seaside-dev April 2009

seaside-dev@lists.squeakfoundation.org

11 participants
13 discussions

Re: [Seaside Commits] Seaside 2.9: Javascript-Tests-Core-lr.35.mcz
by Julian Fitzell 29 Apr '09

29 Apr '09

You're a maniac! Go Lukas! :) On Tue, Apr 28, 2009 at 11:42 AM, <squeak-dev-noreply(a)lists.squeakfoundation.org> wrote: > Lukas Renggli uploaded a new version of Javascript-Tests-Core to project Seaside 2.9: > http://www.squeaksource.com/Seaside29/Javascript-Tests-Core-lr.35.mcz > > ==================== Summary ==================== > > Name: Javascript-Tests-Core-lr.35 > Author: lr > Time: 28 April 2009, 8:42:34 pm > UUID: 1f0cbe5e-13a5-4966-b415-07a4bb89812e > Ancestors: Javascript-Tests-Core-lr.34 > > - pushing test coverage from 74% to 91%

1 0

WAPackage
by Michael Lucas-Smith 23 Apr '09

23 Apr '09

This is sort of a Monticello + Seaside 2.9 question... I'm importing the source code from the mcz files. The dependencies and comments for the package are stored in methods extended on WAPackage. I'd rather not have my Monticello code referencing a Seaside class to fill in the comment and dependencies when I load up an mcz file.. is there some convention Monticello uses to "know" about WAPackage that I could use as part of my import process? Cheers, Michael

3 6

Character >> #seasideInteger
by Philippe Marschall 23 Apr '09

23 Apr '09

Hi In response to [1] there is now #seasideInteger on Character which is supposed to return the Unicode code point. Now there are as well some mean tests ;-) They utf-8 decode a series of bytes into a Character using WACodec and send #seasideInteger to the answer. This has the advantage that only ASCII characters are present in the source code but non-ASCII characters are tested. Additionally it tests your utf-8 implementation ;-) The test does this for characters that take up 1, 2, 3 *and* 4 bytes in utf-8. I'm not sure characters with a code point bigger than 65535 work on every dialect. However the tests should point out such issues fairly quickly. If you feel you're being treated unfairly the Squeak specific tests additionally cover leadingChar for unified Han ideographs ;-) [1] http://code.google.com/p/seaside/issues/detail?id=303 Cheers Philippe

6 24

Re: [Seaside Commits] Seaside 2.9: Seaside-Tests-Core-lr.124.mcz
by Julian Fitzell 22 Apr '09

22 Apr '09

Just so we know... why? Would be nice if the commit log included a rationale (I can see what the change was from the diff :) ). Julian On Tue, Apr 21, 2009 at 11:49 PM, <squeak-dev-noreply(a)lists.squeakfoundation.org> wrote: > Lukas Renggli uploaded a new version of Seaside-Tests-Core to project Seaside 2.9: > http://www.squeaksource.com/Seaside29/Seaside-Tests-Core-lr.124.mcz > > ==================== Summary ==================== > > Name: Seaside-Tests-Core-lr.124 > Author: lr > Time: 22 April 2009, 8:49:31 am > UUID: 56e8a1b9-3174-48a0-b0b9-48d32d252ebb > Ancestors: Seaside-Tests-Core-lr.123 > > - should:raise:description: -> should:raise: > > =============== Diff against Seaside-Tests-Core-lr.123 =============== > > Item was changed: > ----- Method: WAObjectTest>>testError (in category 'testing') ----- > testError > "Make sure #error: signals a subclass of WAError." > > + self should: [ WAObject new error: 'oh dear' ] raise: WAError. > + self should: [ WAObject error: 'oh dear' ] raise: WAError! > - self should: [ WAObject new error: 'oh dear' ] raise: WAError description: 'oh dear'. > - self should: [ WAObject error: 'oh dear' ] raise: WAError description: 'oh dear'! > > > _______________________________________________ > commits mailing list > To unsubscribe, email commits-leave(a)lists.seaside.st > http://lists.seaside.st/listinfo/commits >

