I'm curious to know what DVCS strategy people use in enterprises.
The Github model is based on forks because in open source projects you have some trust issues and your probably don't want anyone to be able to update your code. However, in an enterprise context the trust issue is not as critical and branches might be a more suitable option for DVCS in terms of maintenance (less repos, less permissions to manage).
So, the question is simple: do you replicate Github model and use forks for new features or release with some sort of pull requests mechanism or would you rather work with branches.
This post might help people to get some insights about the strategy they should use if they think about going to DVCS in their own company.
Note that GitHub support both models:
for sensitive repos, you can fork them (clone at the GitHub server side), in order for other collaborators to own the fork, and to push freely to it.
for collaborative repos, with many collaborators, you can create an organisation, in which you manage all your collaborators.
For non-GitHub repos, an extra authorization layer like Gitolite support personal branch precisely for isolating contributor's contributions in their own branch, avoiding the need to provide n clones (forks) for a given repo.
I think you will get some pretty strong opinions on this question! Where I work, we follow a model where
I and one other person have full access to a machine containing the company's official repositories, and
we have several developers who have read access. They will make changes, pull/merge from eachother/us, and occasionally
we will pull revisions we approve and push them to our company's official repositories.
See http://progit.org/book/ch5-1.html for more extended reading on this topic.
Related
I am working on Power Designer 16.1, the repository is installed on a Linux server.
I have an issue with the repository sometimes dropping my models, in most cases I am able to check-in the models to the repository and check-out correctly, but sometimes after I correctly check-in a model it seems that it is being deleted from the repository.
I am not sure but I think the problem may occur when two (or more) people tries to check-in the same model at the same time (or overlapping), and that causes the model to disappear from the repository.
Any idea or workaround to fix this?
Does reading logs of PowerDesigner can help to clarify the reason?
Maybe repository has it's own server logs?
First make sure you upgrade to the latest version. They fix tons of bugs in each patch, and a lot of them have to do with the repository. It is very likely that this will fix it.
But we still found it best to have ONE designated person do check-in on models, to avoid strange issues.
If you use any replication between models, the repository will act even stranger than it already does, so if you can, try removing replicating properties from your models.
Apart from that, we make daily database backups of the repository and we sometimes have to use them too.
Another thing that may help: in the PowerDesigner Repository Administration menu you can select to rebuild indexes, if you have the logon/password for the account that installed the repository. Rebuilding indexes has solved several nasty prob
And now a bit off-topic: in my experience (working with the repository for 3+ years now), it has always been very buggy. Even with the highest patch level installed we have lots and lots of weird issues. This only gets worse when you have 2 repositories in order to test a new release of PowerDesigner. PowerDesigner's habit of hardcoding ID's into models make this very problematic.
I know someone who is often hired by SAP as PowerDesigner Instructor. His recommendation is to stay away from the repository and use Git. It has always been a badly integrated add-on with questionable support. The datamodel for the repository is horrible (which is quite ironic) and it seems that foreign keys and relational integrity are foreign concepts to it.
So we are now in the process of moving off the repository and onto TFS. All PowerDesigner models are XML, so TFS or Subversion or Git etc. will do nicely (but we like the locking of TFS/Subversion for this). Yes, we will have to forget about the nice display of changes, but frankly, having a robust version management instead of the buggy and model-destroying repository is worth it.
Intro:
I'm working for a contractor company. We're making SW for different corporate clients, each with their own rules, SW standards etc.
Problem:
The result is, that we are using several bug-tracking systems. The amount of tickets flow is relatively big and the SLA are deadly sometimes. The main problem is, that we are keeping track of these tickets in our own BT (currently Mantis) but we're also communicating with clients in theirs BT. But as it is, two many channels of communication are making too much information noise.
Solution, progress:
Actual solution is an employee having responsibility for synchronizing the streams and keeping track of the SLA and many other things. It's consuming quite a large part of his time (cca 70%) that can be spend on something more valuable. The other thing is, that he is not fast enough and sometimes the sync is not really synced. Some parts of the comments are left only on one system, some are lost completely. (And don't start me at holidays or sickness, that's where the fun begins)
Question:
How to automate this process: aggregating tasks, watching SLA, notifying the right people etc. partially or all together?
Thank you, for your answers.
You need something like Zapier. It can map different applications and synchronize data between them. It works simply:
You create zap (for example between redmine and teamwork).
You configure mapping (how items/attributes in redmine maps to items/attributes in teamwork)
You generate access tokens in both systems and write them to zap.
Zapier makes regular synchronization between redmine and teamwork.
But mantis is not yet supported by Zapier. If all/most of your clients BT are in Zapier's apps list, you may move your own BT to another platform or make a request to Zapier for mantis support.
Another way is develop your own synchronization service that will connect to all client's BTs as each employee using login/password/token and download updates to your own BT. It is hard way and this solution requires continious development to support actual virsions of client's BTs.
