I would like to label a project and its files in VSS according to a certain previous date. In other words, I would like to be able to select versions of each file in a project as they existed on a particular date and then apply the label to those versions of the files. I have tried pinning files and then labelling the project (and its files), but that failed. This is a very simple task in ClearCase. Is there no simple way to do this in VSS?
The pinning and labeling should work, however due to a display bug where things are displayed in order of date instead of the actual revision number order the label will look like it is associated to the newest revision of the file at the time the label was placed.
You can verify the actual revision a label is associated with a few ways, from the GUI you can select the label and do a diff with the newest version, since it is associated with a different version you should see that the label and the revision right below it do not match and indeed differences exist.
An alternative way to check it is to "pin" at the label. Select the label and pin the file, a pin will show up on the proper revision actually associated with the label.
Finally you can use the command line and do a ss dir -vl"" which will display the actual version of the label, combine this with ss dir -vd"" and you should see that your label and the date revisions are in sync.
Best of luck.
Related
My first question in Stackoverflow, I'm thinking in making a litle program to edit some folder icons on Windows, I searched for the meaning of the number after the path of the icon but find no answer, the system icons have diferent values, i.e:
Music Folder
IconResource=%SystemRoot%\system32\imageres.dll,-108
Pictures Folder
IconResource=%SystemRoot%\system32\imageres.dll,-113
When we manually change an icon from a folder,usually is used(to keep it in a removable media):
IconResource=..\Icons\icon.ico,0
There is any influence in that last number? Or it doesn't matter for that purpose?
.EXE and .DLL files can contain more than one icon. ,0 is the first icon in the file, this syntax can also be used for .ICO files that only contain one icon.
Positive numbers simply refer to the order the icons are stored in the executable file (,0 is the first, ,1 is the second and ,2 is the third etc.). This order is the same as the order of icons in the standard pick icon dialog in Windows (shortcut properties etc.).
A negative number is the resource id of the icon, this is a number chosen by the author of said executable and can be stable over time even if the number if icons changes if the author chooses to use stable resource ids.
See also:
How the shell converts an icon location into an icon
For some reason, source tree does not show the listing of files in my commit. I am on the "Workspace->History" view, It's set to "All Branches", "Show Remote Branches" and "Ancestral Order".
The tree shows up fine.
Under that, it shows me commit description, SHA1, parents, author, date, labels. On the right, it shows me only the diff of only one of files in the commit.
If I drag up the lower border, I can see the Command History.
If I create and stage new files, I do see them appear in the lower portion of the view in "Staged files" and "Unstaged files", and I can click on each file, and the diff on the right shows the differences. But once I commit, I can no longer see the file listing.
I know that all the files are committed as I see them in ls-files, it seems like I'm missing an option or preference somewhere in the tool. Anyone know what I'm referring to?
Wow. I did not realize there was a table with "Filename / Path" headers. On a mac at least, you can resize the contents of the table to a 0 height, so that it completely disappears and there is no way to bring it back without hovering over with a mouse.
I am trying to compare two large Visual Studio 2010 solutions using Beyond Compare. These solutions have 60 projects, the only way I can find to ignore differences in the bin and obj folders of each project is by right-clicking these in the compare results window. I don't want to have to do this 60 times, so is there any other way I can tell Beyond Compare to ignore these folders in each solution?
This is for version 3,
While you are comparing two folders you can go the menu Session->Session settings and go to the tab named Name Filters. There are several boxes to include or exclude files or folders; locate the one on the bottom right and on each line you can declare the folders that you want to exclude. There is an important combo box on the lower left corner where you can specify if you are going to use this settings for just this time or to use them everytime.
Another way to do this is simple look for the File Filters Toolbar (the one with a combo box and a pair of icons with a pair of glasses), in this combo box you can set your filter, for example -bin\;-obj\ and this will exclude the bin and obj folders.
I know this is old, but I came here looking for the same info for version 4.
Version 4 supports the same functionality but expose on the top toolbar in the session, just add the same -bin\;-obj\ in the "Filters:" text box:
I see the option for specifying three files (left, right, and ancestor), but it doesn't seem to be possible to actually display the ancestor.
The bottom pane shows the result of the merge. I'd rather see what the original content was, so I could understand the context for the left and right sides of the conflict. Is this possible? FileMerge has awful help documentation.