3 4

Re: [Seaside Commits] Seaside 2.9: Seaside-Tests-Platform-lr.73.mcz
by Julian Fitzell 22 Apr '09

22 Apr '09

And shouldn't we still be testing that the description has been set? We could signal it again, actually catch it, and then assert the correct description... Julian On Tue, Apr 21, 2009 at 11:49 PM, <squeak-dev-noreply(a)lists.squeakfoundation.org> wrote: > Lukas Renggli uploaded a new version of Seaside-Tests-Platform to project Seaside 2.9: > http://www.squeaksource.com/Seaside29/Seaside-Tests-Platform-lr.73.mcz > > ==================== Summary ==================== > > Name: Seaside-Tests-Platform-lr.73 > Author: lr > Time: 22 April 2009, 8:49:49 am > UUID: bbe5191f-28cf-4bd7-8249-88098970f2fc > Ancestors: Seaside-Tests-Platform-pmm.72 > > - should:raise:description: -> should:raise: > > =============== Diff against Seaside-Tests-Platform-pmm.72 =============== > > Item was changed: > ----- Method: WAPlatformObjectTest>>testError (in category 'testing') ----- > testError > "Make sure #error: signals a subclass of WAPlatformError." > > + self should: [ WAPlatformObject new error: 'oh dear' ] raise: WAPlatformError. > + self should: [ WAPlatformObject error: 'oh dear' ] raise: WAPlatformError! > - self should: [ WAPlatformObject new error: 'oh dear' ] raise: WAPlatformError description: 'oh dear'. > - self should: [ WAPlatformObject error: 'oh dear' ] raise: WAPlatformError description: 'oh dear'! > > > _______________________________________________ > commits mailing list > To unsubscribe, email commits-leave(a)lists.seaside.st > http://lists.seaside.st/listinfo/commits >

2 2

Raising issues versus chatting about them here
by Michael Lucas-Smith 22 Apr '09

22 Apr '09

Just so I understand the preferred procedure here: Should I be raising issues in google code instead of raising issues here, then issues get discussed here if they aren't clear or immediately accepted in google code? I'd be happy to do either. Cheers, Michael

2 1

All the printer tests...
by Michael Lucas-Smith 22 Apr '09

22 Apr '09

Hi All, Float in VisualWorks is a 32bit float, not a 64bit float. Just about all of the printing tests end up failing because the numbers round differently to the expected results. We can declare the numbers as Doubles by using the 1.0d syntax, however that doesn't exist in Squeak, #d gets sent to the Float. Anybody have any suggestions? I'd really like to get all the platform tests passing before I move on. Cheers, Michael

2 1

Seaside2.8a1-damiencassou.586
by Philippe Marschall 19 Apr '09

19 Apr '09

Sorry but no. That will prevent users from loading it directly into plain Squeak. Make a different package for this. Cheers Philippe

2 2

streamed upload?
by Sebastian Sastre 16 Apr '09

16 Apr '09

hi there, do we have streamed uploads? I mean something that won't accumulate bytes in the image. I've see the upload functional test maintains the bytearray in the file instVar. That's absolutely necessary? How can we implement to store uploads of lets say 700MB? sebastian

4 7

hot to not use continuations
by Sebastian Sastre 09 Apr '09

09 Apr '09

hi there, I'm testing not to use continuations in an app. I've set: - Main Class: WAMain - Action Phase Continuation Class: WAActionPhaseContinuation interestingly, after setting that, the non stopping halt previously mentioned started to halt. Also I'm seeing in a filter (I've made for testing) this: handleFiltered: aRequestContext (super handleFiltered: aRequestContext) ifNotNilDo:[:response| self halt. ^ response]. it halts as expected but response has not aWAResponse, it has aFilter (itself). What I'm missing about not using continuations? thanks sebastian PD: that app is for testing. It's trivial: is just a component that renders a title. No tasks, no #call:

4 13

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

seaside-dev April 2009