Identifying files for a hot fix/patch - diff

We (occasionally!) have to issue hot fixes for our product and do this by reissuing the affected files directly rather than with a new installer. The product has a large number of pieces, some managed code, some unmanaged.
Currently development flags which build artifacts (exes, dlls) need to be shipped in a hot fix. We'd like to be able to identify these automatically by comparing them to the previous build. A simple binary diff doesn't work since the version numbers on all the files have changed as stamping the files with a new number if part of the build.
Are there any tools that will do a more intelligent comparison and decide which files should be included? We'd still have a developer check the list, this is more to catch files the developer didn't think of than the other way around.
(Note: changing the hot fix/build process is not an immediate option, whether or not we should be shipping individual files is a different discussion!)

These are the options I see:
On your build machine get a report of the files that were changed and use the directory structure of the file path to determine which dlls were really updated. Not sure if this breaks your "no build process changes" rule or not.
If you want to wait until after the build I would recommend using a binary file diff tool like http://www.romeotango.com/Downloads/FileCompReadMe.txt. Using that you can get back a set of diffs so you just need to get your script that uses the tool to ignore the diff that occurs as a result of the version number. You can figure out the pattern to how the version number appears by using a controlled scenario where you know the two binary files are the same except for the version number and note where the differences are. Do that for a few of your dlls and hopefully a pattern emerges enough so that you can script it.

Related

Merge automatically two hex files into one within Eclipse

I have a bootloader project and an app project within the same workspace in Eclipse.
I'd like to merge the hex files of theses two project into one single hex, so that I can flash in my MCU both project at the same time.
I know this is possible using some tools (https://www.keil.com/support/docs/2666.htm), but I would like that the merge process happens in Eclipse using a post-build command or else.
Thanks
Intel hex are just text files, you can merge them manually in any text editor by using simple copy/paste. Grab the data rows only from one of the files.
This assuming that there's no CRC on the program as whole. Also, you probably shouldn't have multiple rows with the same address in the same file or you might confuse some tools.
I'd recommend doing this through a diff tool such as for example WinMerge, or the one integrated in your version control system. Once you have it working you could write a file handling script for merging them automatically.

Simple yet fast deployment tool or solution for FTP-based server

I have a simple task to which some simple solution should exist yet I cannot come across one.
I have a huge file tree on computer A (development). I have the same (multiple) such file trees on a computer B (let's call it production). Computer B runs FTP and PHP, nothing much else.
I need to move the changed files from the tree on A to the tree on B but as efficiently as possible. I.e. if just one file changes, it will just transfer that one file. It would be enough to "compare" the local and remote trees using last modification dates, nothing else needed.
I tried to use the good old Ant for it but that really does not work as the FTP task is really bad one there (does not preserve modification dates on PUT and so on). What other options are there if I do not want to write the code for such a task myself? I'd expect there is some tool out there that would make a remote dir listing, download it to local computer, select only those changed files and transfer them to the destination. Do you know how I could do it? Some sort of FTP or PHP-based distributed robocopy?
EDIT: I should have added that I mean doing it on a Windows 10 computer syncing to some FTP/PHP server using command-line automated script, not GUI.
Actually I solved the issue using winscp. I managed to integrate it into ant calling it through the task and using the winscp's synchronize command. For my current folder size it is fast enough, let's see later. The FTP command in ant was not useful since it does not preserve the modification dates.

