bob@allosaur.cis.ohio-state.edu (Bob Sutterfield) (09/30/88)
I just got off the phone from a conversation with a technical support representative of a vendor of software that runs on workstations. I'm pretty amazingly frustrated, so I thought I'd share it with y'all :-) Their software has historically run on workstations' native window system (e.g. SunView), but has been ported to X10. We're running their X10 client product on our Sun-3s, and it's in regular classroom use. The classes that use this software chose it for its function, and they're fairly satisfied with it on that basis. Windows-wise, it does OK, even if their user interface still looks exactly like SunView, just reimplemented under X10. We're trying to convert everyone here to X11. One reason for individuals not to convert is if the software they need to use doesn't yet run under X11, and a major sticking point is this vendor's software. So I asked the tech rep when they would be shipping a release of their software that would work under X11. He agreed that X11 is the wave of the future, and pointed out that their company is very interested in following standards, and that they're following developments in the X world very closely. He said that they would support X11 just as soon as the hardware vendors begin to support X11 on their platforms. Since Sun doesn't ship an X11 server, they don't ship an X11 client for Suns. I pointed out that Sun never shipped X10 for their machines, so their entire Sun X10 customer base is running on servers that they got for free from MIT, and that there is already a group of potential customers running free X11 servers from MIT. He wasn't sure whether they had received their X11 tape from MIT yet, but indicated that they still wouldn't be interested in shipping a client product for a machine until the hardware's vendor was shipping a server for it. Has anyone else encountered this sort of situation? Have you had any luck inducing motion? Can you share your technique with us? Any suggestions, other than gnawing on the telephone handset and banging on my keyboard? -=- Zippy sez, --Bob A dwarf is passing out somewhere in Detroit!
pda@stiatl.UUCP (Paul Anderson) (09/30/88)
In article <22992@tut.cis.ohio-state.edu> you write: >I just got off the phone from a conversation with a technical support >representative of a vendor of software that runs on workstations. I'm >pretty amazingly frustrated, so I thought I'd share it with y'all :-) > > ... >Has anyone else encountered this sort of situation? Have you had any >luck inducing motion? Can you share your technique with us? Any >suggestions, other than gnawing on the telephone handset and banging >on my keyboard? 1) Indicate that you are going to post their name to the network for bad support efforts, as you 'dont want anyone else to be taken for a ride by this vendor'. 2) Find another vendor (of anything) and pull out their business card in a face-to-face meeting with your troublesome vendor, flash it at them and indicate 'These folks want my money and are willing to help.' Be prepared to back it up. As soon as your troublemaker comes around to helping you, be very positive in your reaction. (Positive reinforcement of an attititude you find acceptable) 3) Order equipment from another vendor and make sure your troublemaker sees it. Rave on about how quickly it came in, how well it works, etc. Do so very innocently, or better yet, have an accomplice do it. 4) Stop payments on any invoices, threatening to return existing equipment until the situation is resolved 'more than adequately' to your satisfaction. You have been wronged, so be just a _little_ irrational. I don't necessarily use the order given, but basically, I learned a long time ago, "Money Talks, talk walks". Money is his lifeline. Use it. I hate to get hardnosed, but some people just won't hear you until you get that way. Personally, I would like to see you post the vendors name, as I am an X User and I want to stay away from vendors that don't believe that the customer is Number One. If nothing else, these newsgroups are capable of being (and have been) the Consumer Reports of the computer industry. -- Paul Anderson decvax!gatech!stiatl!pda Sales Technologies, Inc 3399 Peachtree Rd, NE X isn't just an adventure, Atlanta, GA (404) 841-4000 X is a way of life...
