Our VB6 guy was part of the last RIF (Reduction in Force). The work he did has been split between me and another developer. We often are both are making changes to projects at the same time. This isn't a problem with CVS since we are working in different areas. However VB6 seems to modify the Reference section and change the paths each time either of us touches a project. Since we don't have the exact same path setup for out source trees we run into merge conflicts on the vbp file all the time.
Is there any way around this other than the obvious method of changing our setup so we have the same directory structures?
I would suggest two things:
Don't commit the .vbp unless you add a file to the project.
Mark the .vbp as read-only and check it in as such into your repo. When users check it out, it should still be read-only which will prevent changes to reference paths (and seemingly random reordering of the file) from being saved. When you have to make a change to the project--make the file read-write, save the change and then make it read-only again before committing.
Depending on your version control system, it is possible to automate this problem away. Both Subversion and Mercurial support hooks - scripts that are triggered by certain events, like check out, update, or commit. We wrote a fairly simple script that was triggered on commit: it looked to see if there was a .vbp in the commit package, and if there was, ran a "normalisation" routine that
put all the .cls/.bas/.frm files at the top of the .vbp, in alphabetic order
put the references section in alphabetic order
lower-cased the reference paths
The rest of the file is left alone, since it's only the first three sections that VB seems to delight in messing about.
Consequently, most of the time when you commit, and haven't made any substantive changes to the .vbp, the hook script restores your .vbp file to a canonical, ordered, state (like a revert), which has the effect of removing it from the commit since it's no longer changed.
In the event that you do add a new file or reference to your project, the consistent alphabetical sorting of the VBP lines means that merge conflicts are avoided since your VC merge algorithm can easily and correctly detect the changes.
We wrote our script in Javascript and execute it using Windows Script Host since, for Windows boxes at least, this removes the dependency on an interpreter like Perl/Python.
Hooking up this script to new VB projects is a 30 second job. The advantage over other manual approaches is that you don't have make any conscious effort to deal with the VBP file. Just commit it with everything else and the script takes care of the rest.
Related
I'm hoping that this is in the right spot.
Basically, I have a set of files which are versioned in an SVN repository, and I am using TortoiseSVN. These files all need to be renamed so that their new name is equal to their current name, plus a prefix added to the front.
Tortoise has the ability to rename the files, but it must be done manually, file by file, through a GUI menu, and I would prefer a less tedious and less time-consuming solution.
I have seen on this site examples which use Windows-based solutions, but I do not expect these to work, as they would not change the filename through TortoiseSVN but through Explorer, which would cause TortoiseSVN to not recognize those files as being equivalent to the originals, even if we recognized them as such.
I saw another question, linked here: (tortoise svn mass rename files merge)
In which the asker says that the files were renamed with a windows tool and thus SVN did not recognize them, and the solution offered was to write a new script which would pick up the old name, replace it with the new name, and repeat for each entry. However, unless I'm misreading it, that solution would require going in and specifying the old and new name for every entry, which from my perspective could be just as time-consuming as doing the manual option in TortoiseSVN.
I considered another possibility in that I could use windows command line to make a copy of every file in the directory, renaming the copies to fit. Then I would SVN delete the old files and SVN add the new files. This would achieve the rename, but then the history association would be disrupted, as it was technically no longer the same file.
The problem I'm running into as I read different questions asking similar queries is that they seem to be built around specifying the start file name and the end file name for each entry. I want to do a blanket rename on every file, without needing to consider the file name before or after the execution.
As such, my investigation has brought me to ask this question: Given a set of known existing Files F, and a known, constant prefix P, can TortoiseSVN be made to automatically update every entry in F with the prefix P, but in a way that TortoiseSVN recognizes as its internal 'rename' process (rather than an exterior process which would, in its perspective, replace, rather than change, the given file), and without having to also define the resultant end name of each file in F (thus no need to specify old and new filename for every entry), so that each file can be updated without needing manual intervention for each one? Thank you for reading.
when writing a file using NSFileCoordinator i need to specify the correct NSFileCoordinatorWritingOptions. Although they are explained in detail, I am not sure when to use which one. The available options are:
NSFileCoordinatorWritingForDeleting
NSFileCoordinatorWritingForReplacing
NSFileCoordinatorWritingForMoving
NSFileCoordinatorWritingForMerging
For example, what option is the correct one if I want to create a file (a plist for example)?
Wich one when I modify a file?
Can someone explain the NSFileCoordinatorWritingOptions for a better understanding?
I agree, documentation is not complete on that front, and hard to understand. And no sample code is available even for basic operations like these.
I try to think of these options in the perspective of other apps that have that specific file open, that helps getting the whole picture.
Pass no option (0) to simply update the file and notify others of your changes.
Let's say you are deleting a file that TextEdit currently displays, by providing the NSFileCoordinatorWritingForDeleting option, you're telling TextEdit to close the file as it does not exist anymore (or it could propose to save it to another place if it's in memory). It acts because of deletion.
