"Fun" with DO and Regex

B

BillMc

Guest
#1
Hello JP_tc_support,

I created an empty directory and added one file: a.b

I tried:

do i in ::a.b
echo %i
enddo

It worked. I then replaced ::a.b with a few other Regex
tests:

::.*a.b Failed
::a.b.* Passed
::a\.b Failed

They should have all passed.

--
Best regards,
Bill
4nt 8.02.106 / tcmd 9.02.151 cp 2.11.32 on xp pro sp3
 
B

BillMc

Guest
#2
. On Mon 28-Jul-08 5:31pm -0600, BillMc wrote:
.
. > I created an empty directory and added one file: a.b
. >
. > I tried:
. >
. > do i in ::a.b
. > echo %i
. > enddo
. >
. > It worked. I then replaced ::a.b with a few other Regex
. > tests:
. >
. > ::.*a.b Failed
. > ::a.b.* Passed
. > ::a\.b Failed
. >
. > They should have all passed.

I should have mentioned that I'm running these tests on the
latest 4nt - the above tests work in tcc10.

--
Best regards,
Bill
4nt 8.02.106 / tcmd 9.02.151 cp 2.11.32 on xp pro sp3
 
B

BillMc

Guest
#4
Here's another example:

do i in ::a.b
echo %i
enddo

do i in a*b
echo %i
enddo

Running under 4nt, you get:

a.b
a.b

Running under tcc you get:

C:\pad\temp\foo\a.b
a.b

Regular expressions and wildcards give different results.

--
Best regards,
Bill
4nt 8.02.106 / tcmd 9.02.151 cp 2.11.32 on xp pro sp3
 
B

BillMc

Guest
#5
BillMc wrote:
| I should have mentioned that I'm running these tests on the
| latest 4nt - the above tests work in tcc10.

By "latest 4nt" did you mean TCC 9.02.151 or 4NT 8.02.106?

I meant 4nt 8.02.106.

Bill
 

rconn

Administrator
Staff member
May 14, 2008
10,580
97
#6
BillMc wrote:

>
> I created an empty directory and added one file: a.b
>
> I tried:
>
> do i in ::a.b
> echo %i
> enddo
>
> It worked. I then replaced ::a.b with a few other Regex
> tests:
>
> ::.*a.b Failed
> ::a.b.* Passed
> ::a\.b Failed
>
> They should have all passed.
I assume you're referring to 4NT v8. The RE library was updated between
v8 and v9; all your tests work correctly in v9.

Rex Conn
JP Software
 
B

BillMc

Guest
#7
BillMc wrote:
I assume you're referring to 4NT v8. The RE library was updated between
v8 and v9; all your tests work correctly in v9.
.
Yes, 4nt v8.

Will you be updating the RE library in v8 at some point?

Thanks,
Bill
 

rconn

Administrator
Staff member
May 14, 2008
10,580
97
#8
BillMc wrote:

>
> Quote:
> Originally Posted by *rconn* View Post <showthread.php?p=1594#post1594>
> BillMc wrote:
> I assume you're referring to 4NT v8. The RE library was updated between
> v8 and v9; all your tests work correctly in v9.
>
> .
> Yes, 4nt v8.
>
> Will you be updating the RE library in v8 at some point?
I don't have any plans to do so at this time.

Rex Conn
JP Software
 
B

BillMc

Guest
#9
. On Thu 31-Jul-08 4:23pm -0600, rconn wrote:
. > BillMc wrote:
.
. >> Yes, 4nt v8.
. >>
. >> Will you be updating the RE library in v8 at some point?
.
. > I don't have any plans to do so at this time.

This seems a bit early to cut off support of 4nt v8, but its
your show.

There are some of us (I suspect I'm not the only one) that
prefer the 4nt v8 (herein just 4NT) interface to that of TCC
v9+ (herein just TCC). So I use 4NT until I need a TCC
feature and just run 'start tcc' where tcc is an alias.

The negatives of TCC to me are that the PgUp or PgDn give a
blank screen (with one entry) for a second or so before
populating with history (its GUI). The Ctrl-PgUp also
displays the directory history (also as a GUI instead of the
nice one in 4NT) - Ctrl-PgDn just beeps. Thanks for not
doing select in a GUI.

