Hi list, I speak for myself and not for IBM.
I should be working on CUPS and Samba printing, but this thread has created an itch I just have to scratch. I'll shed some light on the conspiracy theory, though you won't find a smoking gun :)) I proposed the redbook "Open Source Software for OS/390 UNIX" in 1999 after finishing a couple of redbooks which "dabbled" in open source. To my delight it was funded for the year 2000 and the *next month* Linux hit the scene. We ran an ITSO residency in February and March with Sandor Barany, Ralf Schandl and Egon Terwedow working hard on vim, openss*, groff, gnats, Perl and many others. Jim Tison caught wind of the project and happened to be on the same floor. He chipped in the autoconf and automake chapters. A lot of open source for OS/390 was available for the picking on the Web, especially on the MKS Web site (to MKS' credit, still online at ftp://ftp.mks.com/pub/s390/gnu/). It was a fun project. There was a lot of work done, but it was only a 6 week residency. If you've done software development, you know that six weeks is not a lot of time - oh and we had to write a book as well as work on the code. The first book was published in October of 2000 and was moderately successful as a book. I felt the code was more valuable than the book. Getting legal approval was a lot of work and it probably would never have happened were it not for GNU and Linux and IBM's new religion. For a background on this, see page 4 especially (http://dir.salon.com/tech/fsp/2000/09/12/chapter_7_part_one/index.html?pn=4) of "How Big Blue Fell for Linux" - an excellent article by Andrew Leonard. I proposed an update to the redbook and code which was funded for 2001. By now Linux on zSeries had critical mass and was a whole lot more UNIX-like and fun than z/OS UNIX Systems Services. Regardless, we had a project to complete. This time the residents were Thorsten Brockmeier, Guillermo Freige, Stefan Koesling, Kiran Madnani and Fulvio Malfatto all of whom did a great job. In retrospect, we bit off more than we could chew. We should have focused more on code quality and less on quantity (OpenSSL and OpenSSH have problems. The randomness of OpenSSL is not always adequate. SSH works with version 1 not version 2, sftp and scp do not work I believe.). The book ran into 2002 and was published in March. Both books had CDs in the back; a feature that the ITSO no longer supports. One of the complaints of the first book was that the source and binaries were distributed each as two large tar files with every package in each. So for the second book we split out the source packages and binaries at least on the tools and toys Web site - another chunk of work - especially packaging up tar files for the binaries of individual packages. In October of this year, both the redbook and the code were pulled from IBM Web sites (apparently not as thoroughly as possible :) to be "serviced". Steve Stiert is the Webmaster of Tools and Toys and the good news is that the code is going back on Web soon, per his append. I coincidentally just asked yesterday about the status of the redbook going back on the Web. There are some deletions to be made that would take the excellent editors at the ITSO perhaps a couple of hours to complete and turn the crank to make the book available. So the book should or at least could have been available by now. The holiday season is also the editors' busiest season as ITSO project leaders scramble to get their projects completed by the calendar year and move on to newly funded projects. So there's a queue to get this two hours of work done. The book might have been able to jump the queue, however the snag is that the ITSO no longer includes CDs in hard copies of books. Because of this a lot more deletions have to be made of references to "the CD included with the hard copy of the book" (interestingly the astounding number of 60 hard copies of this redbook were ever sold). Given these factors, I'm guessing the book will be available on the ITSO Web site early next year. I now wanted to step back and comment and just noticed David Froberg's append, so, thanks for that - I'll use it as a springboard: > 1) I really appreciated the web site and redbook because it helped > me greatly to get a handle on porting tools to USS. > 2) Its disappearance forced me to research and experiment more with > porting to USS since a number of tools at the web site really are very > valuable. Thanks, that's good to hear. > 3) The biggest thing that repeatedly struck me, though, was the > static nature of the web site's tools. (And this is NOT > knock against anyone at IBM. As I said in #1, I greatly appreciated > the site.) I also work with the Linux world and > and watch frequent notices and then frequent updates for various > tools (Samba for example is at what? 2.2.7a?). Would it not be nice to > have some web site and process by which the tools were more current? How > this would be done, not sure. Have a place in the site for us to > stick more current ports? Good point - I agree. Because the MKS work and the redbook projects were both outside the domain of traditional z/OS development, and all were finite in terms of project length, the result has been that there were three large drops in 1997-8 (MKS), 2000 (redbook 1) and 2002 (redbook 2), but other than that, there is not a lot of active development. In a way, z/OS and OSS are like oil and water - they are on opposite ends of the development spectrum. OSS's philosophy is "release early, release often". z/OS philosophy is to test the code really, really well before releasing because you can run your business 24x365 on this OS - the bazaar and the cathedral. At least, "the community" of OSS on USS professionals should work on the quality of OpenSSL and OpenSSH. Use the MVS-OE list server as a forum. Al Nichols, John McKown (sp?), Wayne Johnson, Peter Prymmer, Paul Gilmartin and Greg Smith have all contributed (I apologize to others I have forgotten), but in general there does not appear to be as much active OSS development on z/OS as with GNU/Linux (oil and water). If someone does the work, we could probably get a refreshed copy onto tools and toys. Perhaps a SourceForge type vehicle would be a better medium, but without a lot of resource being dedicated, the vehicle is not the issue. If your next thought is that IBM should dedicate more resource to OSS on USS, I would agree, but remember: 1) Oil and water 2) The entire ITSO/redbook organization is in place and funded to ease this tricky job of making computers hum. I would argued such an organization is unparalleled in our industry (where else can you get hundreds of free soft copy books each year?). This is a lot of dedicatd resource. 3) The entire Linux Technology Center (LTC) is funded by IBM on the promise of Linux becoming sustainable (now I could really go off on a tangent, but I'll stop here). In a strange way, having the code and the redbook pulled has been fortuitious in that it has brought OSS on USS back onto the powers'-that-be radar screens. So for the 90% of you guys looking for a smoking gun - get back to work on Linux!!! For the 10% of you who really have to do OSS on USS, 90% of you already have a softcopy of the redbook. So for the remaining 1%, get the copy Mark Post and Georgio Bellusi are going to make available - the next copy will have fewer words, but the same material. OK, my itch is scratched - I'm turning off e-mail and can get back to work on CUPS (hey it runs out of the box in SLES-8). I know this reply has been a bit rambling, but I hope it helps. -Mike MacIsaac, [EMAIL PROTECTED] (845) 433-7061
