Slashdot Mirror

Summarizing about half the posts in this discussion, SuperDuG says:

Personally I don't see a need to switch to 2.0 yet. My site runs just fine on 1.x series.

Another large group (myself included) seems to be waiting until PHP et al are widely considered stable on the new platform, which in the case of PHP will occur sometime after the Zend site has a big fat endorsement on its front page.

The only reason I am looking at Apache 2.0 before that endorsement goes up is the Subversion project. This is a free software replacement for CVS which is getting relatively close to release and which will provide nice little features that CVS doesn't have like versioned moves and renames for files and directories. That one feature alone will make me very happy. The Subversion server requires Apache 2.0 for remote access, therefore 2.0 becomes more immediately interesting to me.

Wouldn't it be a little ironic if version control was what finally drove migration from the 1.3 series to 2.0?

• ArgoUML:http://argouml.tigris.org/
• XDoclet:http://xdoclet.sourceforge.net/
• uml2ejb:http://uml2ejb.sourceforge.net/
• struts:http://www-106.ibm.com/developerworks/jav a/library/j-struts/
• struts:http://jakarta.apache.org/struts/index.ht ml
• ant:http://jakarta.apache.org/ant/index.html
• maven:http://jakarta.apache.org/turbine/maven/
• log4j:http://jakarta.apache.org/log4j/docs/
• middlegen:http://sourceforge.net/projects/middle genmaven
• maven:http://jakarta.apache.org/turbine/maven/
• xalan:http://xml.apache.org/xalan-j/index.html
• xerces:http://xml.apache.org/xerces2-j/index.htm l

The article mentions Viso, and rational rose both of which are expensive windows programs. So does that mean cash strapped geeks are forced to use a windows box, download a crack and use that? No. There are good alternatives that are free and don't need windows. The first one that comes to mind is Gnome's DIA.

To quote Gnomes web site "DIA is a drawing program, designed to be much like the commercial Windows program 'Visio'. It can be used to draw various different kinds of diagrams. In this first version there is support for UML static structure diagrams (class diagrams), databases, circuit objects, flowcharts, network diagrams and more."

Another option is Argo UML which is also free and programmed in java Argo

My traits? Hmmm. I'm not a quick thinker. Hell, quite the contrary. I'm a very slow and belaboured thinker. I have a hard time keeping up in conversations because they change quickly. But I'm a very thorough thinker. I think deeper about any given topic than anyone I know. Perhaps this is why I don't follow conversations well... the MTV style of changing topics every 60 seconds don't suit me well. I'd rather stay on one topic for an hour.

You want to ask me what my favourite joke is? I don't have one. But everyone I've *EVER* worked with says I have a great sense of humour. Very... DEEP. I don't tell jokes. Ever. But I often have something to say that makes people laugh or makes them amused.

Well, at least on that "passion" front I can brag. I have no television, and I don't play games. But I do have six computers at home. They run PostgreSQL, Oracle, Sybase, Perl CGI scripts, shell scripts, and who knows what that I've written. I've also written a Dia diagram to SQL converter (tedia2sql).

I'd recommend rethinking your interview technique.

I'm not sure why, but I find it amusing to picture the author trying to find bugs in Valgrind, actually using Valgrind.

Perhaps it shouldn't seem as funny to me in days when Subversion is self-hosting, and GCC compiles itself.

-Sou|cuttr

1. Nobody should ever be using SCCS these days
2. I wouldn't join in with an attempt to clone BitKeeper because
   • Larry is very good - I wouldn't bet that it can be done better
   • I wouldn't want to take away Larry's revenue stream
3. Something that's been on my TODO list for a long time now is to take a serious look at Arch by Tom Lord.
4. If you do go and invent a new version control system, please don't forget to create a VCS module for it (there are already modules for CVS, RevXML, RCS and Perforce).
5. I use CVS to manage the source repository for CSSC, but I have my eye on Subversion.

ObCVSWhinges:
On the whole CVS is good - in fact, in the most part it's good enough (ever heard the expression "The best is the enemy of the good"? Well, the good is the enemy of the best, too!).

My pains with it are

• It gradually gets slower as you add more files and revisions
• Inadequate hooks for plugging it into your bug tracker, work management system, database, toaster etc. The *info scripts exist but aren't really enough.
• There are obscure gotchas that mean I can't unreservedly reccomend it for use by just anybody (i.e. people without training/skills). These issues include problems with binary files being edited on several branches at once.