Simple and easy to use tool for managing different versions of files

I want to manage different sets of file versions locally on a machine without using complex version control tools like TFS/Git/SVN...etc. here is my use case:
I have a Windows virtual machine that contains many xml, xslt, xsl, txt...etc. files, the virtual machine gets updated with every release of my product.
Often I need to analyze errors in this virtual machine, so I change many files and run the product and start analyzing, let us call these file changes FileChangeSet1.
based on the results above I need to change other files and maybe some of the files in FileChangeSet1 and do another test.
again based on the results, I need to change more files, eventually I end up with FileChangeSet1, FileChangeSet2...FileChangeSet(n)
I want to:
be able to switch between these file change sets easily and quickly, e.g. have a GUI that shows my my tree of FileChangeSets then click one of them and all files of that change are used.
create file change sets from other file change sets e.g. copy FileChangeSet1 in FileChangeSet2 and change only one file in set 2
I don't want to configure and install a complex version/source control system like TFS/Git/SVN where I have to create a database of all my files first.
Making snapshots of the virtual machine is not an option because it is extremely slow.
I think you would not have much advantage with version control tools even because they are made to version text files. For binary files, I think you would end up like managing several diffent copies of the binary files anyway (at least for older tools such as CVS and SVN).
If you are running in linux, you may want to use cmp/diff tools. Take a look on incremental diff and diff tools such as patchutils.
Consider also to create a checksum of huge files to avoid comparing them for nothing.
ps. also take a look on this - http://jojodiff.sourceforge.net/ - haven't tried but it seems simple to use and promising.
Mercurial is the right tool for me. With it I can solve my business case easily as follows:
Install mercurial on Windows, it integrates in the Windows file explorer.
Create a local version control mercurial database by right clicking my root folder.
Now I can open all my files under my root folder in different text editors e.g. notepad++ and modify these files.
When I want to save/remember a specific status I simply commit the files to mercurial by right clicking the root folder, I can provide a commit note.
Later I can change my files in a different way and test how my system reacts to them, again I can commit these files locally.
Over time I have a history of change sets in Mercurial, I can go back to any change set, branch it, merge it...etc.
I have a huge and complex system that contains thousands of files, my root folder is actually the C:\ drive, I can easily and quickly make out of c: a version control database using mercurial.
All with a simple and intuitive GUI, no command line learning needed.

Hybrid version control & sync system?

Is anyone aware of a hybrid version control and synchronising system?
I'm currently a happy mercurial user, but my projects usually contain a mixture of files.
Most of these (code, documentation, ...) I want to be version-controlled. This is why I use mercurial.
However, on the rare occasion I have files that I would like to synchronise between my working copies, but not version control.
For example, I version control the code I write to do image processing. This code can produce a whole bunch of output images which I'd like to have synchronised so I don't have to remember to shuffle them around my various computers, but there's no point having these version controlled.
To clarify - I am aware of extension to mercurial such as bfiles and bigfiles, which are handy for my image example, but I was just wondering if anyone out there knows of alternative ways to handle this. I just want the one system that I can tell "version control all files except those ones, which should be synced but have no history".
cheers!
EDIT: I could do something like adding a hg marksync <filename> that added <filename> to a list of files to be synced, and then adding a hook to hg push/hg pull that would (say) run rsync (or whichever sync tool) in the background, but I wondered if there was a less hacky solution (I think bfiles/bigfiles do something along these lines anyway).
Version Control System (any) doesn't care about synchronization of
not versioned data
besides default pathes
If you want sync any files - use specially designed for this task tools: f.e. rsync
This code can produce a whole bunch of output images which I'd like to have synchronised
Is this DATA or part of your CODE?
If data: Keep out of your versioning system, just don't go there. If it is part of your code (like layout images) check it in. Those are the only ways which are the generally accepted.
A nice solution for the data would be syncing OR generating them. So you might add a step after deployment to a server: GenerateImages().
edit: In addition to the comment made by the thread starter:
If the images are data and you need to process them on a different system don't think about the version control for your code. It is unrelated. The steps which would make sense to me, in order of processing:
Start with updating your image code, check it in versioning. Then deploy (yes this is deployment) the updated code to the cruncher computer. Now code is done.
Then you have tasks which the number cruncher should handle. Like processing the images. So start that processing from either the cruncher itself (probably some queue happens there) or from a central dispatcher.
Then you have the results locally at the cruncher. Now something has to happen with that data, so that's also part of your software. Decide whether you want the cruncher to send them to some central storage, your workstation or another location. Let the software handle that. This is the most hard part as I read through your question. Many solutions are possible from just FTP/network transfers to specific storage solutions. Willing to help but need more info about the real issues, amounts, sizes etc. on these parts.
If the new updated version of the image processor makes the old generated images obsolete implement that also in your code, by for example attaching an attribute to the files generated, a seperate folder or another indication. That way you could request the cruncher after update to re-generate any obsolete files.

