home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.windows.x
- Path: sparky!uunet!utcsri!newsflash.concordia.ca!sifon!thunder.mcrcim.mcgill.edu!sobeco!macx.sts.sobeco.com!user
- From: blimoges@sobeco.com (Bertrand Limoges)
- Subject: Re: XVT toolkit??
- Organization: Sobeco Ernst & Young
- Date: Wed, 11 Nov 1992 16:16:18 GMT
- Message-ID: <blimoges-111192105111@macx.sts.sobeco.com>
- Followup-To: comp.windows.x,comp.windows.x.motif,comp.windows.misc,comp.sys.misc,comp.sys.mac.misc,comp.unix.misc
- References: <1992Nov6.173238.6810@atinc.uucp> <BxCptF.Jw1@world.std.com>
- Sender: usenet@sobeco.com
- Nntp-Posting-Host: macx.sts.sobeco.com
- Lines: 69
-
- In article <BxCptF.Jw1@world.std.com>, kevyn@world.std.com (Kevyn Ford)
- wrote:
- >
- >
- > > I would like inputs from anyone who has used the XVT Portability Toolkit.
- >
- > One of my projects evaluated the XVT Toolkit against using the native GUI
- > for our VAXstations ... as far as producing a portable GUI application XVT
- > is fairly good.
- >
- > However, while there wasn't any one problem that we encountered with XVT, it
- > is not as robust as the native GUI system (which is to be expected) and we
- > found that there were many "small" features lacking that, all together, gave
- > the application an "unfinished" look (you can actually see some of this in
- > their XVT-Design tool itself -- which is brain-dead).
- >
- > We have found that XVT is more suited for "simpler" interfaces, and it will be
- > used as such, but for one particular application we've decided to use the
- > DECwindows/Motif interface.
- >
- > In short, XVT is a good portable GUI system for apps that don't really need
- > anything more than a few pull-down menus and graphical controls.
-
- I've been using XVT for almost a year now, and I must say that it is
- suitable
- for applications that have much more than "a few pull-down menus and
- graphical
- controls". Let's first address the issue the original poster was concerned
- about: portability between Motif, Mac, and OS/2 PM (very dissimilar GUI's).
- Well
- I must say that XVT does allow you to support these platforms, and if you
- know
- one of them well, it will be no problem to learn the XVT API, and it beats
- learning Mac, Windows, PM, Motif, OpenLook, ... Incidently, I started from
- a sound knowledge of the Mac toolbox, and worked from there.
-
- That is not to say that XVT is without problems. As you may guess, any API
- which promises to be portable between so many platforms has to adopt
- somewhat of a common denominator approach, which means that all native GUI
- features are not implemented within the XVT API. The main problem with that
- is not really loss of functionality, but rather loss of adherence to the
- standard human interface for that platform. However, note that you can use
- any feature of the native GUI in a non-portable way, and these can be
- coded in separate modules to minimize the mixing of portable and
- non-portable
- code. BTW, I run some XVT applications on a MacSE, and performance is more
- than acceptable.
-
- XVT does have several problems. It's a product in full evolution, and you
- have to contend with some bugs. However the XVT technical help is very
- good, bug reports are released, and bug fixes are shipped regularly. A lot
- was cleaned up between version 2 and version 3, and you have to look at
- where the product is going. Higher level API's built on top of XVT are
- already available, and OO application framework (a la MacApp) are in the
- pipes. I have had some frustrations migrating from the rich Mac toolbox to
- the XVT API, but it makes me smile everytime I take that code from my Mac,
- and just compile away on our RS6000 with Motif.
-
- I could say more about the pros and cons of XVT, but my "baud rate" at the
- keyboard, and net bandwidth precludes it.
-
- Bertrand Limoges, Sobeco Ernst & Young Inc., Montreal, Canada
-
- My opinions, not those of my employer.
- ---
- We extend and amplify impoverished data on the basis of what we know about
- how
- the world typically behaves, and the result is that prejudice becomes a
- natural part of "rational" thought. --Jeremy Campbell
-