For me, all the hype and PR have tended to obscure what's really in .NET. Also, many of the language interfaces to it (VB, ASP, C#) do not have open source implementations. (Maybe Miguel de Icaza will be done soon:)

That said, there are probably some really good sound technical ideas hidden in .NET after the surrounding marketing has been washed away.

Subversion lets you use different low level layers for actually storing files, pluggin things through an API.

That great idea is compounded by their objective of using WebDAV as a lower level layer.

I like the idea of having an XML description of actions that need to be taken for a version control system. Perhaps .NET has some good ways for doing this, but I'm fearful of a simple open source tool acquiring too much bloated overhead (much in way that SOAP bloats XML-RPC) that could slow it down and make it dependent on more network activity than is always necessary.

OTE: Subversion is now self-hosting -- to obtain a working copy, you must use Subversion itself, not CVS.

Is Subversion stable enough for me to use for my own projects?

Yes and No.

We say "Yes" because we do believe that Subversion is stable and have confidence in our code, so we've been self-hosting since September of 2001--eating our own caviar so to speak>/i>.

We say "No" because if something goes wrong with our svn repository, we've got a horde of active developers who will stay up sleepless nights hunting down the problem and rescuing our data. As altruistic as this horde might be, they don't have time to rescue the data for thousands of people who are storing their data in a pre-alpha product.

We say "No" because there's a good chance that the filesystem might change before we go 1.0, and we don't plan on writing and testing and shipping conversion utilities.

So, as long as you're willing to take those risks, then go right ahead and use Subversion.

In fact, the only thing I see as a major issue with svn (other than their fscking slow-ass site) is the fact that it is pre-alpha code. I want to use it now, dammit. In fact, I'm just starting two new personal projects; I think I'll give it a shot. Worst case, if they do change the filesystem, I'll write a damn conversion program myself.

• 100% java
• BSD licence
• only a few UML v1.3 features not (yet) supported
• import/export from/to Rational Rose using XMI are not (yet) totally working but a third party tool exists to do the translation (can't recall the name)
• SVG output (!)
• you can add all the plugins you want/need on top of the core app (MOF metamodel and GEF graphical framework)
• commercial support provided by Gentleware (if you've got a PHB to convince)

• 100% java
• BSD licence
• only a few UML v1.3 features not (yet) supported
• import/export from/to Rational Rose using XMI are not (yet) totally working but a third party tool exists to do the translation (can't recall the name)
• SVG output (!)
• you can add all the plugins you want/need on top of the core app (MOF metamodel and GEF graphical framework)
• commercial support provided by Gentleware (if you've got a PHB to convince)

I have not tried it myself, but I have heard very good words about subversion, a version control system which is in heavy development right now. The developers have looked at various existing version control systems, and found their pros and cons. Then they designed a sane system, something between CVS and Perforce. Then they started coding.

Subversion is not finished yet, but according to others, it is quite usable already, and the subversion developers are using it themselves.

There was an article in Linux Journal about subversion i February.

I have not tried it myself, but I have heard very good words about subversion, a version control system which is in heavy development right now. The developers have looked at various existing version control systems, and found their pros and cons. Then they designed a sane system, something between CVS and Perforce. Then they started coding.

Subversion is not finished yet, but according to others, it is quite usable already, and the subversion developers are using it themselves.

There was an article in Linux Journal about subversion i February.

But where is Subversion? It seems to be forgotten on the Tigris.org backyard.

It seems to be under very active development. They have been getting out regular releases and have an alpha release planned for June.

Check out its status here

-Bruce

Oh, do I hear "but that isn't fun?"

CVS is not perfect. But maybe, just maybe, by using a technically better versioning tool for a while, it will free up enough programmers to write an improved, free versioning tool? Did anybody think about it this way?

In the long run, maybe, just maybe, the free & improved versioning tool will put BitMovers out of business. RMS himself advocates accepting short-term pain in return for long-term gains. ref: the existence of the LGPL!

Last I heard, Subversion is supposed to be a replacement for CVS. Why aren't more people joining the Subversion effort?

Linus is just trying to make the GNU/Linux kernel better - which benefits everyone. So shut the fuck up, quit complaining about non-free GIFs... er, versioning tools, and go write a better one!

For all the borking I hear on this thread about the "badness" of BK, I have to ask if there are any viable GPL'd alternatives to it?

