Doing an SVN checkout? No web for you!



  • Web browsers are way too dependent on hard disk caching. I had a large SVN checkout to perform and was going to get caught up on some web reading in the meantime. Except somehow disk I/O prevents my browsers from loading web pages and executing scripts in a timely manner. Totally unusable.

    inb4 "TRWTF is Subversion".

    Bonus Discourse WTF: Apparently what I'm posting is very similar to the "How Long Do You Sleep" thread, "How do we format code?", and "Hi, we're Intel and we don't give a shit about standards. Come work for us!" And more. The similar topics list seems to be chosen at random.



  • @mott555 said:

    Subversion

    Is this something you developed in-house?



  • SVN isn't a total TRWTF. I prefer it to Git for the stuff I do myself (where it's just me doing commits)

    Also, that list isn't totally random. Mostly random, yes, but not totally since when I did my series in Sidebar WTF, it correctly found the similarly-titled topics.



  • At my last job we tried Git when we started our iPad project because apparently Git is the preferred source control system for Xcode. We couldn't figure out how to use it, had trouble finding meaningful guides, and too many online Git discussers had the same superiority complex as Linux fanbois that meant they'd rather make fun of us for not knowing how to use it instead of helping us learn the system.

    So after an hour or so we gave up and went back to Subversion. It just works. Well, unless you're using Xcode, because Apple's SVN client and/or Xcode both should have been aborted.



  • I regret I only have one like for this post. I agree with your assessment on the whole fawning over Git (remember who made Git, of course!) and that your take on Apple's SVN client and/or Xcode is also completely correct.



  • @mott555 said:

    too many online Git discussers had the same superiority complex as Linux fanbois that meant they'd rather make fun of us for not knowing how to use it instead of helping us learn the system

    @Arantor said:

    I agree with your assessment on the whole fawning over Git

    I didn't experience this problem, although I never had to deal with the community directly. After reading through the Git Book I understood the system reasonably well. Since then, I've been able to find solutions for any specific problems by searching online.

    Admittedly, the solution is often completely unintuitive, but never too hard to find.



  • That's just it. Our experience has been in trying to learn from the people that use it already, with their self-satisfied smug "if you have to ask, you'll never know" attitude.



  • @Arantor said:

    Our experience has been in trying to learn from the people that use it already,

    Oh, like, in person?



  • I wanted to like this but I'm saving my likes for the Likes thread. So accept this token of my appreciation instead.

    +1



  • Not necessarily in person, but over IRC, forums, chat, whatever.



  • @mott555 said:

    Bonus Discourse WTF: Apparently what I'm posting is very similar to the "How Long Do You Sleep" thread, "How do we format code?", and "Hi, we're Intel and we don't give a shit about standards. Come work for us!" And more. The similar topics list seems to be chosen at random.

    I buttume that it is done like the Suggested topics. At first relevant results are displayed. Once those are exhausted, then it resorts to randomness.



  • @mott555 said:

    It just works.

    What you mean is: "I already know how to use it."

    This has nothing to do with how well it works.

    And this attitude is what people are making fun of. [points and laughs]


  • :belt_onion:

    @mott555 said:

    Bonus Discourse WTF: Apparently what I'm posting is very similar to the "How Long Do You Sleep" thread, "How do we format code?", and "Hi, we're Intel and we don't give a shit about standards. Come work for us!" And more. The similar topics list seems to be chosen at random.

    Psh, all of those Topics contain the word "the". Clearly similar in content.


    Filed Under: I didn't bother to look, maybe they don't, just a guess.


  • No, it's not.

    I already know how to use SVN. And for my tasks it works just fine. It's not that I can't use Git - because I can. It's that SVN suits me better, and separately when I first tried to migrate to Git, I was confronted with a complete wall of what amounts to "if you have to ask, you'll never know". It wasn't about the fact I knew SVN, it was the fact that they were so busy telling me how wonderful it was, they never bothered to help me get started, so I just spent time being told how I was doing it wrong (but never being told how to do it "right")



  • I'm testing the waters with Git at work. I am still failing to see what it has that makes it so much better the subversion.