TFS 2008, remove file from source control but leave it in the project

We are using Scott Hansleman's suggestion for multiple web.configs from his post here. The problem we have is that we have to check out the Web.Config. If we remove it from the project, when we publish, no web.config is pushed. So we need to remove the source control bindings just from the web.config, but leave it in the project, and have the rest of the project still held under source control.
The issue is that source control makes the file read only until you check it out. We need to be able ot overwrite it with the prebuild events, preferably without having to check it out. Is there a way to remove the bindings from that file only, and still leave it as part of the project?
Thanks.
By adding a new file to solution explorer, you will get the little plus sign indicating it is due to be added to source control. Then, right-click and choose "undo pending changes". This will cancel the add but leave the file in your project.
If that doesn't work I suggest one of the following methods:
Use the Attrib task from the MSBuild
Community Tasks project to remove
the read only flag.
Use the Exec
task in MSBuild to invoke
tf.exe and checkout the file.
You should leave the file in source control. Otherwise you'll run into several issues:
changes won't be versioned. 'nuf said.
it can't be branched or merged, even though web.config is one of the files that's most likely to vary between parallel dev/test/production environments
changes you make locally won't propagate to coworkers without manual workarounds
developers setting up an environment for the first time won't get the file at all
Team Builds won't contain the file, so neither will your deployments. (surely you're not deploying directly from the desktop?!)
Note that the state of individual files is stored entirely on the TFS server. ('tf properties' dumps this metadata if you're curious) Only projects & solutions have bindings actually written into the file. And even those are dummy entries that tell VS "don't worry about me, just ask TFSProvider, it'll know who I am and where I'm supposed to be." While there are many other quirks in the VS project system that give me endless headaches, in this case it's your friend. Don't circumvent it.
Best options:
Edit your build script to toggle the read-only attribute before/after modification. If you're using the "copyifnewer.bat" script from the linked blog post, it should literally be one extra line. Even if you want to keep things entirely declarative within the MSBuild makefile, it's barely any work with the help of 3rd party tasks.
Use the File -> Source Control -> Exclude feature. After applying this setting, the file remains under source control, but will no longer be subject to automatic checkouts/checkins by the active solution. In other words, you can edit the file locally to your heart's content without affecting anyone else, but if you want to commit (or shelve) your changes you'll need to do it from Source Control Explorer or the command line.
Option #1 has the advantage of being a very quick fix for your existing setup. The downside comes from maintaining several copies of web.config.* Same reason why copy/pasting code is bad: if you change one, you have to go change all the others -- or worse, forget and let them drift out of sync until strange bugs force you to revisit the issue. This could be improved by changing the process so that there's only 1 "master" web.config and the additional copies only contain differences (via a textual diff engine, XSLT transforms, programmatic manipulation in Powershell, etc). Of course, that's more work.
Option #2 avoids #1's problems with very little overhead. (the engineering process itself is unchanged; only difference is how the Visual Studio UI behaves) This advantage is critical if you make changes to web.config at all frequently. Downside is that there is no built-in way to track variations on the "master" file. If the only diffs are dirt simple, eg a connection string or two, you may find it easiest to stick with just one "master" and let people make ad hoc changes on their dev machines. There are even tools to do this for you, such as Web Deployment Projects (easy) and the IIS Deployment Tool (complex). In any case your actual deployment should be automated and source-controlled, of course! If heavier customizations are required than these tools are capable of, then you'll probably want the hybrid master + transform approach described earlier.
I recently ran into the issue and could not find a good solution. With a little trial and error I was able to figure this out myself.
This works on Visual Studio 2015. I tried to follow the answer above, but 2015 does not have an "Exclude From Source Control Option" I could find. It does have source control / project integration where if you delete the file from source control or the project, it will automatically be removed from both places. This integration is enforced when you have the solution containing the project open.
The problem is when using a web.template.config, web.config is really a build output and should not be in source control. However deleting the file entirely and removing it from the project causes problems because then the file is not part of the Build / Publish steps.
So the work around turns out to be simple:
Close the solution. File --> Close Solution.
In the Source Control Explorer Window, Delete the file.
Check in the Change.
Open your solution and you will see that the Config file remains in the project.
You may see the Triangle / Exclamation Mark Icon warning showing the the config file is not found.
Rebuild the project and click on the file.
The file should now show without the warning.
It should not have the padlock icon next to it. This indicates that it is not under version control.
I did some more checking and found the "Exclude From source Control" feature. The key is that you need to select the item in the Solution Explorer window before selecting the menu item:
Select the Web.config file in the "Solution Explorer".
In the menus select File -> source control -> advanced -> Exclude Web.config from Source Control.
You will now get a Red Circle / White Line icon.
This works if the file is NOT ALREADY in source control.
If the file IS ALREADY in source control you need to do the procedure above to remove it from source control without also deleting it from the project.