Looks like in version 2.5 they have finally added this feature. "Ancestor" now actually means "Ancestor" and there is a separate "merge" option
From the command line you can do something like this:
$ opendiff mine theirs --ancestor base
Where "mine" is my version of the file, "theirs" is the version I'm trying to merge with, and "base" is the common ancestor of "mine" and "theirs.
This command will open File Merge and show the ancestor at the bottom (where the merge is in my example from 2010).
Unfortunately, it's not possible to use the --merge option along with the --ancestor option. So you can't view a three-way merge AND use File Merge to resolve the conflicts at the same time.
I am starting to use File Merge for viewing and understanding the conflict, then editing the conflict markers in a text editor as usual. (sigh...)
To my knowledge, this is not possible with filemerge.
Paid (but very good) app: http://www.deltopia.com/
Open source: http://sourcegear.com/diffmerge/
#mehaase:
That is not exactly correct. Here is the documentation on the new ancestor option from FileMerge Help:
Sometimes, you need to compare two versions of a file that have been modified
independently. For example, say two people branch a file; that is, they make
copies of a file and modify it. In this case, specify an ancestor file, which
is the common ancestor of the modified files.
The ancestor file lets FileMerge choose when to take changes from the left
file or the right file. For example, say the left file and the right file
contain a different version of line 33. If line 33 in the ancestor matches
line 33 in the left file, you know the change happened after the two files
branched off, and FileMerge prefers the newer version of line 33 in the right
file. If line 33 is different among the three files, you know that the files’
editors edited the lines separately. FileMerge displays that difference with
a red border and asks you to choose with edit to use.
So the ancestor is still not displayed, but it helps the merge tool to make a little more informed decisions when it decides to choose the left or right sides by default.
I realise this answer technically doesn't answer the question on how to do it with FileMerge, but I use kdiff3 (available for Windows, Mac and *nix) and it shows the base/ancestor, plus left and right, plus output/result, for a 3-way merge. (http://kdiff3.sourceforge.net/)
It's not a pretty user interface, but (IMHO) it's fairly simple to use and works well. And it's open source. :-)
Often, when I am reading code or debugging, I want the ability to quickly jump around files. I especially want to "go back" to where I was. I know about "Command+T", "Command+Shift+T", and, bookmarks. But, I cannot figure out a way to jump around files quickly.
UPDATE: I do not think I my question was clear enough judging by two answers given. Specifically, I am looking for a way to "jump back" to where I was in a file. I know how to navigate in TextMate (in general). I want to know if TextMate has a "jump back" key binding.
It's subtle.
The command-T thing has the files listed in Most Recently Used order.
So, you can go command-T return to get back to your last file real quick. At first I couldn't find it either.
I don't think there's a go to last edit location as there is in, say, IDEA/RubyMine.
Courtesy of MacroMates.com
2.3 Moving Between Files (With Grace)
When working with projects there are a few ways to move between the open files.
The most straightforward way is by clicking on the file tab you need. This can also be done from the keyboard by pressing ⌘1-9, which will switch to file tab 1-9.
You can also use ⌥⌘← and ⌥⌘→ to select the file tab to the left or right of the current one.
It is possible to re-arrange the file tabs by using the mouse to drag-sort them (click and hold the mouse button on a tab and then drag it to the new location). This should make it possible to arrange them so that keyboard switching is more natural.
One more key is ⌥⌘↑ which cycles through text files with the same base name as the current file. This is mainly useful when working with languages which have an interface file (header) and implementation file (source).
When you want to move to a file which is not open you can use the Go to File… action in the Navigation menu (bound to ⌘T). This opens a window like the one shown below.
Go To File
This window lists all text files in the project sorted by last use, which means pressing return will open (or go to) the last file you worked on. So using it this way makes for easy switching to the most recently used file.
You can enter a filter string to narrow down the number of files shown. This filter string is matched against the filenames as an abbreviation and the files are sorted according to how well they match the given abbreviation. For example in the picture above the filter string is otv and TextMate determines that OakTextView.h is the best match for that (by placing it at the top).
The file I want is OakTextView.mm which ranks as #2. But since I have already corrected it in the past, TextMate has learned that this is the match that should go together with the otv filter string, i.e. it is adaptive and learns from your usage patterns.
If you have a project window open, you can leave frequently-accessed files open (in tabs), and then use ⌘+1-9 to jump to open tabs.