    Also the real WTF is not having an SSD.



  • I wasn't referring to you. (so "you" in the following lines doesn't refer to you, it refers to a hypothetical individual that possess the attitude exhibited in the post I was replying to).

    If you know both tools, then you're in a position to judge which is better suited for a given task.

    If you only know one, find yourself initially confused by the other, give up learning the new tool, and then returning to using the one you're already familiar with, claiming it "just works", you could be holding yourself back.

    And you wouldn't even know it.



  • Except that wasn't the problem we were complaining about. Both of us complained about having a tool that we were familiar with but for whatever reason were looking at a new tool - and ended up being beaten over the head with our own lack of knowledge by its smug self-righteous evangelists. The whole thing of 'it just works for us' is a side effect of that situation.



  • It's not so much git specifically, as distributed version control in general, that people like so much.

    Local commits can be really nice.
    Some people use them alot and find them very helpful.
    You can't do local commits with SVN.
    So people who like local commits won't want to use SVN anymore and want it to go die in a fire forever so they'll never have to live without local commits.

    @russ0519 said:

    Also the real WTF is not having an SSD

    True dat.



  • I wasn't trying to rebut your complaints about evangelist's attitudes, I was filing an original complaint.



  • Originality? Here? PAH!



  • @mott555 said:

    The similar topics list seems to be chosen at random.

    Considering that all topics on this forum get derailed at least once, this is probably an accurate algorithm.


    Filed under: Squirrel!



  • @russ0519 said:

    Also the real WTF is not having an SSD.

    True, but it's difficult to fit 4 different Windows installs and 6 different Linux installs all on the same SSD without spending a TON of money on a massive enterprisey SSD.

    Part of me says TRWTF is that we support Linux with our hardware but that's not my call.



  • After using VSS and TFS for nearly a decade, I can barely tell the difference between SVN (used at work) and Git (used at home). Probably since I use the Tortoise GUIs for both.



  • Local commits can be really nice.

    Yep exactly, it's extremely helpful.
    In my embedded workflow. I can be comitting 10-20 small changes in just a hour. Usually its value adjustments or logic adjustments to firmware operation and I document changes separately in the commit messages. But once I am done I can push the entire series to the main branch. Alternatively I can edit my patch history and actually merge commits together, or even delete them from my branch's history :)

    Then there is branching, git branching, is just amazing.
    One workflow I have with coworkers on a project.
    master - main stable branch
    coworker1
    mine

    Me and coworker both checkout on master at the same time. My coworker makes changes to his branch and now merges into master.
    The master is now ahead of my branch and its changes affect mine.
    Now I want to finally merge my branch to master.
    I could do a merge on master and resolve conflicts in the master branch which I think is silly and "dangerous" for embedded firmware.
    or I git-rebase my branch!
    git will go back in time to where I branched from master, pull the latest commits in from master. Then attempt to apply my commits on top. I can resolve conflicts safely in my branch. Once done I can now merge a conflict free branch into master once I have verified the firmware still works correctly.

    I've seen "hacks" to do rebasing in SVN but it's absolutely nowhere as clean as git which is just a single line command.

    My company was using SVN before I joined, but now even the old guys are loving the painless branching and merging.

    Don't even get me started about tags and how wonderful they are compared to svn. SVN does a "snapshot" or just basically a copy to do a tag. Git literally tags a commit hash. Benefit, huge space savings. Especially now because I don't have to exclusively pull the trunk to avoid the 3000+ tags. I just pull the git repo and tags are in the meta data and I can request them any time.


  • Discourse touched me in a no-no place

    @mott555 said:

    Bonus Discourse WTF: Apparently what I'm posting is very similar to the "How Long Do You Sleep" thread, "How do we format code?", and "Hi, we're Intel and we don't give a shit about standards. Come work for us!" And more. The similar topics list seems to be chosen at random.

