Based on their emphasis on “trivial” and “single-threaded”, I think the point they are trying to get across is that Borg’s performance is worse than something very simple. To put it another way, if you have an algorithm question, and you come up with a solution that performs worse than a naive brute-force solution, then either (a) something is very wrong with your solution that needs fixing, or (b) the problem is actually very well-suited to a simple brute-force solution and your more complex solution should be replaced with a simpler one.
The delivery is unfortunate, since it does come across as “I’m better than you, so your code is bad”. But software engineers are not always the best at communication, and the charitable interpretation here is more like “hey, borg’s performance is worse than a naive implementation, is that a problem?”
To be fair, OP did submit the issue “borg compresses 2.3x faster than casync” to their own issue tracker. But yeah, an issue on the borg tracker should primarily be about borg and not about some other program. It would be different if the issue was more pointed and provided a diagnostic about why borg is slower, but since it does not, it comes off as just humble bragging.
If the author didn’t submit the ticket as a story on here I’d give the benefit of the doubt and say it was a friendly way to help incorporate improvements from their own code into code others use. But given the authored by above, it looks a little more like a humble brag to me.
Isn’t it kind of rude to post this in Borg’s issue tracker?
Based on their emphasis on “trivial” and “single-threaded”, I think the point they are trying to get across is that Borg’s performance is worse than something very simple. To put it another way, if you have an algorithm question, and you come up with a solution that performs worse than a naive brute-force solution, then either (a) something is very wrong with your solution that needs fixing, or (b) the problem is actually very well-suited to a simple brute-force solution and your more complex solution should be replaced with a simpler one.
The delivery is unfortunate, since it does come across as “I’m better than you, so your code is bad”. But software engineers are not always the best at communication, and the charitable interpretation here is more like “hey, borg’s performance is worse than a naive implementation, is that a problem?”
To be fair, OP did submit the issue “borg compresses 2.3x faster than casync” to their own issue tracker. But yeah, an issue on the borg tracker should primarily be about borg and not about some other program. It would be different if the issue was more pointed and provided a diagnostic about why borg is slower, but since it does not, it comes off as just humble bragging.
If the author didn’t submit the ticket as a story on here I’d give the benefit of the doubt and say it was a friendly way to help incorporate improvements from their own code into code others use. But given the
authored byabove, it looks a little more like a humble brag to me.:)
I updated benchmark