dotSoftwaredotDevelopmentdotCustomersdotAbout us
PushOk logoblank
bullet Home
bullet My software
bullet Support
bullet My payments
bullet My info
bullet Subscriptions
bullet Voting
bullet Contact us
fast linksFast Links
news&eventsnews and events

2012-12-21 
Major update of SVN SCC plug-in - versions 1.7.2 are finaly released

2012-12-21 
Major update of SVNCOM version 1.7.2 are finaly released

Lightweight embedded Node.js database with MongoDB API.

Ticket

Search go
PushOk Logo blank
leftTicketright

Question

( CVSSCC )
Type: Public Status:Closed Created: 07 Apr 04 04:00 Updated: 19 May 06 05:00
--> Nathan (user)  at 19 May 06 05:00 writes

I read through those three tickets. They are all about CVS and I'm using
SVN but I tried to relate it to my issue non-the-less.

The first post was about a "do not resolve client name" option that seems
to be specifically in the options of CVS NT. I looked through Subversion
and Apache and didn't find anything related.

The second post explained the different access methods. I have noticed that
using file:/// instead of http:/// is much faster for obvious reasons. But
I need the authentication and eventually we will not have direct access to
the files.

The third post also talked about specific CVS NT options that I do not
have.

Are there any other tips that you might have? I can do any optimizations
on my server that would help out this situation if you know of any.

Let me give you an example of when something is slow. I have set the option
in Visual Studio to automatically check out a file when I start editing it.
As soon as I start typing it first checks out the file and then continues
to let me type. When I am using VSS this is near instant and only sometimes
am I interupted. When I am using PushOk it takes about 5-7 seconds to check
it out.

On a side note. What does it actually do during this process? Since
Subversion is not a check-out/check-in source control model but rather a
edit/merge/check-in model.

Thanks,

Nathan
--> Igor Pushkov (admin)  at 08 Apr 04 04:00 writes

Plug-in can work both with CVS NT (windows server) or with any *unix
server. To work with *unix server you just should configure some
options in plug-in. To do this open its configuration window, page CVS
options and press "Preset for UNIX". Also on the page File Types,
switch on option "treat unicode files as binary".
All this is needed because:
1. *unix work with case sensitive file names. And without option
"Check file names on server" some time can appears problems since IDE
can provide file name in wrong case (because this not have sense for
windows"
2. *unix server do not have support for "unicode" files, and they
should be considered as binary.
Look at this pages for installations details:
http://www.pushok.com/help/cvsscc/index.php?redirect=inst_plugin.htm
http://www.pushok.com/help/cvsscc/index.php?redirect=inst_plugin_unix.htm
http://www.pushok.com/help/cvsscc/index.php?redirect=inst_plugin_win.htm
And the last, basically there not difference which server is used, but
we prefer *unix server. For us it seems to be more stable, but in the
same time thousands of users use CVS NT.
--> Brian Estep (user)  at 07 Apr 04 04:00 writes

I am pretty new to CVS but I am very interested because of recent
frustrations with VSS. I understand that CVS runs best on a Linux/Unix
server. My question is, does your plug-in handle the communications
between the windows desktop (IDE) and the Linux/Unix (CVS) repository or
does the windows desktop need something else installed?
Rate this ticket:
Not useful at all
Partially useful
Useful
Very useful



You are 9752807 visitor since 20 Jan 2003.
2772 visitors today and 4 online right now.
blank left to top right blank

© Copyright by PushOk Software, 2003-2024, webmaster@pushok.com