    Related: http://what.thedailywtf.com/t/unneces/1412?u=pjh

    With bonus hypocrisy at posts #5 and #8, pointed out in #14



  • It looks like it also checks the OP, not just the title.

    All of your "matches" are image OPs, right?


  • Discourse touched me in a no-no place

    @dhromed said:

    It looks like it also checks the OP,

    Well yes - the list changes as you type.

    @dhromed said:

    All of your "matches" are image OPs, right?

    Most of them appear to be...



  • I personally like bitbucket, it's git source control with a usable interface


  • Discourse touched me in a no-no place

    @russ0519 said:

    I'm testing the waters with Git at work. I am still failing to see what it has that makes it so much better the subversion.

    Try doing lots of merging with SVN. Now try doing it with Git.

    I really don't like Git (I shouldn't need a PhD in Git-Usage to do normal development tasks) but it works rather well and is fast.



  • Well you can get a 1TB Samsung SSD for under $450, which would probably fit all your windows installs and linux installs. Sure it's more then $100 for the equivalent hard drive, but if you're doing that much stuff on ur pc, you probably can really benefit from the ssd.


  • I survived the hour long Uno hand

    And yet our middleware team balked at giving me an extra 60GB for a server....



  • @russ0519 said:

    Well you can get a 1TB Samsung SSD for under $450, which would probably fit all your windows installs and linux installs. Sure it's more then $100 for the equivalent hard drive, but if you're doing that much stuff on ur pc, you probably can really benefit from the ssd.

    And it's going to die in what, two years? I could use a SSD for my VMs, but until they can give me a 500GB / 1TB one in a reasonable price and with a reasonable lifespan, I'll pass.



  • This is at work, it's not my call to get new hardware. And every SSD I've ever owned has died within a year.



  • Rule #1 of hard drives: don't buy Seagate

    Rule #2 of hard drives: always mirror SSDs



  • Rule #3

    TB sized SSD are not particularly good.

    Why does that link claiming to be a 1 TB SSD link to a 250 samsung SSD?

    Also, nice referrer link. Subtle.

    http://www.amazon.com/Samsung-Electronics-EVO-Series-2-5-Inch-MZ-7TE1T0BW/dp/B00E3W16OU/ref=sr_1_1?ie=UTF8&qid=1406057358&sr=8-1&keywords=TB+SSD



  • @Matches said:

    Why does that link claiming to be a 1 TB SSD link to a 250 samsung SSD?



  • Yeah, but instead of just linking to the 1TB, he linked to the 250GB and wants me to click.

    Ain't nobody got time for that.



  • Obviously, you're Doing It Wrong™.


  • ♿ (Parody)

    @delfinom said:

    Don't even get me started about tags and how wonderful they are compared to svn. SVN does a "snapshot" or just basically a copy to do a tag. Git literally tags a commit hash. Benefit, huge space savings.

    Aren't svn copies effectively just links to the source, though? IIRC, no data actually gets copied.



  • That's certainly my understanding of what an svn copy is.



  • At the severe risk of re-railing this topic, there's the possibility of configuring browsers to do one of two things:

    • Use no disk cache and use more RAM instead. I've experimented with getting Firefox to do this before, though I have an SSD so it made little difference.
    • Write its cache to a RAMdisk if you have lots to spare.


  • Rerailing? What the hell is wrong with you? Don't you know where you are?



  • I'm sorry. Jeff told me to do it.



  • @faoileag, you appear to have a new acolyte.



  • Meh, I worded it pretty wrong. It isn't the server side that's the issue. But you are you forced to checkout just the trunk of your repo to avoid the tags folder which will pull all of them and make copies of them. A bit of an inconvenience.



  • Yup, that's definitely an inconvenience.


  • ♿ (Parody)

    Except that a DVCS like git has you holding the entire repo locally. Wanting to check out an entire svn tree is legitimately Doing It Wrong.


Log in to reply