If you're overwriting a file (as opposed to updating a file), you want about that same behavior for other apps. That's NSFileCoordinatorWritingForReplacing.
NSFileCoordinatorWritingForMoving says other apps to track the file to it's new location, so that it can be later updated.
NSFileCoordinatorWritingForMerging asks other processes to first commit their changes so that you can then merge your own changes with those.
To answer your question, you should use NSFileCoordinatorWritingForReplacing when creating a new file (even when no file exists, as it was to appear in the mean time from another app, you'd be replacing it with your own, unrelated contents). And NSFileCoordinatorWritingForMerging should be used when updating an existing file with new data, as it allows integrating the latest changes to that file immediately (instead of doing later with conflict resolution).
As I create more applications, my /code/shared/* increases.
this creates a problem: zipping and sending a project is no longer trivial. it looks like my options are:
in Xcode set shared files to use absolute path. Then every time I zip and send, I must also zip and send /code/shared/* and give instructions, and hope the recipient doesn't have anything already at that location.
this is really not practical; it makes the zip file too big
maintain a separate copy of my library files for each project
this is not really acceptable as a modification/improvements would have to be implemented everywhere separately. this makes maintenance unreasonably cumbersome.
some utility to go through every file in the Xcode project, figure out the lowest common folder, and create a zipped file structure that only contains the necessary files, but in their correct relative folder locations, so that the code will still build
(3) is what I'm looking for, but I have a feeling it doesn't as yet exist.
Anyone?
You should rethink your current process. The workflow you're describing in (3) is not normal. This all sounds very complicated and all basically handled with relative ease if you were using source control. (3) just doesn't exist and likely never will.
A properly configured SCM will allow you to manage multiple versions of multiple libraries (packages) and allow you to share projects (in branches) without ever requiring zipping up anything.
I'm new to XCode and I find the file management a huge pain. In most IDEs, you can simply have the project source tree reference a directory structure on disk. This makes it easy to add new files to your project - you simply put them on disk, and they will get compiled automatically.
With XCode, it appears I have to both create the file and separately add it to the project (or be forced to manipulate the filesystem through the UI). But this means that sharing the .xcodeproj through source control is fraught with problems - often, we'll get merge conflicts on the xcodeproj file - and when we don't, we often get linker errors, because during the merge some of the files that were listed in the project get excised. So I have to go and re-add them to the project file until I can get it to compile, and then re-check in the project file.
I'm sure I must be missing something here. I tried using 'reference folders' but the code in them doesn't seem to get compiled. It seems insane to build an IDE that forces everyone to modify a single shared file whenever adding or removing files to a project.
Other answers notwithstanding, this is absolutely a departure from other IDEs, and a major nuisance. There's no good solution I know of.
The one trick I use a lot to make it a little more bearable — especially with resource directories with lots of files in them — is:
select a directory in the project tree,
hit the delete key,
choose "Remove References Only", then
drag the directory into the project to re-add it.
This clobbers any manual reordering of files, but it does at least make syncing an O(1) operation, instead of being O(n) in the number of files changed.
I'm intrigued which IDEs you're using that automatically compile everything in a directory, as no IDE I've ever used does that (at least for C++). I think it's pretty standard to have a project file containing a list of all the files. Often you may want to only include certain files for different targets, have per-file compiler settings, etc.
Anyway, given that that's how it does work, you really shouldn't have too many problems from merge conflicts. The best advice would be commit early and often so that you don't get out of step with other people's changes. Merely adding files to the project shouldn't result in a conflict unless they happen to be added at exactly the same point in the project tree. We've been using Xcode in our team for years and we very rarely get conflicts: only if someone has restructured the project.
Fortunately, because the Xcode file format is text, it's generally quite easy to resolve conflicts when they occur, unlike the Bad Old Days of Codewarrior with it's binary format.
We have a StarTeam view that has two files that are in the "Unknown" state – does anyone understand why they are in this state and/or how we can get them out of the state?
Is deleting them and re-adding them with a different name the only solution?
Note that if you check out these two files (regularly or using “force checkout”), they will always be listed as “Unknown” (annoying).
Thanks.
More info based on Craig's suggestions below:
a) Calculating file status using MD5 checksum: same results ("Unknown" status)
b) These two files in question have only one revision on the server. I'm not sure if this is because our CM group attempted to fix the issue by deleting and recreating the files or if there really was just one revision. The files are text files.
c) I tried deleting the files on my local machine and refreshing the status. When I do this, instead of seeing the two files listed as "Unknown", I see a total of four files listed with a status of "Missing". There are two entries listed for each file - each pair has the same file name, folder path, "modified by", and "file stamp at time of check in". I have no idea why each file is listed twice. If I select each entry in a pair and select "Compare Contents", my diff tool says they are identical.
