Jump to content

Home

Accessing the OJP Source (Pre-SVN Posts)


razorace

Recommended Posts

Yeah, I know what you mean. The setup is terrible but it's the only place that has a hosting deal that allows for our type of project.

 

Anyway, the server seems to be up for me. This seems to happen a lot with new accounts. Personally, I think he's overly paranoid. Using weird security proticals only makes it hard for us to access the repository.

 

I beleive there's a way to grap the raw data and then create a local mirror of the repository but I've never tried it. Keshire is the one to ask about that.

Link to comment
Share on other sites

  • Replies 114
  • Created
  • Last Reply

Ya you can grab the actual module and set up a local "sand box" type deal with it.

 

No fuss no muss. But as far as uploading your changes you'd have to lobby Razor to give you the web access to freeRepository.

 

first Download the raw CVS format (this includes all the ,v files).

 

then make a new module.

 

:local:C:\Location_Of_Raw_CVS\

 

what it should do is make a new module but import it from it from the raw cvs making it act like it was from the actual freeRepository server. From there you can checkout the module from this local version.

 

And here is the output.

 

C:\cvs.exe -q import -m OJP tcvs-vendor tcvs-release CVSROOT=:local:C:\OJPSDK

 

No conflicts created by this import

 

Success, CVS operation completed

 

C:\cvs.exe -q checkout -d jkasdk OJP

CVSROOT=:local:C:\OJPSDK

Link to comment
Share on other sites

Oh yeah, Brad, it sounds like it's probably something on freepository's end, I suggest you email the support email address for freepository (support@freepository.com, I think) about it. This seems to be a common issue with new accounts.

Link to comment
Share on other sites

TortoiseCVS

 

install it. Reboot.

 

unzip the OJP and cvsroot to where ever you want the source files.

 

right click| CVS| make new module

 

set protocol to locally mounted folder

 

set repository folder to directory where the CVSROOT is

 

click fetch list

 

set module to OJP

 

click OK

 

it will set up a sand box that works exactly like the server version.

 

read the tortoise instructions on how to do anythign else. You also need to know how to compile code as that is what it is.

Link to comment
Share on other sites

when you do the fetch list it pulls all the module names from the cvsroot folder.

 

OJPS is not one of them. Your trying to do this inside the OJP folder it looks like.

 

use the openjediproject-raw-cvs folder as the local directory. DO THE FETCH LIST! "SELECT" OJP! hit ok.

Link to comment
Share on other sites

First, you hit new post instead of reply. You can tell when it tells you it can't post without a subject.

 

Second, that would be because its .NET project files. Use Google to search for prjconverter.exe. After that direct all your questions to the coding forum.

Link to comment
Share on other sites

Originally posted by razorace

Yeah, I know what you mean. The setup is terrible but it's the only place that has a hosting deal that allows for our type of project.

 

Anyway, the server seems to be up for me. This seems to happen a lot with new accounts. Personally, I think he's overly paranoid. Using weird security proticals only makes it hard for us to access the repository.

 

I beleive there's a way to grap the raw data and then create a local mirror of the repository but I've never tried it. Keshire is the one to ask about that.

Terrible? Paranoid?

 

That's a bit harsh, razor ace. I consider myself very security-conscious, and am proud of the contribution freepository made or continues to make to thousands of projects across the globe since it went live in 1999.

 

Script kiddies try to crack freepository over a 1000 times per day. The security measures in place on the site protect your code from them.

 

Sserver is actually quite mainstream these days. The sserver connect method is now included by default in CVSNT, which is the binary used by TortoiseCVS. That's why you can now use TortoiseCVS to connect directly to your freepository.

 

When I began designing the secure remote command line interface to freepository, I wanted to use ssh. But that simply _couldn't_ work in the silo'ed fashion necessitated by the many thousands of repositories on the freepository server. Fortunately, Corey Minyard had recently released his sserver reference implementation. This was exactly what freepository needed: a secure pserver-like protocol.

 

I began using sserver when almost no one had heard of it. Today, sserver (natively or via plugin) is available in virtually every mainstream cvs cleint.

