1. Forum
  2. FidoNet
  3. OS2PROG

  • Bug in internal DosFindF

    From Lee Aroner@1:343/41 to Jonathan De Boyne Pollar on Tue Mar 27 14:57:02 2001
    The DOS support in OS/2 does appear to use a badly written FindFirst/FindNext call, however it has a serious bug that
    results in a return of a maximum of about 38,000 files per call.

    JdBP> That's interesting. From what did you deduce this ?


    From an easily reproducible error from a search routine in a DOS
    app I wrote years ago, which we had a discussion about two years
    or so ago.

    I don't recall the details at the moment, but it can be
    reproduced at will. A search on a drive with a known number of
    files will produce a maximum of xxxx files under OS/2, but yield
    all of em under Win or DOS. No error results, just completes
    normally, whistling against the lamppost, as if it had produced
    everything that was there.

    I seem to recall getting an acknowlegement of the problem from
    IBM, and indifference as to fixing it.


    LRA


    -- SPEED 2.01 #2720: This button doesn't do anything. Press it again ...

    --- Maximus/2 3.01
    * Origin: Top Hat 2 BBS (1:343/41)
  • From Eddy Thilleman@2:280/5143.7 to Lee Aroner on Thu Mar 29 02:10:01 2001
    Hello Lee,

    Tuesday 27 March 2001 22:57, Lee Aroner wrote to Jonathan De Boyne Pollar:

    The DOS support in OS/2 does appear to use a badly written
    FindFirst/FindNext call, however it has a serious bug that
    results in a return of a maximum of about 38,000 files per call.

    JdBP>> That's interesting. From what did you deduce this ?

    From an easily reproducible error from a search routine in a DOS app I wrote years ago, which we had a discussion about two years or so ago.

    This demonstrates that you didn't read my original post at all. :)

    I don't call FindFirst/FindNext (not in this app) at all, and the problem is an
    entirely different one and it's not in a DOS box at all. The issue is a lower speed with a compiled Pascal OS/2 program compiled with Virtual Pascal/2, compared with the same Pascal program as DOS program compiled with Borland Pascal and the same program implemented in REXX code.

    Should I repost my message just for you?


    Greetings -=Eddy=-

    email: e.thilleman@freeler.nl
    e.thilleman@hccnet.nl

    ... OS/2 fixes broken windows.
    --- GoldED/2 3.0.1
    * Origin: The only life man has created is the computer virus. (2:280/5143.7)
  • From Lee Aroner@1:343/41 to Eddy Thilleman on Sat Mar 31 09:11:00 2001
    The DOS support in OS/2 does appear to use a badly written
    FindFirst/FindNext call, however it has a serious bug that
    results in a return of a maximum of about 38,000 files per call.

    JdBP>> That's interesting. From what did you deduce this ?

    From an easily reproducible error from a search routine in a DOS app I wrote years ago, which we had a discussion about two years or so ago.

    This demonstrates that you didn't read my original post at all. :)

    ...and the original response, which you didn't read at all, WAS
    NOT TO YOU, IT WAS TO SOMEONE ELSE WHO *DID* COMMENT ON THE DOS
    FINDFIRST/FINDNEXT.

    Got it?

    Feeling a bit snippy today Eddy?

    LRA


    -- SPEED 2.01 #2720: It's as bad as you think and they are out to get you.

    --- Maximus/2 3.01
    * Origin: Top Hat 2 BBS (1:343/41)
  • From Jonathan de Boyne Pollard@2:257/609.3 to Lee Aroner on Thu Mar 29 00:16:14 2001
    The DOS support in OS/2 does appear to use a badly written
    FindFirst/FindNext call, however it has a serious bug that
    results in a return of a maximum of about 38,000 files per call.

    JdBP>> That's interesting. From what did you deduce this ?

    From an easily reproducible error from a search routine in a DOS
    app I wrote years ago, which we had a discussion about two years
    or so ago.

    We did ? I don't remember that. Was it in this echo ? What was the thread subject ?

    » JdeBP «

    --- FleetStreet 1.22 NR
    * Origin: JdeBP's point, using Squish <yuk!> (2:257/609.3)
  • From Eddy Thilleman@2:280/5143.7 to Lee Aroner on Mon Apr 2 12:54:56 2001
    Hello Lee,

    Saturday 31 March 2001 17:11, Lee Aroner wrote to Eddy Thilleman:

    This demonstrates that you didn't read my original post at all. :)

    ...and the original response, which you didn't read at all, WAS
    NOT TO YOU, IT WAS TO SOMEONE ELSE WHO *DID* COMMENT ON THE DOS
    FINDFIRST/FINDNEXT.

    Lesha Tsoorgaev wrote a reply on 19 March 2001 to which you replied. Is that the reply you're referring to? I have read that message and replied to it, because that was a reply to me. :)

    Feeling a bit snippy today Eddy?