I have this same weird issue with the four files whether I use the MD5 checksum compare or non MD5.
If I try to checkout all four of the Missing files, I get two alerts prompting me to merge the files. I say no, the files are now on my local file system and the status is back to where I started - two files listed as "Unknown".
Craig's update:
You're definitely on to something. I moved each of the duplicate items to another directory. That immediately solved the issue in that I now could checkout the four items (two in same dir and two in new dir) without any "Unknown" items. I then deleted the two items that I moved to the new directory.
In doing this, I saw some more info. We somehow have a directory structure like this:
Parent_Dir
--SubDir1
--SubDir1
--SubDir1
--SubDir1 <- Two items were here
--SubDir1 <- Two items were here
--SubDir2
--SubDir3
--SubDir4
--SubDir5
Somehow we have five sub directories with the same name and these two files in question existed in two of the sub dirs with the same name.
The issue appears to be resolved. Do you think I should manually delete the extra sub directories?
Thanks to Craig this issue appears resolved. I have no idea how this situation was created (anyone?) but.. we're good now. Thanks Craig!
It would be a huge mistake to delete the files before you determine where the problem lies. Deleting and re-adding the file would kill the history and any links to the file. It also might not fix the problem, due to the way that StarTeam works internally (with a Native-II repository, anyway). Deleting the file and re-adding the identical file will not actually update anything in the repository, except the pointers to that revision. The revision itself would have stayed there when you deleted the file, and re-adding it would just create a new pointer to that revision.
If you haven't done so already, I strongly recommend telling StarTeam to calculate the file status using an MD5 checksum. Do this in the client via Tools->Personal Options->File->Use file checksums to calculate status. Then try update status again. This is not the default setting (in at least some versions of StarTeam), so it's worth checking. If you have not already done this, it may, by itself, fix the problem.
The first thing to do is to determine if the revision is valid on the server. If the files are text, the easiest way to do that is to compare contents between that revision and the previous revision. If it does turn out that the revision is corrupt, then the best solution would be to check out an earlier revision, and then force check in. This way you preserve the history of the file.
If the file appears to be OK on the server, then test it locally, by comparing contents. If the file is corrupt locally, do feel free to delete the file locally and then check it out again. Unlike deleting the file on the server, you don't lose anything by doing this, except local revisions.
If these suggestions do not fix the problem, I still would not recommend deleting the files on the server. Tell me the results of your investigations back here, and we'll see where we can go from there. It is, in my opinion, nearly always mistake to kill history.
Update
Based on the updated information in the post, I'm getting a better sense of what is going on here. It is likely that there are two items which point to the same file, and have the same name. "Item" is a concept in StarTeam related to the fact that a single file, change request, requirement, etc. can live in multiple places at once. For example, you can have a single file in two different views or projects.
Generally, you do not have the items with the same name in the same folder. But it can happen. And that probably explains the "Unknown" status. When you tell StarTeam to compare the file on disk with the item of the same name on the server, perhaps it can't figure out which item it should look at.
The first thing I would try is to try and drag one of the two items somewhere else. If that fixes the problem in the folder in question, you can delete the item elsewhere, without affecting the item in the folder. If, on the other hand, dragging one of the items elsewhere causes them both to move, it's easy to drag the item(s) back to where they came from.
Update 2
Do you think I should manually delete the extra sub directories?
Yes, but just as with the files, move them first, and make sure the subdirectories you'd like to keep are unaffected before deleting them.
A good way to identify this issue if it comes up again is to click on the root folder, add folder path to the shown columns, then click the all descendants button. Sort the files by name, and look for the unknown files and see if multiple files of the same name are checking out to the same path. Usually this is the result of a mistaken share, addition, or a change in folder check-out path.
As it is undetermined what order StarTeam will check files out, having two different files with the same name pointing to the same local location is a mistake. The 2005R2 native Windows client and Cross Platform client would pull files in a slightly different order, causing views with this issue to produce checkouts with different files.
There are a few ways to legitimately end up with this situation - keeping multiple branches of a single file on the tip and designating which to use by view label is a common one. StarTeam is one of the few VCs that can handle multiple references like this.
Separately, if you can help it, never delete and re-add a file. 99% of the time, it isn't what you want to do. Re-adding a file to the server gives it a whole new archive in the database and it will not be tied in any way to existing shares of that file. You will also lose all of your revision history. Instead, look for ways to move files. And if you have multiple files with the same name and content on your server, it is usually a good idea to make an effort to share the file to the various places it should exist rather than re-add it. This will make sure your changes propagate so you don't need to check-in fixes multiple times.
Were those files added twice? You can add the same file multiple times which could result in this error. Also, check the Reference tab - are these linked files to elsewhere in the repository?
Could that be the files or their status got corrupted on the server? In that case, delete and re-add is the only way.