Link to comment
Share on other sites

It wasn't my intention to offend. Beyond the initial hassle of getting thru the security, the service is great.

 

Unfortunately, that initial hassle is pretty combersome. It seems like every single time someone tries to get their system set up for a direct connection something goes wrong. Weither this is due to human error or the security measures, I can't say, but either way, it really makes it hard to get new people to use the repository.

 

Anyway, you're right. My comment was overly harsh and I applogize.

Link to comment
Share on other sites

Originally posted by razorace

It wasn't my intention to offend. Beyond the initial hassle of getting thru the security, the service is great.

 

Unfortunately, that initial hassle is pretty combersome. It seems like every single time someone tries to get their system set up for a direct connection something goes wrong. Weither this is due to human error or the security measures, I can't say, but either way, it really makes it hard to get new people to use the repository.

 

Anyway, you're right. My comment was overly harsh and I applogize.

I was more suprised than offended.

 

Are the problems experienced similar? The instructions I have posted on the site now cover Eclipse, NetBeans, TortoiseCVS, cvsnt and precompiled sserver clients for Linux and Windows.

 

If there is a common area where everyone experiences trouble, that would be helpful to know.

 

jbminn

Link to comment
Share on other sites

The tortoiseCVS tutorial on the site uses the batch file method that's overly complicated and causes errors even when it works. The method I described on the first page of this thread tells you how to do it without resorting to that.

 

Secondly, I beleive the common problem area is during the log in and checkout process. I think the common error is a "connection refused". I've had multiple people get stuck at that point and have directed all of them to email you to see what the problem is. This happens even when I am able to connect to the repository so I suspect it has something to do with newly secured IDs.

 

I think the biggie is that newbies can't really handle the complicated process of either getting a direct connection or creating a local copy. In most cases, people just want to have a copy of the source but can't do so since trying to download the pure source only gives you the stuff in the root folder.

Link to comment
Share on other sites

  • 2 months later...
Originally posted by jbminn

If there is a common area where everyone experiences trouble, that would be helpful to know.

 

In my experience, I had to compile my own cvs client, since the version that ships with OSX (1.10) doesn't contain sserver support. I can e-mail you the binary if you wish to host it on freepository.com

 

Beyond that, I had the mysterious "connection refused" message, which was only cleared up after e-mailing you. Dunno what I did wrong, or why it needed fixing at that point.

Link to comment
Share on other sites

Originally posted by razorace

Some reminders for those that are actively committing changes to the repository:

 

I have the Mac version building now. The changes needed are very minor - mainly path delimiter changes like #include "..\blah" to #include "../blah" and removing some type redefinitions.

 

How do I go about getting these changes back into the repository?

 

Also, when I did my checkout earlier today, I got a source tree that had both "Basic" and "Enhanced" directories, but only the Basic tree had any source files in it. What should I be pulling down and building against?

Link to comment
Share on other sites

1. Do the changes still allow for it to compile on PCs and linux? If not, we're going to add some #if statements to get it to do so.

 

How about you contact me over email or an IM service and send me a copy of the code for me to test? Or have you already done so?

 

Anyway, once we've figured it out, I'll give you write access to commit the changes.

 

2. Anyway the repository is set up with both the Basic and Enhanced source codes as branched in the same /basic directory. The HEAD branch is Enhanced while the OJP_BASIC leads to Basic.

 

I did this so that it's much easier to merge Basic changes into the Enhanced code. It's probably the wrong way to do it but I'm a pretty big CVS newbie.

Link to comment
Share on other sites

Originally posted by razorace

1. Do the changes still allow for it to compile on PCs and linux? If not, we're going to add some #if statements to get it to do so.

 

They are all separated with #if MAC_PORT/#else/#endif. However, a lot of them can be applied safely to all ports, since they are syntax cleanup fixes that won't affect compilation on the other platforms. It should be clearer when you get the code.

 

How about you contact me over email or an IM service and send me a copy of the code for me to test? Or have you already done so?

 

I'll do so tomorrow or Monday and we'll take it from there.

Link to comment
Share on other sites

  • 11 months later...

Archived

This topic is now archived and is closed to further replies.


×
×
  • Create New...