April 2006 - saga-rg - lists.ogf.org

SAGA conference call
by Shantenu Jha 27 Apr '06

27 Apr '06

Hi All, After a brief hiatus, we're ready to resume our regular conference calls. Suggested time for the next call is 2PM BST on Thursday 20th April, which translates to 8AM CST and 3PM in Amsterdam (and probably close to midnight in Tokyo. Sorry Pascal.) The agenda for the next call is: - Issue List - SAGA and GridRPC How to coordinate activities of GridRPC into SAGA. - SAGA and GridCPR - AOB ? Conference call details to come soon. If you'd like to join, but suggested time doesn't work for you, please let us know. Thanks.

5 15

ISSUE 90
by Andre Merzky 20 Apr '06

20 Apr '06

90) exceptions missing in job! - OPEN, URGENT Well, that says it all. Volonteers? Cheers, Andre. -- "So much time, so little to do..." -- Garfield

1 0

ISSUE 82
by Andre Merzky 20 Apr '06

20 Apr '06

82) Explain sidl.SIDLException ! - OPEN, URGENT Short issue: we need to explain what SIDL.Esception means. Volonteers? Cheers, Andre. -- "So much time, so little to do..." -- Garfield

1 0

ISSUE 76
by Andre Merzky 20 Apr '06

20 Apr '06

76) add specs for literals (attribute and metric names), and add them like enums - OPEN, URGENT Someone needs to go over the spec and make sure, that attribute names and metric names are completely listed. That is done for Streams, but needs to be done for all packages. The respective infos have to taken from the use cases. Volonteers? Also, does anybody see a better way to define attributes and metrics than using comments in SIDL? Cheers, Andre. -- "So much time, so little to do..." -- Garfield

1 0

ISSUE 12
by Andre Merzky 20 Apr '06

20 Apr '06

12) change signature of read/write to have len_out as out param, and to return error code (== EAGAIN on unsucessful read) - DECIDED at GGF16 - is that _really_ the right way? - OPEN, URGENT At last GGF we discussed that issue - sorry for re-opening it. My issue with the issue is, that this would be (almost) the only place where we return an error code. Also, returning an error code dviates from the POSIX version of read, which would flag an error with a negative return value. Sorry, I could not convince myself to change that in the spec - hence this opinion poll on the mailing list: (A) read (in int len_in, in char* buffer, out int len_out, out int error); error is an enum, like EAGAIN, EINTR etc, see man (2) read. Similar signature for other I/O calls (B) read (in int len_in, in char* buffer, out int retval); retval is the length of data read if positive, and flags error if negative, with definitions for EAGAIN, EINTR etc. Yes, that semantically overloads a return value. If no (other) votes come in, that issue will be marked solved with (A) as decided in Athens. I vote (B). Cheers, Andre. -- "So much time, so little to do..." -- Garfield

1 0

Java bindings...
by Andre Merzky 18 Apr '06

18 Apr '06

Hi folx, I made the statement on this list (and elsewhere) a couple of times, that Graemes (think OMII) Java bindings cover a part of the SAGA API. Graeme corrected me recently: its the local implementation which is intented to cover only parts of the API. The Java bindings themself are supposed to cover the _complete_ spec. Sorry for the confusion! I am happy to see a complete coverage of the sepc in Java :-)) To be on the save side: Malcolm, the EPCC implementation _is_ partial (file + jobs), right? Thanks, Andre. PS.: Grames bindings are available on http://argos.e-science.soton.ac.uk/blogs/GeodiseSAGA/ -- "So much time, so little to do..." -- Garfield

1 0

GridForge...
by Andre Merzky 07 Apr '06

07 Apr '06

Dear group, I am happy to announce that both the saga-rg and the saga-core-wg have been (semi-randomly) picked as guinea-pigs for the new GridForge version. What does that mean? GGF is trying the current infamous and often-complained-about version of GridForge with a new, much improved and soewhat customized version. The transition is to be performed in two stages: - stage one (now): a number of group projects get migrated to the new installation, and stress-test it for two weeks - stage two (in two weeks): based on the feedback of the test groups, patches are applied, and all group projects are migrated to the new version According to step one, projects for saga-rg and saga-core-wg have been created on testforge.ggf.org (which hosts the new installation). The GridForge user ID's are _all_ migrated already, so you should be able to use your old login. Also, you should have the correct membership etc. Please let me know if there is something wrong with your account. Of course its also possible to create new accounts. A striking difference between the old and new GF is that the 'help' function is actually useful: please try to consult it whenever you think you are lost etc, and tell us if the help did not tell you what you wanted to know. I made the group chairs and secretaries admins of the respective projects. If someone else wants to play around with project configuration etc, please let me know, and I'll add the respective accounts. Well, the bottom line is: please try to use the new GridForge, and give us feedback! We think its MUCH better than the old one, but we need YOUR confirmation before we go ahead and start adopting it for all GGF projects. http://testforge.ggf.org/sf/sfmain/do/viewProject/projects.saga-rg http://testforge.ggf.org/sf/sfmain/do/viewProject/projects.saga-core-wg Thanks, Andre. PS.: we found the first major glitch: email addresses did not get migrated correctly. I'll try to fix that manually for the saga folx - please let me know if you have trouble with mail. PPS.: the saga-core-group is new - you probably need to join the project, as only a few people are members there as of yet. -- "So much time, so little to do..." -- Garfield

1 0

strawman...
by Andre Merzky 06 Apr '06

06 Apr '06

Hi folx, GGF17 is getting closer, and we had the plan to have a stable strawman API by then. That means we whould try hard to address the remaining issues. I attach the current list of open issues: if you have thoughts to any of them, please post them to the list. Also, we should try to take up phone calls I think, and dedicate them to the issue list. Best regards, Andre. -- "So much time, so little to do..." -- Garfield

1 0