Jump to article
< >

Active GUI element

Static GUI element

Code

WPS object

File/Path

Command line

Entry-field content

[Key combination]

more

Driving into the future?

by Christian Hennecke, © May 2007

The Editor in ChiefChristian Hennecke is the Editor in Chief of the VOICE Newsletter. While being a geographer originally, he runs an IT consulting and service company.

Late last year Scitech decided to offer its SNAP graphics technology for sale after a period of much reduced activity. This caused shock and dismay among OS/2 users as it meant the end for support of current and future video hardware. Several months have passed and now Serenity Systems, Mensys, and eCo Software have announced an agreement regarding development of Panorama VESA, a VESA video driver. So far, so good.

The design as a VESA driver has a number of implications.

On the plus side the driver will work with any VESA-compliant video adapter which enables the developers to invest more time into other things than implementing support for particular cards. Moreover, it keeps development costs at a level that can be handled by Serenity Systems and Mensys.

On the downside more and more cards are not VESA-compliant. Apparently, manufacturers do not consider VESA support important enough.

In addition there are inherent limitations to this driver model. Supporting just the VESA interface will most likely allow only the most basic capabilities of a video adapter. SNAP offered custom resolution, multi-head, zoom, TV-out, and—experimentally—screen orientation (Pivot) support. All this requires using the advanced features of the video adapter's GPU.

Another issue is speed as the VESA mode does not put the GPU's acceleration capabilities to use. Previous VESA drivers like GenGRADD and SNAP in VESA mode left the system barely usable. As a remedy the Panorama developers have come up with a clever buffering strategy. Users of the first alpha version of the new driver report that compared to previous VESA drivers it is usable with, e.g., full window drag enabled, and we have verified this. Nethertheless, this is not comparable to full hardware acceleration.

As a result, the announcement raises a number of questions:

The answers to these questions are of paramount importance as they essentially determine whether eComStation remains a viable solution—for both users and vendors. As was already said in our main editorial: OS/2 users are used to limitations and Tolkienian struggle against decline but even they will only go so far. Let alone users of other systems. Vendors and resellers have to convince them that there is a business case for eComStation.

In the end it all comes to this: Is this a base for a future or a strategic withdrawal?

Editing: James Moe