Q_Liberator malaise

[RalfR]

Im aware that you tested the program, but not its material effects. That is the issue now,

I have just patched the Runtimes with Mark's program and tried a lot programs to verify, that nothing bad has happened or if something has happened. But just on QPC2.

[Artificer]

Has anyone tested this yet?

I have a Q60 80Mhz and had a little bit of time so tried patching 10 qliberated programs with disappointing results.

Of the 10, 3 were not patched, 2 would not run with copyback enabled but one was OK

Of the 7 patched, 3 crashed when tested before copyback was enabled, these were Qdock, q-trans and ScrapIcon.

Of the remaining 4, 3 seemed OK when run with copyback on.

Attached is the log file of cpushaid_bas

CPUSHAID.zip

(830 Bytes) Downloaded 22 times

[Peter]

I have just patched the Runtimes with Mark's program and tried a lot programs to verify, that nothing bad has happened or if something has happened. But just on QPC2.

Many thanks Ralf, it is important to validate there are no side effects. As you know I'm in the process of building a test Q60, which got more difficult than I hoped. But in the long run, providing equipment for people with some QLib background is a better stragtegy than becoming QLib tester myself. It's never just a 5 minutes test, things usually recurse from one issue to the next. I have to concentrate on hardware, which is difficult enough. The QLib malaise has been there for over 25 years and I don't mind a few more weeks or months. It's more important to approach things efficiently than hastily.

[Peter]

Of the remaining 4, 3 seemed OK when run with copyback on.

And they were not okay before the patch?

[pjw]

Has anyone tested this yet?

<>
Attached is the log file of cpushaid_bas CPUSHAID.zip

Great! Thanks a lot. I was just going to ask the same question as Peter: The programs that worked with copyback after patching, did they work with copyback before they were patched?

[pjw]

It would also be interesting to find out about the programs that did run unpatched. Whats special about them?

[Artificer]

And they were not okay before the patch?

Of the 4 that survived patching, nightfall_obj, QFonz_obj and 3D_obj did not run with copyback enabled before patching.

I think ScrapMC_obj may have been OK with copyback before patching but I dont use it often and feel it needs more time in use to be sure.

Cheers

[RalfR]

I have tried QTrans from Dilwyn which works under QPC2.

The program is only available with runtimes, Mark's program found a place, probably the runtimes.

The patched version of QTrans crashes with the error "invalid name" but the name is in unreadable characters.

You should definitely patch the runtimes and test programs that do not have any linked runtimes.

If it really is the runtimes (which we don't know for sure yet), I believe there is a utility program on the Minerva disk that can be used to patch QLib programs so that they don't use the linked runtimes.

[pjw]

Its the runtime, qlib_run that we should look to first, and which ought to be patched. Most programs, I believe, dont include the runtime but expect it to be pre-loaded at boot. If it turns out that that is where the problem lies, then only those programs that include the runtime need to be patched.

However, "we" should try to narrow the problem down: If its in the runtime then, great! the problem is solved, otherwise we need to think of where is best to look next.

(I just saw your post now, Ralf. So it appears we agree )

[Peter]

Most programs, I believe, dont include the runtime but expect it to be pre-loaded at boot.

Opposite for me, all QLiberated programs I use include the runtime.