Memory Use

G

GregTomkins

Active Member
Hi ProgressTalkers,

Boy, it has been going on 5 years since I posted to a Progress forum. What happened to the PEG? I can't get into it anymore. Anyways, here's my real question:

I heard a while back that someone, maybe Peter van Dam, had written something to analyze the PVM heap and help you track down unreleased TT handles and that sort of thing. Does anyone know if there is any truth to this, or if it is even feasible for outsiders such as us?

With all due respect, I'm not really looking for advice about WIDGET-POOL or the perils of dynamic objects or GC in v10 etcetera.

Cheers,
Greg from Vancouver, Canada, home of the next Olympics!
 
The closest I can think of is this from v9stuff, but I doubt its what you are after. There are a couple of dynamic object viewers somewhere on PSDN, and of course, the debugger.

For future reference, pdecode does some lowish-level tools.

My webpage has some more links, not updated recently.

The Peg, which is essentially dead as an online presence is discussed here - if you want to access it you need to go email.
 
Actually, most of the forums are now back on-line, just not fully functional such as searching. The traffic is still there, though.

If you have been away, you should also check out http://www.oehive.org ... no solution for the current question, but lots of other stuff.
 
tamhas said:
Actually, most of the forums are now back on-line, just not fully functional such as searching.
Click to expand...

Not meaning to be, or enjoying - flogging a dead forum, but from the point of view of a formerly paying web user like myself, the Peg is effectively dead.

eg. (several months after crash)

http://www.peg.com/forums/peg/200903/index.html (several days out of date)
http://www.peg.com/techpapers/index.html (page not found)
http://www.peg.com/utilities.html (page not found)

etc.
 
I realize that some people were focused on web access and they have been out of luck and remain in a second class state even now. The bulk of the PEG use is by people accessing it via e-mail and that has remained active throughout. Subscriptions work, email works, the rest will presumably come back eventually.
 
Thanks for the info everybody; much appreciated.

PVD's program looks like it relies on buffers, which might not exist for a TT, I suppose, but in reality always would. I think.

It's too bad PSC has such a dearth of basic tools that other products have and that you can get 100 versions of for free.

As for the PEG, that's also a sad commentary on the state of the Progress community, isn't it? Oh well, at least this forum is still around.
 
I believe that the tools you are looking for are actually built in to OpenEdge Architect in the current version. Can't swear to that because I have never felt the need since I found it easily enough to walk the session tree for the object type in the rare cases when I have been concerned about this. E.g., the NEXT-SIBLING sort of logic found in code like this http://www.cintegrity.com/node/62

Oh well, at least this forum is still around.

Actually, daily volume on the PEG dwarfs all other lists and I believe these days that the daily volume on PSDN is significantly higher than the traffic here and has been growing.
 
Aha, yes, nice paper; I think FIRST-OBJECT is exactly what's needed. Sadly, we're stuck back in 9.something with little hope of upgrading before I retire. It's really breathtaking how costly and difficult a PSC upgrade is for a Win32 client app spread across many customers and desktops.
 
The FIRST-OBJECT for ABL classes is mirrored by other similar constructs for other ABL entities. Like I say, I don't tend to go that way, but I think if you look up FIRST- and NEXT- in the manual you will find a number of different versions.

It's really breathtaking how costly and difficult a PSC upgrade is for a Win32 client app spread across many customers and desktops.

Only costly in terms of license costs if you have failed to pay your maintenance, otherwise it is free.
 
You must log in or register to reply here.
Back
Top