doyle@doyled.b23b.ingr (Doyle C. Davidson) (05/30/91)
I am interested in obtaining some objective data about the performance
of Motif 1.1.* UIL (uid) relative to equivalent functionality
implemented in C.
Some of the things that I want to know is:
1) what kind of startup (load) impact is there
(Does XtCreateWidget() overhead outweight disk/UIL overhead?)
2) what are some limitations with using UIL
ex: limited data in callbacks and event handlers
3) internationalization issues
4) are there techniques that make some UIL file better than others
5) can I use my own widgets with UIL
6) use with respect to "form" builders
7) anything else that's important
Please respond via email to:
doyle@ingr.com
or
uunet!ingr!doyle
and I will eventually post a summary.
Thanks in Advance!
Doyle
-------------------------------------------------------------------
Doyle C. Davidson |
Intergraph Corp. | Intergraph - Everywhere you look!!
Third Party Software |
1 Madison Industrial Park |
Huntsville, AL 35806 | These comments are...
(205) 730-2000 |
| X-clusively my own.
..!uunet!ingr!doyle |
- or - doyle@ingr.com |
-------------------------------------------------------------------marbru@auto-trol.com (Martin Brunecky) (05/31/91)
In article <1991May30.160724.12989@infonode.ingr.com> doyle@doyled.b23b.ingr.com writes: >I am interested in obtaining some objective data about the performance >of Motif 1.1.* UIL (uid) relative to equivalent functionality >implemented in C. > >Some of the things that I want to know is: > 1) what kind of startup (load) impact is there > (Does XtCreateWidget() overhead outweight disk/UIL overhead?) Mrm calls XtCreate widget so everything UIL does is an add-on. > 2) what are some limitations with using UIL > ex: limited data in callbacks and event handlers (no comments, I hate it) > 3) internationalization issues UIL compiler gives you some ways to encode non-ISO-LATIN text If you have the source, you can put this code into your app. > 4) are there techniques that make some UIL file better than others > 5) can I use my own widgets with UIL They say yes. You just have to change compiler tables and rebuild the compiler. If you want to add new data types ..... > 6) use with respect to "form" builders > 7) anything else that's important Look at Wcl (look at FAQ list for more). Using UIL is BAD, the only thing even worse is doing it in "C".... > -- =*= Opinions presented here are solely of my own and not those of Auto-trol =*= Martin Brunecky marbru%auto-trol@sunpeaks.central.sun.com (303) 252-2499 (better avoid: marbru@auto-trol.COM ) Auto-trol Technology Corp. 12500 North Washington St., Denver, CO 80241-2404
connolly@convex.com (Dan Connolly) (06/04/91)
In <1991May31.165653.23064@auto-trol.com> marbru@auto-trol.com (Martin Brunecky) writes: >In article <1991May30.160724.12989@infonode.ingr.com> doyle@doyled.b23b.ingr.com writes: >>I am interested in obtaining some objective data about the performance >>of Motif 1.1.* UIL (uid) relative to equivalent functionality >>implemented in C. >> >> 7) anything else that's important > Look at Wcl (look at FAQ list for more). > Using UIL is BAD, the only thing even worse is doing it in "C".... I'd like to second this piece of advice. These interface languages are all about increasing productivity, no? Wcl does. UIL doesn't. That's all I've got to say. Dan
jhc@ivan.uucp (James H. Coombs) (06/04/91)
In article <connolly.675977637@convex.convex.com> connolly@convex.com (Dan Connolly) writes: >In <1991May31.165653.23064@auto-trol.com> marbru@auto-trol.com (Martin Brunecky) writes: > >>In article <1991May30.160724.12989@infonode.ingr.com> doyle@doyled.b23b.ingr.com writes: >>>I am interested in obtaining some objective data about the performance >>>of Motif 1.1.* UIL (uid) relative to equivalent functionality >>>implemented in C. >>> >>> 7) anything else that's important >> Look at Wcl (look at FAQ list for more). >> Using UIL is BAD, the only thing even worse is doing it in "C".... > >I'd like to second this piece of advice. These interface languages are >all about increasing productivity, no? Wcl does. UIL doesn't. That's all >I've got to say. In my experience, UIL does increase productivity. In fact, we have an interface designer who has shifted from designing exhaustively in a structured graphics editor to writing the UIL herself. This gives her the ability to fine tune the design in the presentation medium. It also means that software engineers don't have to do the work of converting a design into an implementation, and that eliminates some of the inevitable conflicts between a programmer's ideas and the designers ideas. And, of course, we can quickly modify an interface just by changing the UIL, which is a lot safer to work with and a lot faster to compile. We can even try out alternative interfaces from the same executable. These are all big wins. I have looked at Wcl. Last I checked, it was not vendor supported. There may be better tools than UIL, but UIL is definitely better than coding the interface in C or C++ (yes, we use C++ with UIL). I should also stress that UIL is not difficult to work with (although you may need an expert to work through the documentation and then teach everyone else how to use it). We even have a non-programmer who can design and implement a complete interface with UIL. --Jim
jerryl@is.Morgan.COM (Jerry Liebelson) (06/05/91)
According to James H. Coombs: > > In my experience, UIL does increase productivity... > > These are all big wins. I have looked at Wcl. Last I checked, it was > not vendor supported. ... > > --Jim > I realize that to many firms, that lack of vendor support is a big deal, but WCL is so basic in it's concept and implementation (only 1200 lines of code) that I don't think the issue is relevant. We've been using WCL and it works fine. -- |~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~| | Jerry Liebelson | EMAIL: jerryl@is.morgan.com | | Information Systems | uunet!is.morgan.com!jerryl | | Morgan Stanley, Inc. | VOICE: (212) 703-2409 | | 1633 Broadway 36th Floor | FAX: (212) 703-2371 | | New York, NY 10019 | | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
marbru@auto-trol.com (Martin Brunecky) (06/11/91)
In article <54461@inkwell.UUCP> jhc@ivan.UUCP (James H. Coombs) writes: >In article <connolly.675977637@convex.convex.com> connolly@convex.com (Dan Connolly) writes: >>In <1991May31.165653.23064@auto-trol.com> marbru@auto-trol.com (Martin Brunecky) writes: >> >>>> 7) anything else that's important >>> Look at Wcl (look at FAQ list for more). >>> Using UIL is BAD, the only thing even worse is doing it in "C".... >> >>I'd like to second this piece of advice. These interface languages are >>all about increasing productivity, no? Wcl does. UIL doesn't. That's all >>I've got to say. > >In my experience, UIL does increase productivity. In fact, we have an >.....(rest of WHY UIL is better than C deleted) > >These are all big wins. I have looked at Wcl. Last I checked, it was >not vendor supported. No, it is not vendor supported. But, forgive me, that is a very poor argument, considering the size of the Wcl code and the fact it does not do anything "unsupported". I'd be surprized if (even using UIL) your application had less than 5times more (and more involved) Xt code than the entire Wcl. -- Martin -- =*= Opinions presented here are solely of my own and not those of Auto-trol =*= Martin Brunecky marbru%auto-trol@sunpeaks.central.sun.com (303) 252-2499 (better avoid: marbru@auto-trol.COM ) Auto-trol Technology Corp. 12500 North Washington St., Denver, CO 80241-2404