The fact that the company I work for uses SourceSafe is really beginning to piss me off. I think I’m not the only one, but just take a look at this article:
http://www.highprogrammer.com/alan/windev/sourcesafe.html
It fustrates me when only one person can work on a file at any one time. It fustrates me that to edit a file you have to have exclusive checkout. It fustrates me that I can’t do my development work on a branch, allowing for builds to be maintained in an orderly and hierarchical fashion. Not to mention that having to check in untested and incomplete code onto a main branch could, and is likely to, cause problems and errors.
Visual Studio fustrates me that you can’t just create a text file without it being added to the project and it marking it as new content that has to be checked in. Perhaps the minor OCD in me likes the “Pending Checkins” tab to be clear and empty.
Developing in such an environment fustrates me beyond belief. What if my computer crashed? I don’t want to check anything into SourceSafe that is unstable, especially onto the main trunk. I’m aware VSS may or may not have trunking abilities, but I expect they’re shockingly bad. So if I have 5 days worth of work, not checked in, my HDD goes tits up, what then? I’ve lost everything.
This could be saved by using a decent source control package, like CVS which I have used before and love. I’ve not used Subversion (SVN) before, but hear its basically better than CVS. The problem is, I’m working for a company that bums Microsoft software, and doesn’t seem to have time for alternative, and often much better software.
The fact that practically the entire website I write for is based on ASP.NET (apart from the legacy classic ASP stuff which they are replacing with ASP.NET stuff) really grates my bones. The database we use (MS-SQL) is designed badly. They use sprocs (Stored Procedures) for the sole reason that “they don’t have to rebuild a query when it’s wrong”. Thats a poor excuse in my opinion. Why on earth would you BUILD a bloody website anyway?! Until I worked here, you wrote a website, and it just appeared. Building a website?! Pah, to me it seems like an useless step implemented by Microsoft. A website shouldn’t be compiled code. As far as I know (and please correct me if I’m wrong), it’s no more or less secure than a nice neatly laid out PHP script with Apache being secured properly.
And what the hell is runat=”server”? Looking at it, it means “run this on the server”. The whole “custom control” thing is beyond my belief. The code that gets generated, as far as I know, is hardly ever W3C compatible. The Javascript it generates that I have seen is disgusting. Take for example the ID’s that it generates (and this isn’t the longest one I’ve seen):
XXXComboBox1_Menu1_XXXComboBox1_Menu1_p_TreeViewItem_ComboBox_TreeView_TreeView_item_1
In my experience, making an ID this bloody long is very bad practise. I understand that IDs have to be unique, but that’s just ridicuous. A good ID would be something like:
ComboBox1_TreeViewItem1
Or even better for optimisation:
cmbx1_tvi1
Using a good W3C compilant function like “document.getElementById” becomes slow when using such long ID names, especially if the DOM is large.
ASP.NET I find is really difficult to write any kind of decent web application. I would find it a hell of a lot easier with some simple PHP, Javascript and AJAX where required. And I certainly wouldn’t use XML, I’d use JSON for the AJAX (or should it be AJAJ?). XML is slow for Javascript to parse, anything over about 500k becomes laborious for the average user’s PC. JSON is quick to transfer because it’s plain text with very little overhead (all these XML tags make things many times bigger), and parsing it is just like parsing any normal Javascript - for 99% of browsers, nice and fast.
Fustration vented, I don’t feel any better now though because I know once I click “Publish” here, I’ll be straight back into the fray of all the above crap that I have to put up with at work.
