side, I mean... >> the core project. >> >> To me it seems like it was a great attempt to have something similar >> to ODBC but for all platforms. >> But I'm not seeing a wide usage of it in the wild. >> >> Seems like everybody else is happy with their JDBC, DBI and similar >> "DB layers" on the language side. >> >> >> Regards, >> >> Esteban A. Maringolo >> >> >

Dec. 12, 2013

Nginx Reverse Proxying

by Esteban A. Maringolo

Hi there, I continue with my server setup... I'm trying to setup nginx as the HTTP front end. I have two servers, one for the Zinc SeasideAdaptor running on 8080 and another server running on 8081 (maybe there is a way to run both on the same server). ZnZincServerAdaptor startOn: 8080. ZnServer startDefaultOn: 8081. ZnServer default delegate: (ZnWebSocketDelegate map: 'ws' to: MyWsHandler new). In the Seaside side I have three "entry points": /app : which is a standard Seaside application /app-api/v1/ :which is a WARestfulHandler (REST API, with Basic HTTP Auth) /assets (File Directory for static files) I plan to put a nginx in front of those services, mainly for performance purpose of static file serving, and to avoid the exposure of my server directy to internet. Supposing the hostname is app01.mydomain.com ? My requirements are 1) I want that http://app01.mydomain.com/app be forwarded to http://localhost:8080/app, the same with app-api 2) Same for websockets, maybe to http://ws01.mydomain.com to http://localhost:8081/ws 3) http://app01.mydomain.com/assets be served statically. I know this is a request which isn't directly related with Pharo nor Seaside, but given the fact it's a common setup I hope somebody will help me with this. Sorry for the crosspost. Regards, Esteban A. Maringolo

Dec. 12, 2013

Re: [Pharo-users] [launcher] fixed layout of tollbar buttons

by btc＠openinworld.com

Dec. 12, 2013

Re: [Pharo-users] Glorp PostgreSQL Drivers in Ubuntu 12.04

by Sven Van Caekenberghe

On 12 Dec 2013, at 16:24, Esteban A. Maringolo <emaringolo(a)gmail.com> wrote: > 2013/12/12 Sven Van Caekenberghe <sven(a)stfx.eu>: >> The class NativePostgresDriver is in a package called GlorpDriverPostgreSQL. >> It seems to have come from >> MCSqueaksourceRepository >> location: 'http://squeaksource.com/PostgresV2' >> user: '' >> password: â' > > SqueakSource... how quaint! ;-) Yeah, PostgresV2 exists on StHub as well, with some updates from Mariano it seems. > Thanks Sven. It works now. Your pragmatism outsmarts most of us :D ;-) > Best regards! > > > Esteban A. Maringolo >

Dec. 12, 2013

Re: [Pharo-users] Glorp PostgreSQL Drivers in Ubuntu 12.04

by Esteban A. Maringolo

2013/12/12 Sven Van Caekenberghe <sven(a)stfx.eu>: > The class NativePostgresDriver is in a package called GlorpDriverPostgreSQL. > It seems to have come from > MCSqueaksourceRepository > location: 'http://squeaksource.com/PostgresV2' > user: '' > password: â' SqueakSource... how quaint! ;-) Thanks Sven. It works now. Your pragmatism outsmarts most of us :D Best regards! Esteban A. Maringolo

Dec. 12, 2013

Re: [Pharo-users] [launcher] fixed layout of tollbar buttons

by Benjamin

I was referencing to the LaTeX command \hfill http://en.wikibooks.org/wiki/TeX/hfill Ben On 12 Dec 2013, at 14:48, btc(a)openinworld.com wrote: > Thanks Ben. But why is it required? I think I am missing some basic info how layouts work on the Morph side, and the few references to #hFill in the image don't tell me much. > > btw, adding a method #hFill would make it look like less of a hack. > > cheers -ben > > Benjamin wrote: >> >> The empty column is used as a \hFill :) >> >> Ben >> >> On 12 Dec 2013, at 13:57, btc(a)openinworld.com wrote: >> >>> Hi Damien, Benjamin.V.R. >>> >>> I see in "PharoLauncher-Spec-DamienCassou.17" that the layout of the toolbar buttons is fixed, which is really great. However in trying to learn from that for future reference, seeing it relies on an empty column being added "seems" like a hack to workaround some issue with Spec. Should that be logged as a bug? Or is there some greater architectural background for this that I am missing? >>> >>> PhLCommandTooldbar>>dynamicLayout "(newer)" >>> "Create a spec layout with 1 row, and as many columns as there are command classes" >>> ^ SpecLayout composed newRow: >>> [ :r | >>> r newColumn: [ :col | ]. "<--------FIX" >>> self commandClasses do: >>> [ :commandClass | >>> r add: commandClass specId width: self class toolbarHeight >>> ] >>> ] >>> >>> >>> I tried digging into Spec to look at this myself, and my naive intuition felt it was something to do with initialization of LayoutFrames, but that was my first exposure to LayoutFrame and I quickly got lost. However I was reminded of Igor commenting "i only know that if you using proportional layout, then if you want submorphs to be laid down properly, you must use addMorph:fullFrame: instead of just addMorph:" [1] I tried this with a few hacks, but tit was really shooting in the dark and unsuccessful. >>> >>> [1] https://www.mail-archive.com/pharo-dev@lists.pharo.org/msg08838.html >>> >>> cheers -ben >>> <PharoLauncher-fixed-button-layout.png> >> >

Dec. 12, 2013

Re: [Pharo-users] Is OpenDBX alive?

by Esteban Lorenzano

is not incredibly used, but it is not completely dead :) anyway... one of our many many projects is to drop opendbx (by moving that layer to pharo itself)... but that's for the future (probably far future). Esteban On Thu, Dec 12, 2013 at 2:30 PM, Esteban A. Maringolo <emaringolo(a)gmail.com>wrote: > As a side question about my dependency issues... > > Is the OpenDBX project still active? Not its Smalltalk side, I mean... > the core project. > > To me it seems like it was a great attempt to have something similar > to ODBC but for all platforms. > But I'm not seeing a wide usage of it in the wild. > > Seems like everybody else is happy with their JDBC, DBI and similar > "DB layers" on the language side. > > > Regards, > > Esteban A. Maringolo > >

Dec. 12, 2013

Re: [Pharo-users] [launcher] fixed layout of tollbar buttons

by btc＠openinworld.com

Dec. 12, 2013