Upgrading Software

[crelf]

QUOTE(Pablo Bleyer @ May 6 2007, 11:19 PM)

IMHO, if you are not having major problems with 8.2, I'd suggest that you stick with it and avoid 8.2.1.

I think that's esoterically true for any upgrade of any software - you need to consider the motivation behind upgrading, and not do it "just because a new version came out". I had a customer once (no names) who we about a week away from production (yes, that's right, I said production), and a new major version of LabVIEW came out - they thought it was kind-of like using Adobe Acrobat Reader: all you need to do is install the new version and everything will work just fine, right? Sure enough, it didn't, and the production timetable needed to be adjusted

So, here's the start of a list of what I consider to be valid reasons for upgrading any software (it's just a start, so please feel free to add your own suggestions!)

1. You've got an issue that the new version will fix (and hopefully not create other issues that you can't workaround)
   
2. There's features in the new version that you need to use
   
3. You need support from the software company, and they don't support the older version anymore
   
4. Newer required external components (drivers, OS's, hardware, etc) don't work with your old version, forcing your hand to upgrade
   

[TG]

QUOTE(crelf @ May 6 2007, 04:42 PM)

I think that's esoterically true for any upgrade of any software - you need to consider the motivation behind upgrading, and not do it "just because a new version came out". I had a customer once (no names) who we about a week away from production (yes, that's right, I said production), and a new major version of LabVIEW came out - they thought it was kind-of like using Adobe Acrobat Reader: all you need to do is install the new version and everything will work just fine, right? Sure enough, it didn't, and the production timetable needed to be adjusted

So, here's the start of a list of what I consider to be valid reasons for upgrading any software (it's just a start, so please feel free to add your own suggestions!)

1. You've got an issue that the new version will fix (and hopefully not create other issues that you can't workaround)
   
2. There's features in the new version that you need to use
   
3. You need support from the software company, and they don't support the older version anymore
   
4. Newer required external components (drivers, OS's, hardware, etc) don't work with your old version, forcing your hand to upgrade
   

To this list I would also add

* You've had a good "read" of comments on stability and features (or from a lack of bad comments) on independent forums.

* You have carefully considered the benefits (or the implications of not keeping up), and compared this to the potential risk of upgrading

and are comfortable with taking the plunge.

At the end of the day, upgrading LV is tedious (would take at least several hours to upgrade 5 legal rigs here) but also prone to some form of risk.

The "unknown" is a fearful thing. I am aware this is not a courageous approach,

however past experience with upgrades from "All" software vendors usually translates into going slow and carefully.

[Chris Davis]

Chris,

I don't know if you meant to post this in the lounge, it sounds like a response to a different thread. I personally think this topic might make a good wiki article. I'm sure you'll get lots of responses so I'll throw my two cents in too.

Upgrading versions of software, LabView or any other software, is a time consuming thing. It should never be done in or around "crunch" time. Having a source code control system is a must to make the process of falling back to an older revision easier. Use a virtual operating system to test out your "this upgrade is a piece-of-cake/walk-in-the-park/5-minute-job" theroy.

Finally, my rules to upgrading new software

1. Backup

2. Backup

3. Backup

If all else fails, remember to use your backups.

[crelf]

QUOTE(chrisdavis @ May 7 2007, 01:29 PM)

I don't know if you meant to post this in the lounge, it sounds like a response to a different thread.

It is, but I didn't want to hijack the thread: I figured a general upgrading thread was importnant to see on it's own.

QUOTE(chrisdavis @ May 7 2007, 01:29 PM)

I personally think this topic might make a good wiki article. I'm sure you'll get lots of responses so I'll throw my two cents in too.

That's an excellent idea: are you volunteering? I'm a bit busy writting the http://wiki.lavag.org/index.php/Custom_LabVIEW_configuration_file' target="_blank">Custom_LabVIEW_configuration_file wiki at the moment, and if anyone has a spare ten minutes, I agree that this thread could be an inpotant wiki article...

[Chris Davis]

QUOTE(crelf @ May 7 2007, 07:50 AM)

That's an excellent idea: are you volunteering? I'm a bit busy writting the http://wiki.lavag.org/index.php/Custom_LabVIEW_configuration_file' target="_blank">Custom_LabVIEW_configuration_file wiki at the moment, and if anyone has a spare ten minutes, I agree that this thread could be an inpotant wiki article...

I'll work on it, although I'm not quite sure how to start a new wiki article, I've just edited one that already existed. Guess I'll have to learn.

[crelf]

QUOTE(chrisdavis @ May 8 2007, 10:00 AM)

Guess I'll have to learn.

http://wiki.lavag.org/index.php/Help:Starting_a_new_page' target="_blank">Here's a good place to start.