Sixx update
Locked
 
|
FYI,
Norbert --- Name: SIXX-NorbertHartl.169 Author: NorbertHartl Time: 08/04/2010, 12:02:34 UUID: e0611da1-3501-4156-9c59-1aaaccb465ce Ancestors: SIXX-DaleHenrichs.168 removed all usages of the convenient "text" selectors in stream. Grease stream does not support selectors like cr or tab: This makes some of the code in Sixx a little bigger but also more clean. Now Sixx can be used with GRCodecStream to output encoded strings directly |
|
|
Norbert,
did you run the SIXX tests ... I'm using Seaside 3.0.0-rc.1 and a lot of the SIXX tests are failing with what I assume to be line terminator issues ... There is GRPlatform current newline that should be used to get the line terminator correct for the platform (in GemStone streams I think that #cr actually outputs Character>>lf, since Character>>lf is the native newline for GemStone)...At least that's my guess ... Dale Norbert Hartl wrote: > FYI, > > Norbert > > --- > Name: SIXX-NorbertHartl.169 > Author: NorbertHartl > Time: 08/04/2010, 12:02:34 > UUID: e0611da1-3501-4156-9c59-1aaaccb465ce > Ancestors: SIXX-DaleHenrichs.168 > > removed all usages of the convenient "text" selectors in stream. Grease stream does not support selectors like cr or tab: > > This makes some of the code in Sixx a little bigger but also more clean. Now Sixx can be used with GRCodecStream to output encoded strings directly > |
|
|
On 04.08.2010, at 23:01, Dale Henrichs wrote: > Norbert, > > did you run the SIXX tests ... I'm using Seaside 3.0.0-rc.1 and a lot of the SIXX tests are failing with what I assume to be line terminator issues ... > > There is GRPlatform current newline that should be used to get the line terminator correct for the platform (in GemStone streams I think that #cr actually outputs Character>>lf, since Character>>lf is the native newline for GemStone)...At least that's my guess ... > Gotcha :) What do we learn from this? There is _really_ no change without side effect. I'll take care about that. Using GRPlatform would make the fix very easy but it would make a lot of things grease dependent. If you'd say it is feasible to force all users (that want to use sixx) to install grease by default or that grease is a standard component from now on than I'm happy to use it. While I'm at it I'll take a look at the problems that prevented xml parser to be integrated in the 1.0-beta.8 line up. If you didn't take care in the meantime. Norbert > Norbert Hartl wrote: >> FYI, >> Norbert >> --- >> Name: SIXX-NorbertHartl.169 >> Author: NorbertHartl >> Time: 08/04/2010, 12:02:34 >> UUID: e0611da1-3501-4156-9c59-1aaaccb465ce >> Ancestors: SIXX-DaleHenrichs.168 >> removed all usages of the convenient "text" selectors in stream. Grease stream does not support selectors like cr or tab: >> This makes some of the code in Sixx a little bigger but also more clean. Now Sixx can be used with GRCodecStream to output encoded strings directly |
|
|
Dale,
you were absolutely right. It was the line ending that made the tests fail. All tests are back to green. Norbert On 04.08.2010, at 23:17, Norbert Hartl wrote: > > On 04.08.2010, at 23:01, Dale Henrichs wrote: > >> Norbert, >> >> did you run the SIXX tests ... I'm using Seaside 3.0.0-rc.1 and a lot of the SIXX tests are failing with what I assume to be line terminator issues ... >> >> There is GRPlatform current newline that should be used to get the line terminator correct for the platform (in GemStone streams I think that #cr actually outputs Character>>lf, since Character>>lf is the native newline for GemStone)...At least that's my guess ... >> > Gotcha :) What do we learn from this? There is _really_ no change without side effect. I'll take care about that. Using GRPlatform would make the fix very easy but it would make a lot of things grease dependent. If you'd say it is feasible to force all users (that want to use sixx) to install grease by default or that grease is a standard component from now on than I'm happy to use it. > > While I'm at it I'll take a look at the problems that prevented xml parser to be integrated in the 1.0-beta.8 line up. If you didn't take care in the meantime. > > Norbert > > >> Norbert Hartl wrote: >>> FYI, >>> Norbert >>> --- >>> Name: SIXX-NorbertHartl.169 >>> Author: NorbertHartl >>> Time: 08/04/2010, 12:02:34 >>> UUID: e0611da1-3501-4156-9c59-1aaaccb465ce >>> Ancestors: SIXX-DaleHenrichs.168 >>> removed all usages of the convenient "text" selectors in stream. Grease stream does not support selectors like cr or tab: >>> This makes some of the code in Sixx a little bigger but also more clean. Now Sixx can be used with GRCodecStream to output encoded strings directly > |