Your GUI approach seems natural when operating from within
TCMD. But for standalone or detached TCC, the integrated
approach of 4NT is more natural.

BTW, yesterday and today (I haven't rebooted), start
followed by TCC worked but a PgUp, PgDn or Ctrl-PgUp puts
the session out of apparent focus (the blue bar at top turns
light blue) - but hitting Escape to close the invisible
window shows that's its not really out of focus.

--
Best regards,
Bill
4nt 8.02.106 / tcmd 9.02.151 cp 2.11.32 on xp pro sp3
 

rconn

Administrator
Staff member
May 14, 2008
10,580
97
#10
BillMc wrote:

> . On Thu 31-Jul-08 4:23pm -0600, rconn wrote:
> . > BillMc wrote:
> .
> . >> Yes, 4nt v8.
> . >>
> . >> Will you be updating the RE library in v8 at some point?
> .
> . > I don't have any plans to do so at this time.
>
> This seems a bit early to cut off support of 4nt v8, but its
> your show.
I haven't cut off v8 support, but I have to balance the usefulness of
spending a week incorporating a new RE library for v8 (to address a
problem that nobody had reported until now, and which can be trivially
worked around by providing a path), versus spending that time on v9 & v10.


> The negatives of TCC to me are that the PgUp or PgDn give a
> blank screen (with one entry) for a second or so before
> populating with history (its GUI). The Ctrl-PgUp also
> displays the directory history (also as a GUI instead of the
> nice one in 4NT) - Ctrl-PgDn just beeps. Thanks for not
> doing select in a GUI.
You've either got a really sizeable history or a slow machine - my
history window (on a mid-range machine) pops up completely populated
(several hundred entries) in no more than .25 seconds.


> Your GUI approach seems natural when operating from within
> TCMD. But for standalone or detached TCC, the integrated
> approach of 4NT is more natural.
>
> BTW, yesterday and today (I haven't rebooted), start
> followed by TCC worked but a PgUp, PgDn or Ctrl-PgUp puts
> the session out of apparent focus (the blue bar at top turns
> light blue) - but hitting Escape to close the invisible
> window shows that's its not really out of focus.
The popup window is probably appearing behind the TCC window (old
Windows bug).

Rex Conn
JP Software
 
B

BillMc

Guest
#11
. On Fri 1-Aug-08 11:16am -0600, rconn wrote:
. > BillMc wrote:
.
. >>> BillMc wrote:
.
. >>>> Yes, 4nt v8.
. >>>>
. >>>> Will you be updating the RE library in v8 at some point?
.
. >>> I don't have any plans to do so at this time.
.
. >> This seems a bit early to cut off support of 4nt v8, but its
. >> your show.
.
. > I haven't cut off v8 support, but I have to balance the usefulness
. > of spending a week incorporating a new RE library for v8 (to address
. > a problem that nobody had reported until now, and which can be
. > trivially worked around by providing a path), versus spending that
. > time on v9 & v10.

Your priorities seem reasonable. It just seems that TCC v9 has enough
problems that those of us still using 4nt v8 are wanting normal fixes
taken care of - even if they are just now appearing.

BTW, v9 has the issue of inconsistency between the output for normal
wildcards and regular expressions (which shows full path) - v8 doesn't
have that bug.

. >> The negatives of TCC to me are that the PgUp or PgDn give a
. >> blank screen (with one entry) for a second or so before
. >> populating with history (its GUI). The Ctrl-PgUp also
. >> displays the directory history (also as a GUI instead of the
. >> nice one in 4NT) - Ctrl-PgDn just beeps. Thanks for not
. >> doing select in a GUI.
.
. > You've either got a really sizeable history or a slow machine - my
. > history window (on a mid-range machine) pops up completely populated
. > (several hundred entries) in no more than .25 seconds.

My four year old computer is a 2.8 GHz Pentium 4. I have less than
7,000 entries in history. It takes a little over 2 seconds to
populate - its nearly instant (around 1/4 sec) in 4nt v8.