I am in a position to get a VC system in place at work. We have looked at a lot of commercial stuff, but they leave me kinda dry and with a lot less money. The are very proprietory, usually only work on Windows, and don't work & play well with others.

I think I would love to implement CVS, however there are some problems with this solution. Although it is "Free", GPL'd, open, & x-platform, it is also somewhat difficult to setup, use, & maintaine from an enterprise view.

There are Subversion and Arch. CVS/NG is being talked about.

McVoy managed to piss off ESR, who is, as you all know a strange mix of valuable open source contributor and condemnable weapons idiot.

Regards, Marc

Using BitKeeper troubling.. why cant someone just beat linus over the head with cvs.. and script that up for him? have perl will travel

Linux has made it very clear that CVS doesn't get anywhere near to to meeting the needs of the kernel developers. There are a couple of interesting free revision control systems, arch and subversion on the horizon, which may get adopted at some point in the future however.

john

Not true! The DeltaV specification, "Versioning Extensions to WebDAV" was published as RFC 3253 (Standards Track), in March, 2002.

This specification provides versioning, configuration management, workspaces, and logical change tracking capabilities.

First out the door with an implementation is the Subversion project, which is developing an Open Source replacement for CVS.

While we're at it, two other WebDAV specifications nearing completion are:
WebDAV Access Control Protocol
DAV Searching and Locating Protocol

For easily installable Java applications take a look at Java Web Start. I guess installation/execution of Java apps cannot get much easier than it gets with Java Web Start. Just install the Web Start software which comes equipped with a JRE and click the link to the software you want to launch. Web Start takes care of getting the newest version, caches the latest version and then executes it using Java's extensive security mechanisms. You want to get access to otherwise restricted areas of the user's computer? Go ahead and digitally sign your application - the user can now grant the application the rights it needs. For an example application, look at the Web Start version of TopCoder or the UML modelling software Argo UML.

Java came a long way, baby - and there will be a lot of way ahead. With the next releases of Java focussing more and more on speed improvements, your speed problems should go away as well. And, BTW, there currently are Java apps which run very fast - take a look Borland JBuilder - a Java IDE written in Java. And it runs smooooth!

.bbr fear women playing with delete functions - next time it could be you.

Might want to look into subversion which also supports atomic operations.

It's intended as a conceptual replacement for CVS, but without all the things that people hate about CVS (i.e. cvs doesn't track metadata, renaming files is evil, etc).

Check out subversion.
It's CVS, but better and based on WebDAV for RPC and BerkeleyDB for storage.

Cheers,

pointer

In opening, what is wrong with CVS? If it's such a big piece of shit, where are CVS' authors to address the problem?

With arch, subversion, Bitkeeper, Perforce, and Starteam there is a large group of choices for someone setting up a new project. Having the choice is a good thing, but there is a lack of good information on the topic.

I'm looking to get a RCS setup soon for an open source project, and this article reminded me that maybe CVS isn't a given. To decide, I can go read the docs for each, set each one of them up, and find the one that best suits me. Man, that's gonna suck.

Someone needs to sit down with each of these (and any others of significance) and really do a comparison. Investing the time to get your brain wrapped around a new RCS is very expensive.

I think this article is basically ZDNet trolling again. After all, the more "controversial" the article, the more hits they get = more ad revenue.

So today's developers will use one of three languages: Java, C# or VB.Net.

Strange, a lot of projects I'm familiar with don't use any one of those languages. I think it depends who you talk to.

I think the author believes in two common fallacies:

1. C++ has some plus signs after it, so it must be a replacement for C
2. All problems in systems programming are trivial and have already been solved, and will never need solved again, so there's no need for really low-level languages.

I'm sure the argument is a lot more valid for big corporations, but they've always been bastions of VB and "4GL's" (even when 4GL was just a marketing term). Basically, /. has been trolled again.

I've been using Apache 2 on Linux and FreeBSD for about 2 months now (got into it while playing around with Subversion, another project that seems to be making excellent progress), and IMHO it is really going to rock the server world. Some major plusses:

• ./configure; make; make install (almost). No more APACI, thankfully.
• APR. It's already starting to be used by other projects.
• Totally rewritten mod_cache, mod_proxy, etc. Works much better now!
• Will actually work on Windows (well, some may not see this as a benefit, but whatever).

People have been complaining that Apache 2 is slow to come out, but from what I've seen lurking on the mailing list, it's because they want to ensure the quality of this release. They've also been talking about how they want a lot of beta testers, because (<rumor mode on>)they want to release soon, maybe even from 2.0.32. So get out there and beta test it!