|
|
In reply to this post by NorbertHartl
Norbert,
I'm happy to have a SIXX dependent Grease ... getting SIXX to work with the latest XML Parser would be a bonus ... Dale Norbert Hartl wrote: > On 04.08.2010, at 23:01, Dale Henrichs wrote: > >> Norbert, >> >> did you run the SIXX tests ... I'm using Seaside 3.0.0-rc.1 and a lot of the SIXX tests are failing with what I assume to be line terminator issues ... >> >> There is GRPlatform current newline that should be used to get the line terminator correct for the platform (in GemStone streams I think that #cr actually outputs Character>>lf, since Character>>lf is the native newline for GemStone)...At least that's my guess ... >> > Gotcha :) What do we learn from this? There is _really_ no change without side effect. I'll take care about that. Using GRPlatform would make the fix very easy but it would make a lot of things grease dependent. If you'd say it is feasible to force all users (that want to use sixx) to install grease by default or that grease is a standard component from now on than I'm happy to use it. > > While I'm at it I'll take a look at the problems that prevented xml parser to be integrated in the 1.0-beta.8 line up. If you didn't take care in the meantime. > > Norbert > > >> Norbert Hartl wrote: >>> FYI, >>> Norbert >>> --- >>> Name: SIXX-NorbertHartl.169 >>> Author: NorbertHartl >>> Time: 08/04/2010, 12:02:34 >>> UUID: e0611da1-3501-4156-9c59-1aaaccb465ce >>> Ancestors: SIXX-DaleHenrichs.168 >>> removed all usages of the convenient "text" selectors in stream. Grease stream does not support selectors like cr or tab: >>> This makes some of the code in Sixx a little bigger but also more clean. Now Sixx can be used with GRCodecStream to output encoded strings directly > |
|
|
On 05.08.2010, at 00:22, Dale Henrichs wrote: > Norbert, > > I'm happy to have a SIXX dependent Grease ... getting SIXX to work with the latest XML Parser would be a bonus ... > Guess what I'm doing at the moment ;) Norbert > > Norbert Hartl wrote: >> On 04.08.2010, at 23:01, Dale Henrichs wrote: >>> Norbert, >>> >>> did you run the SIXX tests ... I'm using Seaside 3.0.0-rc.1 and a lot of the SIXX tests are failing with what I assume to be line terminator issues ... >>> >>> There is GRPlatform current newline that should be used to get the line terminator correct for the platform (in GemStone streams I think that #cr actually outputs Character>>lf, since Character>>lf is the native newline for GemStone)...At least that's my guess ... >>> >> Gotcha :) What do we learn from this? There is _really_ no change without side effect. I'll take care about that. Using GRPlatform would make the fix very easy but it would make a lot of things grease dependent. If you'd say it is feasible to force all users (that want to use sixx) to install grease by default or that grease is a standard component from now on than I'm happy to use it. While I'm at it I'll take a look at the problems that prevented xml parser to be integrated in the 1.0-beta.8 line up. If you didn't take care in the meantime. >> Norbert >>> Norbert Hartl wrote: >>>> FYI, >>>> Norbert >>>> --- >>>> Name: SIXX-NorbertHartl.169 >>>> Author: NorbertHartl >>>> Time: 08/04/2010, 12:02:34 >>>> UUID: e0611da1-3501-4156-9c59-1aaaccb465ce >>>> Ancestors: SIXX-DaleHenrichs.168 >>>> removed all usages of the convenient "text" selectors in stream. Grease stream does not support selectors like cr or tab: >>>> This makes some of the code in Sixx a little bigger but also more clean. Now Sixx can be used with GRCodecStream to output encoded strings directly |
|
|
In reply to this post by Dale Henrichs
On 05.08.2010, at 00:22, Dale Henrichs wrote: > Norbert, > > I'm happy to have a SIXX dependent Grease ... getting SIXX to work with the latest XML Parser would be a bonus ... > Ok, there is now SIXX-NorbertHartl.170 that has the Character cr replaced by GRPlatform current newline. And now there is XML-Parser.g-NorbertHartl.73 that upgrades the XML parser to a state where namespace handling has been added. You may have seen my mail to the pharo list. The squeaksource version after the one I've ported has some encoding issues that is an issue either in pharo monticello or in gemstone. I need to clarify how to update the ConfigurationOfXMLSupport. I will update it then for the new versionse. And I tend to integrate Sixx into the ConfigurationOfXMLSupport. Norbert > Dale > > Norbert Hartl wrote: >> On 04.08.2010, at 23:01, Dale Henrichs wrote: >>> Norbert, >>> >>> did you