preece%vger@XENURUS.GOULD.COM (Scott E. Preece) (10/03/88)
> In article <22992@tut.cis.ohio-state.edu> you write: > >I just got off the phone from a conversation with a technical support > >representative of a vendor of software that runs on workstations. I'm > >pretty amazingly frustrated, so I thought I'd share it with y'all :-) > > > > ... > >Has anyone else encountered this sort of situation? Have you had any > >luck inducing motion? Can you share your technique with us? Any > >suggestions, other than gnawing on the telephone handset and banging > >on my keyboard? > > From: emcard!stiatl!pda@gatech.edu (Paul Anderson) > 1) Indicate that you are going to post their name to the network for bad... > 2) Find another vendor (of anything) and pull out their business card... > 3) Order equipment from another vendor and make sure your troublemaker... > 4) Stop payments on any invoices, threatening to return existing ... > > Personally, I would like to see you post the vendors name, as I am an > X User and I want to stay away from vendors that don't believe that the > customer is Number One. If nothing else, these newsgroups are capable > of being (and have been) the Consumer Reports of the computer industry. ---------- You guys are being pretty hard on a vendor for not be willing to ship something based on non-final versions of X. People who sell software like to have happy customers; as a group they don't usually come home from sales meetings saying "Wow, I really put it to those guys today..." MOST customers want stable, solidly supported products. Universities are a very special case, often willing to take interim versions because they want to be at the leading edge and because they have students around who like to play with things. Most commercial customers would not want to buy software that (a) was built on an interim release of X and (b) had to run on a non-supported server. The earlier release that ran on v10 in all probability (1) was begun when they thought v10 would be the real thing, (2) was done as an in-house exercise, and (3) was made available to those customers who really wanted it ( possibly with a warning that it was a preliminary version?) because it was recognized that v11 was going to be a long way away. Software vendors hate with a passion having to do work over again; they would rather spend their dollars on producing the next generation of the product than on upgrading the internals because there's a new release of X. I'd say your vendor was (1) unusually honest, (2) unusually sensible, and (3) doing the right thing for the bulk of the customer base. The Sun server isn't that far way now; once it's out, ask your vendor again when they will have support for X11; I suspect they'll have an answer then. -- scott preece gould/csd - urbana uucp: uunet!uiucuxc!urbsdc!preece arpa: preece@Gould.com
mlandau@DIAMOND.BBN.COM (Matt Landau) (10/04/88)
Let me add another voice from the vendor's point of view. As much as I believe that the relatively wide availability of X11, and the relatively large number of people writing publically available software for X is A Good Thing, I also have to recognize that not much of that software is what you'd call "commercial quality" as far as either performance or robustness is concerned. Not that all software has to be of commerical calliber to be good -- I use, and write, public domain software for X11. But one expects more from a vendor selling software for money than from a university grad student or an interested hacker cranking things out for the fun of it, and X just isn't quite "there" enough to make writing and selling software reasonable yet. For example, my real job is to work on a complex, featureful document authoring system for workstations. Sure, we'd love to ship an X11 version of the system today, and customers have certainly asked about availability. We have to put them off, telling them "not quite yet." Why? A couple of reasons. For one thing, the performance of the MIT sample servers (which is what most people are using) just isn't good enough to support really complex software. We know - we've been there. We were a beta site for X11R1, and we've had prototype versions of our product under X11R2 for months now. They just don't perform well enough that you'd want to use them, and we're not willing to sell them under those conditions. (And if we did sell them as they are right now, you can be we'd have people lambasting us for selling "slow, buggy software" instead of for delaying on shipping the X version; basically, if you're a software developer, you can't win :-) Another thing that delays people -- well, us at least -- is waiting for the ICCCM conventions to gel. When we ship an X product, we want to be as sure as possible that it's a good X citizen, interacts correctly with ICCCM compliant window managers and other applications, etc. As it is, the rules are changing fast enough that anything we implement today probably isn't going to work by the time the ICCCM spec gets out of Consortium review. We just don't believe it's wise to ship a system today that's going to work correctly with some window managers, some of the time. Sure, we take a bit of a black eye for not shipping X11 versions of our software as soon as some people would like. But we'd take the same black eye in any case, because the customers wouldn't like what it's possible to ship right now. So what're you gonna do? We opted for waiting until we can ship systems we're happy with. Other venders may make other choices, but don't fault the ones who hold out for quality. [Needless to say, these are my opinions, and not necessarily those of BBN.] -- Matt Landau Waiting for a flash of enlightenment mlandau@bbn.com in all this blood and thunder