8.WebDAV support for Digest authentication

That's interesting, because AFAIK MSIE still doesn't support Digest authentication for WebDAV shares (a.k.a. "Web Folders"). (At least, IE5.5 doesn't, and IIRC neither does IE6). It seems to get confused and try to access the site using FrontPage extensions instead, which of course doesn't work because it's running Apache 2.0. That makes it hard to interoperate MSIE with other WebDAV products (like Subversion), at least if you're using Digest auth. I'm glad to see at least someone is actually trying to implement web standards, instead of mixing them together with proprietary stuff. Anyway, just my $0.02.
---
Windows 2000/XP stable? safe? secure? 5 lines of simple C code say otherwise!

What about ArgoUML? It's written in Java and I'd say it's successful....

Let's look at this a little more objectively. Hosting kernel.org costs about $80,000 a year (Larry McVoy posted this number to lkml about a month ago) at the least. It's an ftp site. That's bandwidth, not any warm bodies doing admin, not any fancy database stuff, nothing fancy just an ftp server and a minimal web site. Sourceforge has to cost 20 times more, probably more, to run. I have no idea what the numbers are but it has a staff and a huge amount of resources to manage and keep running. Personally, I'd assume that it's in the neighborhood of $5million+ a year, that's just my half-assed guess though. That's some substantial output for most companies, at IBM you can't spend that kind of money without producing something, people notice chunks that big. At most places, that kind of funding simply isn't available for something like that. At some point the free ride has to end, or something has to come out of it, or something has to change. Even a company like MS would see $5mill on the books in red ink and not black and there would have to be some reason to justify it and goodwill towards the community might not be enough.

Then with subjects like these, things rise up. Well they should trim dead stuff out of the tree, trimming the "dead" stuff is silly becuase it might be useful to people, that's the whole premise, if it's in use anywhere then it's not really dead. It might be dead to you and me, but that guy who is using it might want it. They should do x, y, or z to better support projects like q. They could do this or that. I think the most alarming propect is that there will be code in SF and it could be lost because of a policy change. I can get over most things, the changes to the mailing lists, and various other things they've done, it's free and you get what you pay for but a big part of the justification has been to promote interaction with developers to give VA a community they have close ties with and to promote open source software development. The idea of losing code is appauling, SF no longer serves a big part of its purpose at that point. That's what brings credibility in to question, what are they doing to prevent that from happening? Can I buy a set of DVDs that have SF backed-up on to them? Or is this it, the policy change is that there won't be any warning of future policy changes and those might cost you your code. I understand that they might have to sell stuff, or charge for services or do lot's of different things. I also understand that services like SF are prime for pirates and porn hustlers and others to use to propagate data and they need to protect themselves. It's time to look to tigris, Savannah, and Berlio more seriously.

I wonder if there is something we could add to licenses that would prevent a place like SF from shutting down and taking your code with them.

The subversion developers are coming close. The current version supports almost everything CVS does, plus more. See the homepage, for more details. Also, something not noted on the webpage is that someone has actually got a cvs2svn repository converter which can convert repositories. Note that it doesn't follow branches or tags yet, but is able to convert the entire mainline gcc development tree to an svn repository.

Now I'm confused!

I've been using CVS for years and read with great interest the recent Linux Journal article about the Subversion project to created a CVS replacement that is better than CVS.

Then I see a Slashdot story about arch.

Now, my FearLessLeader starts using Bitkeeper.

Should I move from CVS and, if so, which is best?

Subversion was deliberately designed to address CVS's shortcomings, not to break new ground. Our philosophy was essentially conservative: CVS basically works, but has some bugs and maintainability problems. Let's keep the model and fix the problems. Result: Subversion.

The ideal situation is a world where both models have good, free implementations. Then we'll all very quickly find out which model works better. :-)

-Karl

Ken

They are. Try this for size. Still in development, but it's now self-hosting (subversion uses subversion as their VC now).

Argo and Dia are the two I know that are OSS - I have played with both, but I didn't do enough real work for either one to tell you if either will work for you - Argo seems a little more baked...

ArgoUML looks pretty cool. Somebody should tell those guys that Java doesn't have multiple inheritance of classes (without use of inner classes) though. Probably a C++ coder came up with the nice java example on examplethis page.