    Yep, I felt a bit snappy, sorry about that. :)


    Greetings -=Eddy=-

    email: e.thilleman@freeler.nl
    e.thilleman@hccnet.nl

    ... Borg, James Borg. Vodka Martini, Gin is irrelevant.
    --- GoldED/2 3.0.1
    * Origin: Windows is great if you can make it work. (2:280/5143.7)
  • From Lee Aroner@1:343/41 to Jonathan De Boyne Pollar on Wed Apr 4 07:58:12 2001
    The DOS support in OS/2 does appear to use a badly written
    FindFirst/FindNext call, however it has a serious bug that
    results in a return of a maximum of about 38,000 files per call.

    JdBP> JdBP>> That's interesting. From what did you deduce this ?

    From an easily reproducible error from a search routine in a DOS
    app I wrote years ago, which we had a discussion about two years
    or so ago.

    JdBP> We did ? I don't remember that. Was it in this echo ? What was the thread
    > subject ?


    As I recall, it was in the general OS/2 echo, but maybe not.
    (Getting old sucks...).

    Seems to me it came up as part of a discussion about using the
    DOS FCB hack for mass deletions. At least that was what I joined
    in for...

    If I get a chance in the next day or 2 (and don't forget), I'll
    run the app on one of my drives and see if I can figure out where
    the wall is.

    LRA


    -- SPEED 2.01 #2720: My other tagline is being used on my CRAY.

    --- Maximus/2 3.01
    * Origin: Top Hat 2 BBS (1:343/41)
  • From Lee Aroner@1:343/41 to Eddy Thilleman on Wed Apr 4 07:44:16 2001
    This demonstrates that you didn't read my original post at all. :)

    ...and the original response, which you didn't read at all, WAS
    NOT TO YOU, IT WAS TO SOMEONE ELSE WHO *DID* COMMENT ON THE DOS
    FINDFIRST/FINDNEXT.

    Lesha Tsoorgaev wrote a reply on 19 March 2001 to which you
    > replied. Is that the reply you're referring to? I have read
    > that message and replied to it, because that was a reply to
    > me. :)

    And I replied to HER because she brought up the DOS call...


    Feeling a bit snippy today Eddy?

    Yep, I felt a bit snappy, sorry about that. :)

    I have days like that too. Don't let it get ya...

    LRA


    -- SPEED 2.01 #2720: ACME Exploding Tagline...5....4....3...<BOOM!>..

    --- Maximus/2 3.01
    * Origin: Top Hat 2 BBS (1:343/41)
  • From Bob Jones@1:343/41 to Lee Aroner on Wed Apr 4 08:03:26 2001
    From an easily reproducible error from a search routine in a DOS
    app I wrote years ago, which we had a discussion about two years
    or so ago.
    ...
    As I recall, it was in the general OS/2 echo, but maybe not.
    (Getting old sucks...).
    ...
    Seems to me it came up as part of a discussion about using the
    DOS FCB hack for mass deletions. At least that was what I joined
    in for...

    If I get a chance in the next day or 2 (and don't forget), I'll
    run the app on one of my drives and see if I can figure out where
    the wall is.

    You may be able to find it on this BBS.... two years is pushing the archive limit for my current message base. You may have to switch over to my read-only
    archive section.... [Drats... Read-only archives aren't currently on-line.]

    [I really should take this to the local message area.... But if it is still in
    the current echo's archive, you can just bring up the message and quote it from
    the BBS.]

    Hmmm... A quick search of this area for DOS FCB didn't find it.... You may have given me a reason to restore read access to the old BBS message archives. Take care.....

    Bob Jones, Sysop, Top Hat BBS

    --- Maximus/2 3.01
    * Origin: Top Hat 2 BBS (1:343/41)
  • From Lee Aroner@1:343/41 to Bob Jones on Wed Apr 4 13:27:00 2001
    From an easily reproducible error from a search routine in a DOS
    app I wrote years ago, which we had a discussion about two years
    or so ago.
    ...
    As I recall, it was in the general OS/2 echo, but maybe not.
    (Getting old sucks...).
    ...
    Seems to me it came up as part of a discussion about using the
    DOS FCB hack for mass deletions. At least that was what I joined
    in for...

    If I get a chance in the next day or 2 (and don't forget), I'll
    run the app on one of my drives and see if I can figure out where
    the wall is.

    You may be able to find it on this BBS.... two years is
    > pushing the archive limit for my current message base. You

    Hmmm... A quick search of this area for DOS FCB didn't
    > find it.... You may have given me a reason to restore read
    > access to the old BBS message archives.
    > Take care.....

    Don't fret it Bob. I suspect Johnathan has access to full
    archives through Andy, if he really wants it.

    LRA


    -- SPEED 2.01 #2720: AWWW, come on, just this one last little feature.

    --- Maximus/2 3.01
    * Origin: Top Hat 2 BBS (1:343/41)
  • From Eddy Thilleman@2:280/5143.7 to Lee Aroner on Fri Apr 6 05:01:45 2001
    Hello Lee,

    Wednesday 04 April 2001 14:44, Lee Aroner wrote to Eddy Thilleman:

    And I replied to HER because she brought up the DOS call...

    OK.


    Greetings -=Eddy=-

    email: e.thilleman@freeler.nl
    e.thilleman@hccnet.nl

    ... I am Bene Gesserit of Borg: Fear is irrelevant.
    --- GoldED/2 3.0.1
    * Origin: If it ain't broke, don't FixPak it. (2:280/5143.7)