paulsh@denali.gwd.tek.COM (Paul Shearer) (10/04/88)
Bob, Use the x10tox11 converter and you won't have to wait for his port. You can run X10 clients on your X11 server. Paul Shearer Tektronix
bob@allosaur.cis.ohio-state.edu (Bob Sutterfield) (10/05/88)
It has been suggested that I share the name of that dozing vendor in this forum. That seems inadvisable and perhaps rude, at least until I've exhausted all possible avenues in private discussion with the company itself. I think I appreciate the quality-control issues that may be at stake in the company's development cycle, but I still maintain that if they are shipping software that runs on a freely- available though commercially unsupported server platform now (X10.4) then they should not have any qualms about doing the same with X11. In article <8810040110.AA04391@denali.GWD.TEK.COM> paulsh@denali.gwd.tek.COM (Paul Shearer) writes: >Use the x10tox11 converter and you won't have to wait for his port. >You can run X10 clients on your X11 server. x10tox11, while a nifty idea, is simple-minded with the protocol and slow. That slowness, combined with the already sludgy performance of X11R2 on a Sun-3/50 (though it's much better with gcc and the Purdue mods), would make the software very unpopular. But perhaps it's a way, at least, to get on down the road. I'll suggest that the more adventurous users give it a try. Another problem is that this particular X10 client causes x10tox11 to die with a segmentation fault when it exits. It's ugly, but it seems that the work gets done before then. -=- Zippy sez, --Bob Didn't I buy a 1951 Packard from you last March in Cairo?
UEJIOWH%ASLCLU.DECnet@GE-CRD.ARPA (10/06/88)
Bob Sutterfield's comments concerning Software Houses was right on the mark!! I just finished talking to a SW vendor who claimed that they were waiting for the X Protocol and Xlib to settle down enough to use. They also were waiting for HW vendor support before moving to X11 even though we are using their X10 product on the MIT tape. I personally think they are waiting for either the HW vendors or their customer's to pay them to do the port. Maybe if I were in their shoes, I would do the same. Wayne Uejio These are my own opinions and not those of General Electric.
toddb@tekcrl.crl (Todd Brunhoff) (10/06/88)
>> x10tox11, while a nifty idea, is simple-minded with the protocol and slow. Not so much anymore. It has been 60% rewritten since r2 to (try) to cooperate with window managers, do more optimal buffering and flushing, and cache a great deal of information. Many things work now that did not work before, such as pikapix, xwd, xwud. All things run faster. The worst performance degredation I have seen on the current version is 30%. In some cases there is none at all (mathscribe, a very slick WYSIWYG algebra system does some plotting that starts appearing 6 seconds after the computing starts (T0) and finishes at 11 seconds after T0. Under the protocol converter, it starts appearing 10 seconds after T0 but finishes at 11 seconds after T0). Many of the simple-minded problems have been fixed, like handling all uses of UnmapTransparent (I count 3); not getting upset when an x10 client does a GrabMouse on an unmapped window (illegal in X11); adding delayed write to slow clients (like X11 has); ridgedly adhering to x10's ability to paint text, propogating non-device events up the tree like x10 does (x11 does not), etc, etc. >> Another problem is that this particular X10 client causes x10tox11 to >> die with a segmentation fault when it exits. It's ugly, but it seems >> that the work gets done before then. Ugly indeed. If you wish, I can send you the newest version now (it will also be available when R3 appears), to see if this helps. I'd dearly love to fix any remaining problems. --------------- Usenet: {ucbvax,decvax,allegra,uw-beaver,hplabs}!tektronix!crl!toddb {CS,ARPA}net: toddb%tekcrl.crl.tek.com@relay.cs.net c--Q Q US: Todd Brunhoff; Visual Systems Lab; Tektronix, Inc. ` Box 500 MS 50-662, Beaverton OR 97077 - Phone: (503) 627-1121
bcripe@hpcvlx.HP.COM (Brian E. Cripe) (10/07/88)
> but I still > maintain that if they are shipping software that runs on a freely- > available though commercially unsupported server platform now (X10.4) > then they should not have any qualms about doing the same with X11. You have never been bitten, learned your lesson, and changed your policy? Oh to be so fortunate. Brian Cripe Hewlett-Packard bcripe@hp-pcd