UML is the language, not the essay. If you don't know what you want to say, any language will do - but if you do know what you want to say, at least in Software Design - UML gives you a way of describing it precisely.

In my current organisation, we took a couple of "difficult" projects, where the customer's needs weren't clear, and converted them over to UML. The results were amazing. In one case, we were able to make the customer gain a new appreciation of the complexity, and could re-negotiate the contract. In another scenario, we saved a huge amount of time in "usability" design, where software features were grouped around business workflows.

Where it failed - and consistently failed - was in teams where everyone did not switch over to UML (or converted under duress). We had cases where developers would rewrite the explicit instructions of UML into plain English, and loose the level of detail in the design.

If you want a real example of the power of UML, have a look at Argo UML. The tool really takes UML one step further, allowing you to see the Java code generated, and do full roundtrip in a single editor.

English is an imprecise language, and is very unsuited for expressing program functionality. It's even worse for documenting requirements. Thats where UML fits in, and in MHO, it does work.

The problem is that there isn't a decent multi-patch versioning system out there

Uh, yes there are. Perforce, aide-de-camp, bitkeeper, and others all do this just fine. I haven't used squeak much, but I think this is also how the built-in version control in their smalltalk image works as well. Every change management system that uses changesets works pretty much exactly this way.

CVS basically sucks, which is why some people are trying to replace it. It only gets used because it is popular and free, not because it is technically superior. The only thing it is better than is RCS/SCCS. Every other possible solution is no worse, and usually much better, than CVS.

Those who wish to play around with UML might be interested in Argo UML. It is a free Open Source UML modeling system. It has a few shortcomings but it can give you some hands on experience with a pretty good UML CASE tool.

A missing piece in the acticle is the importance of Apache 2.0 for WebDAV. WebDAV is a HTTP-extension making the HTTP-server a real fileserver - Apache 2.0 comes with full WebDAV support. As WebDAV is quite flexible and allows stuff like meta-data, versioning and different authentication mechanisms (that are unfortunately not finalized yet) it is a possible successor for both NFS and SMB/Cifs.
WindowsXP supports the mounting of WebDAV shares, as does Linux with the help of the DAV filesystem driver. And Apache could be the standard fileserver... scary.

Has interaction with the open source community contributed to any changes in Sun's internal development practices and/or toolset and/or do you see this happening in the future? I'm speculating that perhaps the toolset being developed at Tigris may be funded indirectly by Sun via CollabNet with an eye towards internal use in addition to use in Sun's collaboration with the community on projects like OpenOffice, Netbeans, and JXTA.

I'm excited about several potentially significant projects that may have their first "stable" releases next year. Everyone knows about OpenOffice, Apache 2 and Mozilla (I'm surprised that neither article mentioned the last two). Here are a few others:

• Subversion version control rethought, could replace CVS as free software tool of choice
• E capability secure programming
• Reptile reputation-based content aggregator

So they've gathered a lot of requirements and done a lot design, but, as the lady says, "Where's the beef?" Why do all this work to replace open source tools that already exist? Why not take those tools and contribute back to them, or if the project owners don't like the contributions, take the things they like and build from there?

Perhaps this from their web page: "The Advanced Computing Laboratory at Los Alamos National Laboratory is providing $860,000 of funding for Software Carpentry, which is being administered by Code Sourcery, LLC." And this: CodeSourcery also provides training and strategic consulting services for companies considering the adoption of free or open source software.

Why use the MIT License? Why develop in Python? Why require that the submissions to the design competition not contain any source code? They require a language but not any source? Am I being too cynical in seeing how, after all this contributed design, coding, testing, etc is done, at taxpayer expense, Code Sourcery is now in a tremendous position as the sole-source solution for support and training to the shops that choose to use these tools? And to notice that choosing the MIT license allows them to take and wrap up all the source code into their products and not give back anything? These are questions that either are in the FAQ but not clearly answered, or not spoken about. Even the SC site itself has been retired and archived.

Thanks, I'll stick with XUnit, Bugzilla, cvs (and subversion when it's ready. For build config tools, well, if you do cross-platform C and C++, then autoconf or its successor, but that's just for one language and set of development requirements.

I'll be interested to see if anything widely used comes out of this exercise. So far, of the all the tools implementations promised for "Summer of 2001", we have QMTest 1.0. The rest? Late and unfinished.

Ok, maybe HTTP/1.1 isn't quite as good as you might want, but if you add in WebDAV, it becomes an incredible replacement for FTP. And CVS. And NFS...

