home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!wupost!cs.utexas.edu!rutgers!spcvxb.spc.edu!terry
- From: terry@spcvxb.spc.edu (Terry Kennedy, Operations Mgr.)
- Newsgroups: vmsnet.vms-posix
- Subject: Re: name of VMS product has changed...
- Message-ID: <1992Aug20.011559.3656@spcvxb.spc.edu>
- Date: 20 Aug 92 05:15:59 GMT
- References: <2415@mitech.com> <1992Aug18.091028@mccall.com> <1992Aug18.182842.3645@spcvxb.spc.edu> <1992Aug19.153401@mccall.com>
- Organization: St. Peter's College, US
- Lines: 74
-
- In article <1992Aug19.153401@mccall.com>, tp@mccall.com (Terry Poot) writes:
- > I must have missed the "do not install Posix 1.0" letter. :-)
-
- I think there's an implicit "do not install" for anything that's V1.0 8-)
-
- > Do you suppose that hyping it to the marketplace 6 months before it is ready
- > might be putting pressure on developers to get it out early, rather than taking
- > the time to make it work? As far as multi-platform support, it will certainly
- > take more resources, but often times a bug will be uncovered on one platform
- > that was latent on another. I'm not sure it intrinsically affects quality,
- > except to the extent that it increases the pressure to skip the testing, because
- > the developers are spread thinner. (So what does DEC do? Lay people off, of
- > course.)
-
- I've said this before (and trust me, it hasn't made me many friends in VMS
- Engineering), but I'll mention it again here:
-
- You can't test quality into a product - it needs to be in there from the
- start. I know that DEC can produce some excellent software (sometimes even
- in V1.0 products - eXcursion is a good example). DEC needs to spread the word
- about successful development methods around the company.
-
- I think that part of the problem is that VMS folks deal with very small
- pieces of the OS (of course there are exceptions: Andy, Kathy, Keith, etc.).
- I once asked a VMS engineer what he did and his answer was "I support the
- set characteristics subset function of the terminal class driver". I _think_
- he was joking. When the number of interfaces doubles, the communication
- squares.
-
- However, even FT isn't as successful as it should be. DEC seems to select
- sites which match DEC's internal use of VMS (whatever that is - it doesn't
- seem to include using BACKUP or LAT) [note: old comment - these problems seem
- to be mostly fixed now]. Then there are the DEC Cooperative Marketing Part-
- ners or whatever they're called this week - folks that let DEC sell their
- software, like Joiner Software (Jnet). Those folks are only concerned with
- making sure _their_ package didn't break. And then there are what I call the
- "Uncooperative marketing partners" [that's a joke] like Executive Software.
- ES is a VMS FT site. Somehow I don't think DEC asked them to do that out of
- the goodness of DEC's heart - but there would be enough folks mad at _DEC_
- if, say, Diskeeper didn't work on a new VMS release that it's a necessity
- that they be a test site.
-
- When someone who does a thorough job happens to sneak in (and I spoke to
- one such site which was _very_ disillusioned with the whole thing) the re-
- ports of "BACKUP being completely unusable" were ignored because "if it was
- a real problem one of the other sites would have reported it".
-
- I think the problem is that while the folks who write the code have pride
- in their work, and they don't mind being told "gee, that's ugly code" every
- now and again, they're so isolated that they never get the FT feedback, or
- it's mangled so much they can't make sense of it.
-
- That's just a guess - I'm not, nor have I ever been, a VMS FT site. How-
- ever, I have tested lots of other stuff for DEC in the past. RSTS/E V10.0
- (since released) had a "Field Test panel" a few Symposia ago where the FT
- sites reported experiences to the attendees (with DEC's blessing). One of
- the comments I made there was that V10.0 was the first release to support
- printing to LAT queues (like LATSYM with VMS). I reported a problem with it
- crashing the system when started. I had a workaround for it in 15 minutes
- and I had a patch which fixed it completely at the end of that day. Since
- then I have had _zero_ problems with the RSTS version (6 RSTS systems com-
- peting for 8 printers with 6 VAX systems). Whenever a printer is stuck, it's
- invariably the VMS symbiont that's wedged. How long has LATSYM been part of
- VMS? V4.4? Earlier? And it still doesn't work properly?
-
- I'll close with a comment on timely SPR answers. I filed a SPR against
- VMS V4.1, complaining that it didn't operated the TSV05 in high speed mode.
- I got the usual "fixed in a future release". I forgot about it. Years (many)
- went by. Well, guess what! It's fixed. In VMS V5.4-1. Amazing turnaround. It
- was a missing register save.
-
- Terry Kennedy Operations Manager, Academic Computing
- terry@spcvxa.bitnet St. Peter's College, Jersey City, NJ USA
- terry@spcvxa.spc.edu +1 201 915 9381
-