I review all of the changes (I've just made a pass). You are correct that the ticket author should review; I didn't make this explicit but included it under the "anyone who is interested". Please make sure that the ticket author agrees with your realization of the ticket. Bill On Jul 22, 2009, at 8:21 AM, Richard Treumann wrote:

Thanks

I was hoping for something better.

I made some edits to my chapter. I checked all approved tickets and what I found matched the spread sheet Bill Gropp provided a while back. I made edits as called for by the identified tickets. If somehow I missed one, only the ticket owner has a chance of catching me out on that.

Nobody else should have changed my chapter. Obviously I reviewed each edit as I made the PDF files. I guess that review is basically automatic. No chapter author will share his PDFs without doing his own review.

I attached each pdf file to the ticket that inspired it but I did not make any additional notice to ticket owners. Since these changes are each a response to a ticket and the ticket owner is motivated to get his ticket closed, the onus best falls on the ticket owner. If the chapter author and a representative ticket owner agree that the ticket has been fully addressed then who else is logical as a reviewer?

It is good that Bill Gropp is able to act broadly as a backup reviewer so he should be notified. I sent an email saying all my PDFs as called for by the spread sheet are now attached to appropriate tickets.

Making the chapter author rather that the ticket author the key gatekeeper seems backwards to me. We will close any completed ticket soon but the chapters will change more times. Shouldn't the checkpoints of the process link up with the clear state transitions a ticket can move through?

Dick

Dick Treumann - MPI Team IBM Systems & Technology Group Dept X2ZA / MS P963 -- 2455 South Road -- Poughkeepsie, NY 12601 Tele (845) 433-7846 Fax (845) 433-8363

<graycol.gif>"Bronis R. de Supinski" ---07/22/2009 08:51:39 AM------------- Forwarded message ---------- Date: Tue, 23 Jun 2009 16:25:37 -0500

<ecblank.gif> From: <ecblank.gif> "Bronis R. de Supinski" <bronis_at_[hidden]> <ecblank.gif> To: <ecblank.gif> "MPI 2.2" <mpi-22_at_[hidden]> <ecblank.gif> Date: <ecblank.gif> 07/22/2009 08:51 AM <ecblank.gif> Subject: <ecblank.gif> Re: [Mpi-22] MPI 2.2 Document Update (fwd) <ecblank.gif> Sent by: <ecblank.gif> mpi-22-bounces_at_[hidden]

---------- Forwarded message ---------- Date: Tue, 23 Jun 2009 16:25:37 -0500 From: William Gropp <wgropp_at_[hidden]> To: Bronis R. de Supinski <bronis_at_[hidden]> Cc: "Supalov, Alexander" <alexander.supalov_at_[hidden]>, Adam Moody <moody20_at_[hidden]>, David George Solt <david.solt_at_[hidden]

...

, Jesper Larsson Traeff <traff_at_[hidden]>, Rolf Rabenseifner <rabenseifner_at_[hidden]>, Torsten Hoefler <htor_at_[hidden]>, Jeff Squyres <jsquyres_at_[hidden] , Rajeev Thakur <thakur_at_[hidden]>, George Bosilca <bosilca_at_[hidden] , Richard Graham <rlgraham_at_[hidden]>, Richard Treumann <treumann_at_[hidden]

Subject: Re: MPI 2.2 Document Update

Bronis,

1. Number of reviews. This is a different review, so it should be the chapter author, anyone else who is interested (which should be the author of a different chapter if the chapter author created the change, as will be the usual case), and me. Unlike the reviews of the proposals, this is primarily to ensure that the edits implemented the change.

2. The chapter author is responsible for getting the reviews done. My reasoning is that the chapter author is responsible for any fixes, so there's no reason to introduce a third party.

Bill

On Jun 23, 2009, at 11:43 AM, Bronis R. de Supinski wrote:

...

Bill:

Some general questions:

1. How many reviews of the PDF are required? Is it 4 again?

2. Who is responsible for getting the PDF reviews done? The chapter author? The ticket owner?

Having the chpater author makes sense to me since they clearly know when it is ready. That also allows the owner to be a reviewer when the owner and author are different. However, that does not really work for the tickets with multiple authors. Maybe in that case it makes the most sense for the last author to be responsible? Or perhaps this problem makes it most reasonable to make the author responsible but to require the owner to review the ticket before requesting other reviews if the owner is not the author?