WebDAV servers support (by default) Locking of files, Put of files, PropPatches of custom properties, PropFinds of those custom properties,Moves, Copies, and Delete. Plus Get, Put, Head, etc. that HTTP/1.1 provides.

Furthermore, there are standards defined for:

• Access Control Lists (so you can decide WHO has acess to WHICH of your files with a fine level of granularity).
• Versioning (AKA DeltaV) checking in/out, etc.
• Searching (AKA DASL) using XML-based grammars.

Furthermore, MANY clients support saving and locking of files over WebDAV. In particular:

Microsoft Web Folders uses a WebDAV server as a file system and ships with Win 98 +

Mac OSX has WebDAV support built into the File System

Several Open Source Developers are working on a quite-functional Linux WebDAV file system that you should check out.

Adobe Photoshop 6.0 does saving over WebDAV. So does Macromedia's Dreamweaver UltraDEV. So does Microsoft Office. Many others do as well.

Mozilla's Composer WILL HAVE WebDAV support (eventually, see bug #13383)

Wish to try it out? Got apache? A level 2 DAV server (sometimes) ships with Apache. It's called ModDAV, and it seems to be quite easy to setup.

Dav on

Then, in theory anyway, you instantly have the ability to PUT files there, LOCK them, etc.

Read more at The mod_dav page

-marick

P.S. Want more functionality?

Check out Sharemation for a free (5Meg) WebDAV account that has DeltaV, ACL, DASL support. (And yes, I work for Xythos Software, we host sharemation and sell the Web File Server it's based on.)

Or go check out Tigris' Subversion a highly capable free DeltaV enabled DAV server.

Have you guys considered simply joining an existing next generation source control system rather than starting from scratch on a completely new one?

David

I recently had the job of evaluating what SCM system to use for our company. We were using CVS at the time.

I believe that the complaints your bosses have about logging and concurrent editing can all be fairly easily fixed in CVS.

The major gripes we had with CVS were:
. slow (see below)
. merging between branches was miserable, because the system didn't keep track of what had already been merged
. renaming files lost all the history
. windows interface was cumbersome

We actually worked with a system layered on top of CVS that allowed us to submit batches of files at once, in a single transaction. This was the major cause of slowness, and CVS didn't really support transactions, so in some sense we were just fooling ourselves. The other major cause was doing a 'cvs update' on a large tree could be slow.

Most of the problems we had could've been fixed
if we spent the time to fix it. Some (transactions, renaming) we couldn't really fix at all. But, when I looked at everything I wanted to fix in CVS, I found that I had just described Perforce's feature set, and when I looked at how they implemented things, they did it like I would've. Plus, comparing the cost of Perforce (relatively cheap) to the time it would've taken us to implement the same features ...

So, we switched to Perforce, and I've been (more or less) happy since. The branching structure is a little weird (compared to ClearCase's, which is the most intuitive I've seen), but we're learning to live with it.

At a previous company, we used ClearCase. This was also a fine product, and it does a few things that no other product does, but it's very expensive, and a major hassle to administer.
ClearCase (at least in the mode we used) implements its own filesystem, and can provide a level of security that the others just can't. But, is this worth paying 10+ times the amount for it?

I also looked at AccuRev. This was about on a par with Perforce, and had one or two features that looked really cool. But, in the end, Perforce won mostly because we went with the product that had bigger market share and more people had used it before.

VSS wasn't an option because we're a mostly linux based shop, and because I had heard many of the complaints that others are making as well.

PVCS I think is mostly an also-ran in this day and age. I think most new source-system users use one of the other previously mentioned systems.

One new open source project (Subversion) looked promising, but it was too immature for us to use.

Bitkeeper also looked interesting, but not enough so to beat out Perforce or AccuRev.

Another thing you might want to consider is how well the SCM integrates with a change mgmt system (or bug/task database). Perforce has a simplistic change database built in, but it's good enough for what we want it to do, and it can also be used with Bugzilla and a few other systems. Of course, CVS and at least ClearCase can do these as well. I've found Bugzilla somewhat cumbersome to use on limited inspection. Other freeware systems (GNATS, for example) are very weak.

I'm not sure what your point is (sorry) but if you're at all interested in source control software, you might want to check out Subversion. I don't think their scope is larger than what is said in their goal statement; "to build a version control system that is a compelling replacement for CVS", but perhaps this is of interest to you... Personally I find their reliance on Apache slightly odd, but I guess a web server is always fun to run anyway. I'm looking forward to the first alpha release which should come this fall.