You can have a look a Slack : https://slack.com/
It's a great tools for group conversations
Talk, share, and make decisions in open channels across your team, in
private groups for sensitive matters, or use direct messages
one-to-one.
you can have a lot of integrations tools, and you can use Zapier https://zapier.com/ with it to programm triggers.
With differents channels you can notifying the right people partially or all together in group conversation :)
The obvious answer is to create integrations between all of the various BT's. Without knowing what those are, it's hard to say if that's entirely possible. Most modern BTs have an API and support integrations. Some, especially more desktop based ones, don't. For those you probably have to monitor a database directly.
Zapier, as someone already suggested, is a great tool for creating integrations and may already have some of the ones yo need available. I love Slack and it has an API, but messages are basically just text and unless you want to do some kind of delimiting when you post messages to its API, it probably isn't going to work.
I'm not sure what budget is, but it will cost resources to create the integrations. I'd suggest that you hire someone to simply manage these. Someone who's sole responsibility is to cross-populate the internal and the external bug tracking system and track the progress in each. All you really need is someone with good attention to detail for this, they don't have to be a developer. This should be more cost effective than using developer resources on this.
The other alternative is simply to stop. If your requirements dictate that you use your clients' bug tracking software for projects you do for them, just use their software and stop duplicating the effort. If you need some kind of central repository or something for managing work maybe just a simple table somewhere or spreadsheet with the client, the project, the issue number, the status and if possible a link to the issue in the client's BT. I understand the need and desire for centralizing this, but if it's stifling productivity, then the opportunity costs are too high IMO.
If you create an integration tool foe this, you will indeed have a very viable product. This is actually a pretty common problem.
I'm currently experimenting with Bazaar, trying to find a configuration that works well for the project I am currently working on with a team of developers. Our model is currently centralised but I am open to persuasion about changing this if I can see significant benefits. Despite advocating it, I will not be able to change the version control system... some people don't know what's good for them.
At present, the project is divided into tiers. It's a web project and consists of a data access layer, a business/domain logic layer and a web layer (and a bunch of other application level projects such as services that that sit on the domain).
At present I find a bazaar repository with a number of working trees in there, one for each of the tiers I have mentioned.
Question part 1
Is there a better alternative to using working trees inside a respository for each tier?
What I have considered:
Putting everything into one fat branch/working tree (I am told this was purposely avoided, because of the necessity to check out everything). Developers are creating their own local setups by checking out subsets of the multiple repositories. For example if I am working on one of the services, I checkout the service, the business layer and the data access layer to a local directory. In the current setup though, I can checkout a top tier application which doesn't affect anything else, make a change to it and commit it back, without checking out the entire repository, which is nearly 1GB in size.
What I would like to remedy:
The problem is really that my web tier is reliant on a version of the
business layer, which in turn is reliant on the data access layer.
With a project organised like this, I have no idea which version of
the business and data access layers were current for a given commit
on the web layer. This feels very bad to me. I want to be able to
checkout versions of our application relative to some commit to one
of the working trees. At the moment we are attempting to keep track of
this across all of the working trees with tagging, but it feels
pretty wrong to me.
Question part 2
If I branch one of these tiers for a release branch, and find that a change in the root of the branch needs to be in that release, how do I push only those required changes into the branch?
So assuming that the structure of one working tree (or possibly a real branch) exists for each of these projects, we would like to have a branch for each of these, which contain a version for the a particular release. So the website tree has a website_rls1 branch, which encapsulates the state of development for that particular release. General development for a later release is going to happen concurrently. So if one file out of many happens to be relevant to this release, I would like to merge that single file into my release branch. What is the preferred method of doing this. As I understand it, Bazaar does not support merging/pulling individual files.
What I have considered:
Just putting the files I want into my local checkout of the release branch and committing
What I would like to remedy:
Following this consideration is going to kill off version information for this file. I want to keep the version information in tact, but only merge in a subset of changes. I essentially want to push individual files up to the child branch, or something to the effect of.
Any thoughts appreciated.
#1 There are two plugins for managing multiple dependent bzr repositories:
https://launchpad.net/bzr-externals
https://launchpad.net/bzr-scmproj
They have different approaches and are suited to different situations.
You may also want to consider solving this problem at the build level. Use a continuous integration system like Jenkins and a dependency resolution system like Ivy or Maven. QA should be done using automated builds from this system so that bugs that are filed can refer to a particular build number which should include in its logs the versions of the various dependencies used to produce that build.
#2 There isn't really a way to do that merge. Bazaar doesn't track cherry-pick merges (yet). A better approach would be to make the original fix in the release branch and then merge it back into its parent.
I work in a shop that's adopted TFS for source control. We're happy with the integration in VS and the other features it offers, so it's likely we're not going to switch to another platform. However, features for team collaboration and documentation are lacking compared to other solutions, most notably SVN + Trac.
Has someone been able to integrate TFS 2010 with TRAC?
