ESET and BEGINLIBPATH/ENDLIBPATH

[KO Myung-Hun]

Hi/2.

ESET is a great feature of 4OS2. BTW, it does work with BEGINLIBPATH/ENDLIBPATH. Even if I modify them with ESET, its result is not applied actually.

[Steven Levine]

Hi,

Interesting defect. What you are seeing is the combination of a number of interacting defects/omissions.

get_list() does not supply a "varname=" prefix when returning values for BEGINLIBPATH etc. This confuses other code with expects the variable name to be returned as part of the returned string.

cmd_eset() does not check for nLength 0 (i.e. the variable name length), probably because it assumes that this can not occur.

put_list() does not notice that the supplied string does not contain a variable name or equal character. It probably should report this as an error even though the condition should never occur.

Given the above, put_list() does insert the string in the environment, but not in a useful way.

The fix should be relatively easy. I plan to modify get_list() to return the missing variable name and equals character.

[Steven Levine]

Please try

​ftp://ftp.netlabs.org/incoming/4os2-3.08-shl-2014-09-15.zip

The fix needed to go into cmd_eset(), but that's details.

[KO Myung-Hun]

I've tested. And I've confirmed that eset worked well with beginlibpath/endlibpath. But I've found another problem. After ESET BEGINLIBPATH/ENDLIBPATH, LIBPATHSTRICT is changed.

set beginlibpath=x:\path
set libpathstrict=T
eset beginlibpath [just enter, do not change contents]
set libpathstrict

Then, you can see [T:\path]. This is true for ENDLIBPATH, and does not occur original 3.08.

[Steven Levine]

​ftp://ftp.netlabs.org/incoming/4os2-3.08-shl-2014-09-24.zip

should resolve this. I neglected to nul terminate the value returned by DosQueryExtLIBPATH.

[KO Myung-Hun]

I've confirmed. It works well. Thanks.

Ah, BTW, my mail to you was blocked.

---

Connected to 209.86.93.229 but sender was rejected...Remote host said:>>>550 IP 203.252.1.122 is blocked by EarthLink?. Go to earthlink.net/block for details.

---

Last time, there was no problem.

[Steven Levine]

Corrected by PerForce changeset 37556.