They all lack many essential features. They all have web-based GUIs that are tighly coupled with the back-end logic; that is, they have no back ends. Thus the default GUI is the only GUI you can ever realistically put on top of it. A lot of people are missing out on the MVC model these days. What you need is a programmable back end accessible through a cross-platform API (based on CORBA, SOAP, XML-RPC, UNO, anything that strikes your fancy). Then you can leverage the back-end support for clients. One can be a powerful reporting tool with graphing capabilities. Another one can be a wxWindows-based portable GUI for modern desktops. Another one can be a common-denominator HTML-based GUI for browsers. Etc.

Current GUIs are all crude and cluttered and obviously designed by programmers with no interface design background (and by that I don't mean graphical design, but functional design). Many are ad-hoc systems thrown together using PHP. Presumably the poor devils think that by slapping it on SourceForge or Freshmeat it will magically bloom into a usable product. Nuh-uh.

Another common problem with these systems is that they're fundamentally bug-tracking systems. When you get to a certain point in development, you realize that a better all-embracing concept is the idea of issues -- a generalization of problems that aren't specifically related to code. There is a popular fork of Bugzilla, for example, called IssueZilla.

The only system that was mildly interesting was Keystone, which provides some interesting form-based extensibility -- basically, if I remember correctly, the schema is malleable, so you can add stuff like time estimation numbers, completion progress, or other metadata that would be useful in your project. Also Keystone supports the notion of subtasks: any bug "slip" can have another slip as its parent. This is more elegant than Bugzilla's dependency system. Unfortunately, Keystone sports a GUI from hell. (Applying CSS to it might sound fun, but it isn't; their HTML isn't very CSS-friendly, so to do anything radical you have to delve into their HTML generation code).

We currently use Bugzilla. It's currently the best system out there, but that doesn't say much. We are pretty excited about Scarab -- this is a project where the developers actually sat down and designed it beforehand (wowee).

CVS has some big flaws, but it is quite nice. I haven't seen a source code control system that didn't have problems. Anyway I think you should look at subversion they are directly addressing CVSs big flaws. It looks like the authors know a lot about CVS, and like CVS, so whatever they build will probably not suck more then CVS...

...except they have a lot of dependence on Apache and the DAV module. So that part at least sucks differently then CVS, and maybe more. Hopefully subversion will get far enough along that I can find out for myself though.

If McVoy would stop playing silly license games with Bitkeeper so he can try to become the next Sourceforge (sorry, but you lost), then the world would probably beat a path to his door.

Bitkeeper does look cool. I don't think subversion can do the same sort of hierarchy of repositories that bitkeeper can. Anyway I don't think McVoy wants to be the next SourceForge (are they making money?), he want to be the next PerForce, CodeSafe, or whoever else has made a ton of money directly off version control software.

In the meantime, are there any robust and free alternatives to cvs?

No, unless by "robust" you really mean "alpha quality, not trusted to be self hosting yet". Try again in six months :-)

I've been using CVS for five years, and although it took me a while to learn its twists and turns, I now think that it is an absolutely excellent piece of software. I have been forced to use other major alternatives at various companies I've worked for (the most notable being Perforce and MS SourceSafe) and CVS is vastly superior to the others I have used. (I have not used Bitkeeper, so I can't compare it.)

Granted, it is missing a few important features such as the ability to flags files as having no revision history (so that large binaries can be stored without killing the server on commits and updates) and the ability to rename files and directories. But those are relatively minor complaints compared to all that it does do well.

The project I'm excited about is Subversion.

If you want to avoid feeding the M(assive) S(hark), you can build on top of CVS. Set up a web page for uploading to a document directory, and have a server-side daemon grab files that appear in the directory and shove them into CVS. However, there is more to version control than just storing multiple versions of a document, and even CVS doesn't do too well on non-text docs, like you mentioned.

I suggest you take a look at SubVersion, written by (among others) the guy who developed CVS. The alpha is due out on (or shortly after) the 15th of this month. The intent of SubVersion is to be everything that CVS was, plus most of what it wasn't. SubVersion, among other things, versions file meta-data (including directories), and allows for file renames. Apropos to your question, SubVersion allows for client-side diff plugins, so that it can intelligently and efficiently store revision differences. Also apropos to the question, SubVersion supports WebDav... you can't get much more user-friendly than that.