EDIT:
It's been asked that I clarify my intent here. It's very simple. I'm just asking if the TRAC project management and bug/tracking system can be used in conjunction with TFS. And, if so, what would it take?
Remember, I'm not looking for a Sharepoint solution--I've already got that. I'm asking if it's possible that TRAC and TFS can get along.
(Long rambling clarification on what I'm looking to get out of TRAC removed. The question is simply "Can TRAC work with TFS?")
Not so much of a trouble a the Trac side. There is FOSS everywhere, a lot of modularity and flexibility. No quite the same at the other side. I've read about the trouble even with one-time migration from SVN to TFS. Despite the source is all open and well documented, there no evidence of good support, that should tell you much about the chance for getting even more - constant synchronization.
Facts: MS SQL server is the base for TFS. No connector available for MS SQL server as a Trac db backend, although there are several python bindings to MS SQL server available, or the option to connect via ODBC. But just an option, nothing ready AFAIK.
I'm not aware of any well documented open TFS API as foundation for migration and integration. And I'm not convinced this will ever change. At Redmond (Microsoft) they are reportedly only considering what seems important to themselves: "helping customers with IBM Rational ClearCase and ClearQuest tools." And most probably it this behavior will persist and SVN/Trac keeps very low on the ToDo for them.
[Edit2]
While TFS has some support for bidirectional communication, these scenarios are not recommended. It mostly aims at integration, read: sucking information in, not communicating with other information systems like Trac.
[Edit]
Just for sub-task of repository browsing you could try to write code to push a duplicate of changes to another (SQLite|MySQL|PostgreSQL) repo that Trac supports right now. But I consider this is rather wasteful and ugly, and fact remains, that it's hard, if possible at all, to do the same tracking without such big code duplication. Ultimately, if you want to live without the actual check-in source changes you must at the very least send information about the meta-data like resource ID's (for link generation) to find the data in TFS.
I'm looking into that right now. So take the following as half-educated advise to the best of my knowledge and feel free to correct/discuss.
I have a Mailman managed list with years of history that I want to migrate into a web-based forum. Things it would be nice to do:
Keep the mailing list going for those who are used to and prefer it to a web interface, but have it integrate with the web-forum activity.
Take the web-based forum posts, and send them out to the mailing list.
I have run sites based on phpBB, Drupal, Simple Machines, etc., and am able to do a little bit of coding if that was what was required to integrate some package into Mailman, or replace it entirely. But I'm unaware of what is available, commercial, or open-source, that could accomplish this. I am also open to replacing Mailman, if the candidate package can perform its functions reasonably well and integrate well into its own forum.
Even if the phpBB/mail2forum option is using the older version, it is a nice option. OpenSceneGraph just implemented this for their long running mailing list, and it seems to work flawlessly. It's the only option I've seen that allows for forum posts and mailing list messages to work very seamlessly, and has the means to allow for tags to separate forum subjects from mailing list posts, etc.
Their v2 dev works with phpBB3, too - so if you're willing to live on the bleeding edge, that might be an option.
So far, here's what I've looked into:
Drupal, with its Mailmanager and Listmanager modules, though I've been having trouble turning on the IMAP feature so it can talk to my mail box (clients can access it fine), and the Drupal's forum module isn't really up to the feature level I'd like.
PhpBB, mail2forum, though it looks like currently it only works with the older phpBB version, so not really an alternative for me.
The only one that seems to really "work" is FUDforum and its maillist.php module, which integrates directly at the procmail level, nice. I can take all my Mailman archives and "formmail -ds" them directly into a chosen forum.
I've started a bounty for this issue, I'm looking for something, even commercial, that really integrates the email interface into the forum experience for the end users, particularly the handling of accounts. With FUDforum I'm still going to have to resolve the separation of the Mailman accounts from the FUDforum accounts.
I ran across this on the web: http://mail2forum.com/
I haven't used it but it looks promising and has both a 1.2 stable version and a 2.0 development version, so it's not a derelict SourceForge project with 1 developer and no commits, or anything :)
Theoretically my org may use it at some point in the next year or two, but we have to finish our own Listserv to Mailman transition first.
Well, based on what you've said (first, that you're able to do a bit of coding and second that FUDForum would work for you except for the issue of maintaining list membership, have you considered that:
Mailman stores its user information in a plain text file and
FUDForums stores its user information in a single sql table
The obvious solution would be to declare one of these master (probably FUDForum, since it looks as if it holds a proper superset of the Mailman info) and have a little script/cron job that copied changes from the master to the slave.
A passing note -- neither of these systems appears very secure (actually, they both look leaky as all get out) and combining them may well reduce the collective security even further. If you are doing anything even remotely confidential you should rethink your goals, and in any case you should take appropriate steps to protect your system from attack.
I know an SME which switched from the mailing list to phpBB!
If you want to keep your mailing list, you have also to maintrain it.
Finally the people I know diabled the mailing list service.