run the SIXX tests ... I'm using Seaside 3.0.0-rc.1 and a lot of the SIXX tests are failing with what I assume to be line terminator issues ... >>> >>> There is GRPlatform current newline that should be used to get the line terminator correct for the platform (in GemStone streams I think that #cr actually outputs Character>>lf, since Character>>lf is the native newline for GemStone)...At least that's my guess ... >>> >> Gotcha :) What do we learn from this? There is _really_ no change without side effect. I'll take care about that. Using GRPlatform would make the fix very easy but it would make a lot of things grease dependent. If you'd say it is feasible to force all users (that want to use sixx) to install grease by default or that grease is a standard component from now on than I'm happy to use it. While I'm at it I'll take a look at the problems that prevented xml parser to be integrated in the 1.0-beta.8 line up. If you didn't take care in the meantime. >> Norbert >>> Norbert Hartl wrote: >>>> FYI, >>>> Norbert >>>> --- >>>> Name: SIXX-NorbertHartl.169 >>>> Author: NorbertHartl >>>> Time: 08/04/2010, 12:02:34 >>>> UUID: e0611da1-3501-4156-9c59-1aaaccb465ce >>>> Ancestors: SIXX-DaleHenrichs.168 >>>> removed all usages of the convenient "text" selectors in stream. Grease stream does not support selectors like cr or tab: >>>> This makes some of the code in Sixx a little bigger but also more clean. Now Sixx can be used with GRCodecStream to output encoded strings directly |
|
|
Norbert,
We've bumped into WideStrings creeping into Monticello packages before ... I consider this a bug ... if non-ascii characters are to be stored in an mcz file, then UTF8 should be used to encode the strings ... As I think about, I assume that it _is_ possible to create QuadByteString-based mcz files from GemStone:), but using utf8 is only a solution if everyone follows the convention... Integrating SIXX into the XML config is probably a good idea ... once you've got it integrated ... I'll phase out the separate config that is GemStone-specific. Dale Norbert Hartl wrote: > On 05.08.2010, at 00:22, Dale Henrichs wrote: > >> Norbert, >> >> I'm happy to have a SIXX dependent Grease ... getting SIXX to work with the latest XML Parser would be a bonus ... >> > Ok, there is now SIXX-NorbertHartl.170 that has the Character cr replaced by GRPlatform current newline. And now there is XML-Parser.g-NorbertHartl.73 that upgrades the XML parser to a state where namespace handling has been added. You may have seen my mail to the pharo list. The squeaksource version after the one I've ported has some encoding issues that is an issue either in pharo monticello or in gemstone. > > I need to clarify how to update the ConfigurationOfXMLSupport. I will update it then for the new versionse. And I tend to integrate Sixx into the ConfigurationOfXMLSupport. > > > Norbert > >> Dale >> >> Norbert Hartl wrote: >>> On 04.08.2010, at 23:01, Dale Henrichs wrote: >>>> Norbert, >>>> >>>> did you run the SIXX tests ... I'm using Seaside 3.0.0-rc.1 and a lot of the SIXX tests are failing with what I assume to be line terminator issues ... >>>> >>>> There is GRPlatform current newline that should be used to get the line terminator correct for the platform (in GemStone streams I think that #cr actually outputs Character>>lf, since Character>>lf is the native newline for GemStone)...At least that's my guess ... >>>> >>> Gotcha :) What do we learn from this? There is _really_ no change without side effect. I'll take care about that. Using GRPlatform would make the fix very easy but it would make a lot of things grease dependent. If you'd say it is feasible to force all users (that want to use sixx) to install grease by default or that grease is a standard component from now on than I'm happy to use it. While I'm at it I'll take a look at the problems that prevented xml parser to be integrated in the 1.0-beta.8 line up. If you didn't take care in the meantime. >>> Norbert >>>> Norbert Hartl wrote: >>>>> FYI, >>>>> Norbert >>>>> --- >>>>> Name: SIXX-NorbertHartl.169 >>>>> Author: NorbertHartl >>>>> Time: 08/04/2010, 12:02:34 >>>>> UUID: e0611da1-3501-4156-9c59-1aaaccb465ce >>>>> Ancestors: SIXX-DaleHenrichs.168 >>>>> removed all usages of the convenient "text" selectors in stream. Grease stream does not support selectors like cr or tab: >>>>> This makes some of the code in Sixx a little bigger but also more clean. Now Sixx can be used with GRCodecStream to output encoded strings directly > |
«
Return to GLASS
|
1 view|%1 views
| Free forum by Nabble | Edit this page |