Anyway, I am asking because it is not clear to me which tickets I am expected to review and I would like someone to poke me on the ones that I should review...

Bronis

On Thu, 18 Jun 2009, William Gropp wrote:

...

I've made an update as well as fixing the macros. One thing to remember (this is noted in the README-2.2), the "color" macro only applies until the end of the current TeX group - that makes it

pretty

...

...

painful within a table or other environment. There's really no easy work around without changing color packages.

For creating the PDFs, if you can't easily extract the relevant pages, just post the entire chapter - but in that case, include a comment that indicates which pages have changes. An alternative would be to use TeX/LaTeX as Knuth intended; there are many tools to select subsets of pages that will work with TeX DVI files, and most of the PDF tricks can be preserved as well. If someone is interested in pursuing that option, let me know.

And the Makefile for the Appendix was missing - I've fixed that now.

Bill On Jun 18, 2009, at 10:14 AM, Supalov, Alexander wrote:

...

Thanks for corrections. Originally I tried to use the change macros around the change table lines, but the document would not be built then. I assumed that the macros might not be functional inside some LaTeX entities. This is why I opportunistically surrounded the whole tables by macros.

As to the PDF - change in the tables, especially their removal, makes all other pages slip. I guess in this case it may be advisable to provide a full section chapter. I can cut that out using Adobe Acrobat, but I'm not sure everyone has access to this fairly expensive product or its equivalent. Is there a free utility that can extract indicated pages out of a PDF file?

I was not able to produce a PDF file for the appendix alone. Maybe there's some make target for this?

From: William Gropp [mailto:wgropp_at_[hidden]] Sent: Thursday, June 18, 2009 4:10 PM To: Supalov, Alexander Cc: Adam Moody; David George Solt; Jesper Larsson Traeff; Rolf Rabenseifner; Torsten Hoefler; Jeff Squyres; Bronis R. de Supinski; Rajeev Thakur; George Bosilca; Richard Graham; Richard Treumann Subject: Re: MPI 2.2 Document Update

Thanks, Alexander.

This brings up a few more instructions:

The PDFs should be of the pages with changes or deletions; to make it easier on the reviewers, just those pages (and a previous or following page if the change is at the top or bottom) should be included. If you can't extract the pages, or there are too many affected pages, just post the chapter's PDF.

In looking over Alexander's changes, I realized that I needed to specify the use of the \MPIupdate macros more precisely. They should be used around only the change - for example, if the change is within a table, then only the affected items within the table should be marked. I've modified some of Alexander's changes to move the update macros within the tables.

This brought up another problem - The color highlighting is currently overwritten by some of the MPI 2.1 merge/update macros. I will modify them so that the proper color is displayed. That may take a few days.

Bill

On Jun 18, 2009, at 5:49 AM, Supalov, Alexander wrote:

...

Dear Bill,

What kind of PDFs should be attached to the tickets to complete the work? Please clarify.

FYI, I've fixed my issues (## 4, 40, 63) in the chap-appLang taken off https:// svn.mpi-forum.org/svn/mpi-forum-docs/trunk/ MPI-2.2 , revision 241. Also added two changeLog references requested by Rolf:

William Gropp Deputy Director for Research Institute for Advanced Computing Applications and Technologies Paul and Cynthia Saylor Professor of Computer Science University of Illinois Urbana-Champaign

---

...

...

...

Intel GmbH Dornacher Strasse 1 85622 Feldkirchen/Muenchen Germany Sitz der Gesellschaft: Feldkirchen bei Muenchen Geschaeftsfuehrer: Douglas Lusk, Peter Gleissner, Hannes Schwaderer Registergericht: Muenchen HRB 47456 Ust.-IdNr. VAT Registration No.: DE129385895 Citibank Frankfurt (BLZ 502 109 00) 600119052

This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.

William Gropp Deputy Director for Research Institute for Advanced Computing Applications and Technologies Paul and Cynthia Saylor Professor of Computer Science University of Illinois Urbana-Champaign

William Gropp Deputy Director for Research Institute for Advanced Computing Applications and Technologies Paul and Cynthia Saylor Professor of Computer Science University of Illinois Urbana-Champaign

_______________________________________________ mpi-22 mailing list mpi-22_at_[hidden] http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-22

<ATT00001.txt>

William Gropp Deputy Director for Research Institute for Advanced Computing Applications and Technologies Paul and Cynthia Saylor Professor of Computer Science University of Illinois Urbana-Champaign *