. >> Your GUI approach seems natural when operating from within
. >> TCMD. But for standalone or detached TCC, the integrated
. >> approach of 4NT is more natural.
. >>
. >> BTW, yesterday and today (I haven't rebooted), start
. >> followed by TCC worked but a PgUp, PgDn or Ctrl-PgUp puts
. >> the session out of apparent focus (the blue bar at top turns
. >> light blue) - but hitting Escape to close the invisible
. >> window shows that's its not really out of focus.
.
. > The popup window is probably appearing behind the TCC window (old
. > Windows bug).

It is. However, in v9, when I drag the console half way down the
screen and hit PgUp or PgDn, the history window shows up starting
above the screen. Resizing it lets me see the edge of the blue bar so
I can drag it. Moving it away from the console doesn't stick (an
"old" tcc v9 bug).

For now, I'm staying with v8 because of the slow response time and
lack of usable history windows.

--
Best regards,
Bill
4nt 8.02.106 / tcmd 9.02.151 cp 2.11.32 on xp pro sp3
 
Aug 10, 2008
1
0
#12
. >> This seems a bit early to cut off support of 4nt v8, but its
. >> your show.
.
. > I haven't cut off v8 support, but I have to balance the usefulness
. > of spending a week incorporating a new RE library for v8 (to address
. > a problem that nobody had reported until now, and which can be
. > trivially worked around by providing a path), versus spending that
. > time on v9 & v10.

>Your priorities seem reasonable. It just seems that TCC v9 has enough
>problems that those of us still using 4nt v8 are wanting normal fixes
>taken care of - even if they are just now appearing.

Rex:

Your priorities are reasonable, yes. But it still leaves the owners of 4NT8 like this gentleman and myself out in the cold when it comes to fixing small nagging bugs.

If you don't intend to keep updating the old 4NT/TC/TCI code because it saves you time and effort, how about giving old users of those pieces of software the option to upgrade free or at a very, very reduced price to the latest version?

I'll apologize in advance for being blunt, but I prefer to be honest about this: I've been supporting you and your company for years and have been enjoying the usefulness of the software, but lately I've noticed a significant lack of return for dollar spent for new versions. Each "upgrade" seems to be a re-hash of the same software with some bugs fixed and the GUI twiddled around with. (to me, the end user, it doesn't matter much that 4NT/TC merged into one product.) To you, perhaps it matters more, as the complexity of your codebase should have dropped.

Perhaps my perception that the value of the software per upgrade has been dropping is incorrect. Maybe it's just me, but I don't get that feeling. :)

Cheers,

Mark
 

rconn

Administrator
Staff member
May 14, 2008
10,580
97
#13
swish wrote:


> Your priorities are reasonable, yes. But it still leaves the owners of
> 4NT8 like this gentleman and myself out in the cold when it comes to
> fixing small nagging bugs.
>
> If you don't intend to keep updating the old 4NT/TC/TCI code because it
> saves you time and effort, how about giving old users of those pieces of
> software the option to upgrade free or at a very, very reduced price to
> the latest version?
I never said I wasn't ever going to update the v8 code; I said I wasn't
going to update it for the one particular bug report (which had never
generated a real problem, and which has a trivial workaround).


> I'll apologize in advance for being blunt, but I prefer to be honest
> about this: I've been supporting you and your company for years and have
> been enjoying the usefulness of the software, but lately I've noticed a
> significant lack of return for dollar spent for new versions. Each
> "upgrade" seems to be a re-hash of the same software with some bugs
> fixed and the GUI twiddled around with. (to me, the end user, it doesn't
> matter much that 4NT/TC merged into one product.) To you, perhaps it
> matters more, as the complexity of your codebase should have dropped.
First, the complexity vastly increased in v9 (more than 100K lines of
new code in v10). V9 is far more than a "twiddle" of the GUI!

Each new version for the past few years has had a minimum of 200 new
features (v9 has a lot more). Each user will make his own decision on
whether the new features are important to him; if not there's certainly
no requirement to upgrade on each version. (Unlike many companies which
charge more to upgrade older versions, we've always charged the same
price to upgrade from any version to the latest.)

Rex Conn
JP Software